Luca Saba

Eytan Raz 
Editors

Neurovascular
Imaging
From Basics to
Advanced Concepts

1 3Reference
Neurovascular Imaging
Luca Saba • Eytan Raz
Editors

Neurovascular Imaging
From Basics to Advanced Concepts

With 799 Figures and 102 Tables

Editors
Luca Saba Eytan Raz
Department of Radiology Neurointerventional Radiology Section
AOU of Cagliari Department of Radiology
Monserrato, Cagliari NYU Langone Medical Center
Italy New York, NY
USA

ISBN 978-1-4614-9028-9 ISBN 978-1-4614-9029-6 (eBook)

ISBN 978-1-4614-9030-2 (print and electronic bundle)
DOI 10.1007/978-1-4614-9029-6
Library of Congress Control Number: 2015956531

# Springer Science+Business Media New York 2016

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or
part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way,
and transmission or information storage and retrieval, electronic adaptation, computer software, or
by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained
herein or for any errors or omissions that may have been made.

Printed on acid-free paper

This Springer imprint is published by SpringerNature

The registered company is Springer Science+Business Media LLC New York
Luca: This book is dedicated to Tiziana
Eytan: To the women of my life, my mother Daniela and
my wife Lia
Preface

Vascular neurology, vascular neurosurgery, and interventional neuroradiology

are independent fields with dedicated training programs. Neuroimaging, and in
particular what we call “neurovascular imaging,” is a unifying factor which
can be considered the intersection of these three medical specialties. With this
book, we aim to cover thoroughly the imaging techniques, potentialities, and
present and future applications as applied to the vascular diseases of the central
nervous system from the imaging point of view.
In recent years, several advances in imaging technology have dramatically
expanded the frontiers of human investigation in the neurovascular field. We
hope that this book is a timely and complete addition to the growing body of
literature on this topic.
Our aim is to cover all aspects of neurovascular disease in a comprehensive
way. The illustrations and high quality of the images will be the key to success.
This is the first collection to be published dedicated to the field of vascular
neuroimaging.

Cagliari, Italy Luca Saba

New York, USA Eytan Raz
December 2015

vii
Quotes

Do not go where the path may lead, go instead where there is no path and
leave a trail.
Ralph Waldo Emerson (1803–1882)

The foundation of every state is the education of its youth.

Diogenes (412 BC–323 BC)

Life is like riding a bicycle. To keep your balance you must keep moving.
Albert Einstein (1879–1955)

Live as if you were to die tomorrow; learn as if you were to live forever.
Mahatma Gandhi (1869–1948)

ix
Acknowledgments

It is not possible to overstate my gratitude to the many individuals who helped

to produce this book; their enthusiasm and dedication was unbelievable. In
particular, I would like to thank Professor Mario Piga for his help and advice. I
am deeply appreciative of the support offered by my colleagues at the Azienda
Ospedaliero-Universitaria di Cagliari. My gratitude goes to all authors who
have contributed to this project. Without their incredible work, this would not
have been possible.
I would like to express my appreciation to Springer for their professional-
ism in handling this project and for the patience. Most particularly, I would like
to thank Sandra Fabiani, Sunaina Dadhwal, Neha Thapa, and Gregory Baer for
their tireless dedication and advice during each stage in the production of this
book. Your help was wonderful and made producing this project an enjoyable
and worthwhile experience.
Finally, again, I would like to acknowledge Tiziana! Thank you, for
everything.

Cagliari, Italy Luca Saba

First and foremost, I want to thank Luca Saba for inviting me to help him in this
project and for his efforts in making this project going along the way. Luca,
you are one of the most hardworking persons I have ever met!
I want to thank Springer for the patience and enthusiasm that they applied to
this project, especially Neha Thapa and Sunaina Dadwhal. Thank you very
much!
I also wanted to thank all the teachers I had in my medical career so far,
especially those teachers that did that without clamor, those that taught ethics
before medicine, and those that taught me a method, a way of living, and
secondarily a way of working.

New York, USA Eytan Raz

xi
Contents

Volume 1

Part I The Basics: Embryology, Pathological Basis of

Vascular Diseases, and Imaging Techniques . . . . . . . . . . . . . . . . . . 1

1 Cerebrovascular Development and Evolution . . . . . . . . . . . . 3

Maksim Shapiro and Eytan Raz
2 Atherosclerosis: Biological and Pathological Basis . . . . . . . . 33
Michele Porcu, Eytan Raz, and Luca Saba
3 Nonatherosclerotic Vascular Disease: Biological and
Pathological Basis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Pierleone Lucatelli, Beatrice Sacconi, and Carlo Catalano
4 Essentials of Transcranial Doppler Ultrasound . . . . . . . . . . . 47
Jonathan D. Kirsch
5 DSA Imaging Protocol and Technique . . . . . . . . . . . . . . . . . . 67
Eytan Raz

Part II Carotid and Vertebral Artery: Anatomy, Diseases,

and Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

6 Anatomy of the Neck Arteries . . . . . . . . . . . . . . . . . . . . . . . . . 87

Nasim Sheikh-Bahaei, Tomasz Matys, and Jonathan H. Gillard
7 Imaging of Carotid Atherosclerosis . . . . . . . . . . . . . . . . . . . . 95
Jie Sun, Niranjan Balu, and Chun Yuan
8 Carotid Artery Dissection . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Olivier Naggara, Myriam Edjlali-Goujon, Eric Bodiguel,
Marie Pierre Gobin-Metteil, Denis Trystram, Christine
Rodriguez-Regent, Jean-Louis Mas, Jean Francois Meder, and
Catherine Oppenheim
9 Nonatherosclerotic Nondissection Diseases of
Carotid Artery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
Santosh Kumar Kannath and T.R. Kapilamoorthy

xiii
xiv Contents

10 Imaging of the Pathology of the Vertebral Arteries . . . . . . . . 163

David Chiao and Max Wintermark

11 Carotid Artery Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191

Roberto Montisci and Luca Saba

Part III The Basics of Intracranial Arterial Circulation . . . . . . . . 203

12 Intracranial Atherosclerosis . . . . . . . . . . . . . . . . . . . . . . . . . . 205

Xinyi Leng and David S. Liebeskind

13 Imaging of Spontaneous Nontraumatic Intracerebral

Hemorrhage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Stephen Quinet and Patrick Turski

14 ICA–ECA Collaterals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293

Maxim Mokin and Adnan H. Siddiqui

15 Imaging Biomarkers of Angiogenesis and the

Microvascular Environment in Cerebral Tumors . . . . . . . . . 303
Alan Jackson, Ibrahim Djoukhadar, and David J. Coope

Part IV Stroke Imaging .................................... 327

16 Patterns of Ischemic Stroke: From Lacunar to Territorial

to Multiple Embolic to Watershed Hypotensive . . . . . . . . . . 329
Jessica Rotman and Robert Zimmerman

17 Territorial Strokes as a Tool to Learn Vascular

Territories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
Behroze Adi Vachha and Pamela Whitney Schaefer

18 Hemorrhagic Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383

Julius Griauzde, Elliot Dickerson, and Joseph J. Gemmete

19 Medical Therapy of Acute Stroke . . . . . . . . . . . . . . . . . . . . . . 413

Keith G. DeSousa and Albert S. Favate

20 Endovascular Treatment of Stroke . . . . . . . . . . . . . . . . . . . . . 425

Paolo Machi and Paolo Garofalo

21 Sickle Cell Disease and Stroke . . . . . . . . . . . . . . . . . . . . . . . . 439

Akifumi Fujita, Chie Asai, Yu-Ming Chang, Nadja Kadom,
Martin H. Steinberg, Naoko Saito, and Osamu Sakai

22 Hypoxic–Ischemic Encephalopathy (Preterm, Term,

and Adult) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469
Mauricio Castillo and Francisco Chiang
Contents xv

Part V Aneurysms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495

23 Aneurysmal Subarachnoid Hemorrhage . . . . . . . . . . . . . . . . 497

Frédéric Clarençon, Nader-Antoine Sourour, Vincent Degos,
Aurélien Nouet, Federico Di Maria, Eimad Shotar,
Joseph Gabrieli, Lise Le Jean, and Jacques Chiras
24 Blister Aneurysms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521
James V. Byrne and Svein Harald Mørkve
25 Surgical Treatment of Aneurysms . . . . . . . . . . . . . . . . . . . . . 535
Saul F. Morales-Valero, Shanna Fang, and Giuseppe Lanzino
26 Endovascular Treatment of Brain Aneurysms . . . . . . . . . . . 551
Luca Quilici and Edoardo Boccardi
27 Postop Evaluation of Aneurysms (Including Vasospasm) . . . 583
Michael Mayich, Brian P. Walcott, Christopher
J. Stapleton, and Daniel Thomas Ginat

Volume 2

Part VI Other CNS Vascular Diseases . . . . . . . . . . . . . . . . . . . . . . . . 603

28 Brain Arteriovenous Malformations . . . . . . . . . . . . . . . . . . . 605

Daniel Sahlein and Nathan Manning
29 Dural AVF and Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . 641
Pierre Guedin
30 Carotid Cavernous Fistula . . . . . . . . . . . . . . . . . . . . . . . . . . . 657
Timothy R. Miller, Ravishankar Shivashankar,
Gaurav Jindal, and Dheeraj Gandhi
31 Imaging of the Neurovasculitides . . . . . . . . . . . . . . . . . . . . . . 683
Adam J. Davis
32 Miscellaneous Vascular Malformations (Cavernous
Malformations, Developmental Venous Anomaly,
Capillary Telangiectasia, Sinus Pericranii) . . . . . . . . . . . . . . 719
Luiz Celso Hygino da Cruz and Cintia Elias Pires
33 Miscellaneous Vascular Diseases . . . . . . . . . . . . . . . . . . . . . . . 751
Matylda H. Machnowska and Richard I. Aviv
34 Vascular Loop Syndromes . . . . . . . . . . . . . . . . . . . . . . . . . . . 781
Divyata R. Hingwala and Kesavadas Chandrasekharan
35 Migraine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 791
Mark Kruit
xvi Contents

36 Moyamoya Disease (Spontaneous Occlusion

of the Circle of Willis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 817
Akira Yamamoto, Tomohisa Okada, and
Jun C. Takahashi
37 Brain Death Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865
M. Sawicki, Joanna Wojczal, Bozena Birkenfeld, and
Lech Cyrylowski

Part VII Veins Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 897

38 Jugular Vein Thrombosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 899

Nicholas A. Koontz, Richard H. Wiggins III, and
Lubdha M. Shah
39 Imaging of Cerebral Venous and Sinus Thrombosis . . . . . . . 935
Jennifer Linn
40 The Role of Diagnostic Imaging Techniques for Detection
of Extracranial Venous System Abnormalities Associated
with Central Nervous System Disorders . . . . . . . . . . . . . . . . 953
Kresimir Dolic and Robert Zivadinov

Part VIII Pediatrics ........................................ 981

41 Pediatric Vascular Malformations . . . . . . . . . . . . . . . . . . . . . 983

Sachin K. Pandey and Darren B. Orbach
42 Pediatric Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1009
Daniel Tibussek, Gabrielle deVeber, and Manohar Shroff
43 Congenital Vascular Syndromes and Diseases . . . . . . . . . . . . 1041
Sarah Milla, Jennifer Vaughn, and Nilesh K. Desai

Part IX Spine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1061

44 Spinal Vascular Imaging: Technique . . . . . . . . . . . . . . . . . . . 1063

Maria Isabel Vargas, Fabrice Bing, Joanna Gariani, and
Jean-Louis Dietemann
45 Spinal Vascular Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1095
Tibor Becske and Eytan Raz
46 Spinal Vascular Malformations and Treatment . . . . . . . . . . . 1105
Srinivasan Paramasivam
47 Spinal Cord Infarction and Differential Diagnosis . . . . . . . . 1125
Srikanth R. Boddu, Alessandro Cianfoni, Kyung-Wha Kim,
Mohammad Amin Banihashemi, Emanuele Pravatà,
Y. Pierre Gobin, and Athos Patsalides
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1185
About the Editors

Prof. Luca Saba received his M.D. from the University of Cagliari, Italy, in
2002. Today, he works in the A.O.U. of Cagliari. Professor Saba’s research
fields are focused on multi-detector-row computed tomography, magnetic
resonance, ultrasound, neuroradiology, and diagnostic in vascular sciences.
His works, as lead author, are more than 170 publications in high-impact-
factor, peer-reviewed journals such as American Journal of Neuroradiology,
Atherosclerosis, European Radiology, European Journal of Radiology, Acta
Radiologica, Cardiovascular and Interventional Radiology, Journal of
Computer Assisted Tomography, American Journal of Roentgenology, Neuro-
radiology, Clinical Radiology, Journal of Cardiovascular Surgery, Cerebro-
vascular Diseases, Brain Pathology, Medical Physics, and Atherosclerosis.
He is a well-known speaker and has spoken over 45 times at national and
international level conferences.
Dr. Saba has won 15 scientific and extracurricular awards during his career.
Dr. Saba has presented more than 500 papers and posters in national and
international congresses (RSNA, ESGAR, ECR, ISR, AOCR, AINR, JRS,
SIRM, AINR). He has written 21 book chapters and is Editor of 10 books in
the field of computed tomography, cardiovascular, plastic surgery, gynecolog-
ical imaging, and neurodegenerative imaging.
He is a member of the Italian Society of Radiology (SIRM), European
Society of Radiology (ESR), Radiological Society of North America (RSNA),
American Roentgen Ray Society (ARRS), and European Society of Neurora-
diology (ESNR) and serves as reviewer of 40 scientific journals.

xvii
xviii About the Editors

Eytan Raz currently works at NYU Langone Medical Center in New York
City, where he is an Assistant Professor working for the Neuroradiology and
Interventional Neuroradiology Sections. He graduated from Sapienza
University of Rome School of Medicine in 2006, where he also obtained his
Radiology Specialty in 2011. He completed his fellowships in Diagnostic
Neuroradiology and Interventional Neuroradiology at NYU Langone
Medical Center. He is the author of more than 60 publications in
high-impact-factor, peer-reviewed journals such as Radiology, AJNR, AJR,
JNIS, Neurology, and Brain Pathology. His fields of expertise include
neurovascular diseases, traumatic brain injury, and multiple sclerosis. In
2012, he was appointed Editor for the Fellows’ Portal of AJNR. He is a
reviewer for more than 10 journals and has presented more than 70 abstracts
at international conferences. He is a member of the Radiological Society of
North America (RSNA), the American Society of Neuroradiology (ASNR),
and the Italian Society of Neuroradiology (AINR).
Contributors

Chie Asai Departments of Radiology, Boston University, Boston Medical

Center, Boston, MA, USA

Richard I. Aviv Division of Neuroradiology, Department of Medical Imag-

ing, School of Medicine, Sunnybrook Health Sciences Centre, University of
Toronto, Toronto, ON, Canada

Niranjan Balu Department of Radiology, University of Washington, Seattle,

WA, USA

Mohammad Amin Banihashemi Department of Neurological Surgery,

Weill Cornell Medical Center, New York, NY, USA

Tibor Becske Neurointerventional Radiology Section, Department of

Radiology, NYU Langone Medical Center, New York, NY, USA

Fabrice Bing Department of Radiology, University Hospital of Strasbourg,

Strasbourg, France

Bozena Birkenfeld Department of Nuclear Medicine, Pomeranian Medical

University, Szczecin, Poland

Edoardo Boccardi Ospedale Niguarda – Ca’ Granda, Milan, Italy

Srikanth R. Boddu Department of Neurological Surgery, Weill Cornell

Medical Center, New York, NY, USA

Eric Bodiguel Service de Neurologie, Faculté de Médecine, Centre

Hospitalier Sainte-Anne, Université Paris Descartes, Sorbonne Paris Cité,
Paris, France

James V. Byrne Nuffield Department of Surgical Sciences, Oxford Univer-

sity, John Radcliffe Hospital, Oxford, UK

Mauricio Castillo Division of Neuroradiology, Department of Radiology,

University of North Carolina, Chapel Hill, NC, USA

Carlo Catalano Department of Radiological, Oncological and Anatomo-

Pathological Sciences, Sapienza University of Rome, Rome, Italy

xix
xx Contributors

Kesavadas Chandrasekharan NM Medical Centre, Mumbai, Maharashtra,

India
Sree Chitra Tirunal Institute for Medical Sciences and Technology, Trivan-
drum, Kerela, India
Yu-Ming Chang Departments of Radiology, Boston University, Boston
Medical Center, Boston, MA, USA
Francisco Chiang Division of Neuroradiology, Department of Radiology,
University of North Carolina, Chapel Hill, NC, USA
David Chiao Department of Radiology, School of Medicine, University of
Virginia, Charlottesville, VA, USA
Jacques Chiras Department of Interventional Neuroradiology,
Pitié-Salpêtrière Hospital, Paris VI University, Paris, France
Alessandro Cianfoni Department of Radiology, Medical University of South
Carolina, Charleston, SC, USA
Frédéric Clarençon Department of Interventional Neuroradiology,
Pitié-Salpêtrière Hospital, Paris VI University, Paris, France
David J. Coope Wolfson Molecular Imaging Centre, University of Manchester,
Withington, Manchester, UK
Lech Cyrylowski CT and MRI Clinic, “Nasz Doktor” Health Care Institu-
tion, Szczecin, Poland
MRI Clinic, “Euromedic” Health Care Institution, Szczecin, Poland
Luiz Celso Hygino da Cruz Department of Radiology, Clinics CDPI/DASA
and IRM, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil
Adam J. Davis Neuroradiology, Department of Radiology, NYU Langone
Medical Center, New York, NY, USA
Vincent Degos Department of Anesthesia and Critical Care, Neuro-intensive
Care Unit, Pitié- Salpêtrière Hospital, Paris VI University, Paris, France
Nilesh K. Desai Department of Radiology, Texas Children’s Hospital, Baylor
College of Medicine, Houston, TX, USA
Keith G. DeSousa Department of Neurology, NYU Langone Medical Cen-
ter, New York, NY, USA
Vascular Neurology, University of Miami, Miami, FL, USA
Gabrielle deVeber Children’s Stroke Program, Division of Neurology,
The Hospital for Sick Children, University of Toronto, Toronto, ON, Canada
Federico Di Maria Department of Interventional Neuroradiology, Pitié-
Salpêtrière Hospital, Paris VI University, Paris, France
Elliot Dickerson Department of Radiology, Radiology Resident, University
of Michigan, Ann Arbor, MI, USA
Contributors xxi

Jean-Louis Dietemann Department of Radiology, University Hospital of

Strasbourg, Strasbourg, France
Ibrahim Djoukhadar Wolfson Molecular Imaging Centre, University of
Manchester, Withington, Manchester, UK
Kresimir Dolic Buffalo Neuroimaging Analysis Center, Department of
Neurology, University at Buffalo, State University of New York, Buffalo,
NY, USA
Clinical Department of Diagnostic and Interventional Radiology, Clinical
Hospital Center Split, University of Split, Split, Croatia
Myriam Edjlali-Goujon Service de Neuroradiologie, Centre Hospitalier
Sainte-Anne, INSERM UMR894, Université Paris Descartes, Sorbonne
Paris Cité, Paris, France
Shanna Fang Mayo Medical School, Mayo Clinic, Rochester, MN, USA
Albert S. Favate NYU Langone Medical Center, New York, NY, USA
Akifumi Fujita Departments of Radiology, Boston University, Boston
Medical Center, Boston, MA, USA
Joseph Gabrieli Department of Interventional Neuroradiology,
Pitié-Salpêtrière Hospital, Paris VI University, Paris, France
Dheeraj Gandhi Department of Diagnostic Radiology, Neuroradiology,
University of Maryland Medical Center, Baltimore, MD, USA
Joanna Gariani Department of Radiology, Geneva University Hospitals,
Geneva, Switzerland
Paolo Garofalo Medicine School, University of Parma, Parma, Italy
Joseph J. Gemmete Department of Radiology, Neurosurgery, and Otolaryn-
gology, University of Michigan, Ann Arbor, MI, USA
Jonathan H. Gillard Department of Radiology, School of Clinical Medicine,
University of Cambridge, Cambridge Biomedical Campus, Cambridge, UK
Daniel Thomas Ginat Department of Radiology, University of Chicago,
Chicago, IL, USA
Y. Pierre Gobin Department of Neurological Surgery, Weill Cornell Medical
Center, New York, NY, USA
Marie Pierre Gobin-Metteil Service de Neuroradiologie, Centre Hospitalier
Sainte-Anne, INSERM UMR894, Université Paris Descartes, Sorbonne Paris
Cité, Paris, France
Julius Griauzde Department of Radiology, Radiology Resident, University
of Michigan, Ann Arbor, MI, USA
Pierre Guedin Department of Diagnostic and Therapeutic Neuroradiology,
Hôpital Foch, Suresnes, France
Divyata R. Hingwala NM Medical Centre, Mumbai, Maharashtra, India
xxii Contributors

Alan Jackson Wolfson Molecular Imaging Centre, University of Manchester,

Withington, Manchester, UK

Gaurav Jindal Department of Diagnostic Radiology, Neuroradiology,

University of Maryland Medical Center, Baltimore, MD, USA

Nadja Kadom Departments of Radiology, Boston University, Boston Med-

ical Center, Boston, MA, USA

Santosh Kumar Kannath Neurointervention Center, Department of Imag-

ing Sciences and Interventional Radiology, Neurointervention Center, Sree
Chitra Institute of Medical Sciences and Technology, Trivandrum, Kerala,
India

T. R. Kapilamoorthy Department of Imaging Sciences and Interventional

Radiology, Sree Chitra Institute of Medical Sciences and Technology, Trivan-
drum, Kerala, India

Kyung-Wha Kim New York Presbyterian Hospital, Weill Cornell Medical

College of Cornell University, New York, NY, USA

Jonathan D. Kirsch Department of Diagnostic Radiology, Yale University

School of Medicine, New Haven, CT, USA

Nicholas A. Koontz Department of Radiology, University of Utah Health

Sciences Center, Salt Lake City, UT, USA

Mark Kruit Department of Radiology, Leiden University Medical Center,

Leiden, The Netherlands

Giuseppe Lanzino Department of Neurologic Surgery, Mayo Clinic, Roch-

ester, MN, USA

Lise Le Jean Department of Anesthesia and Critical Care, Neuro-intensive

Care Unit, Pitié- Salpêtrière Hospital, Paris VI University, Paris, France

Xinyi Leng Department of Medicine and Therapeutics, The Chinese Univer-

sity of Hong Kong, Prince of Wales Hospital, Hong Kong, SAR, China

David S. Liebeskind UCLA Department of Neurology, UCLA, Los Angeles,

CA, USA

Jennifer Linn Institute of Neuroradiology, University Hospital Carl Gustav

Carus, Dresden, Germany

Pierleone Lucatelli Department of Radiological, Oncological and Anatomo-

Pathological Sciences, Sapienza University of Rome, Rome, Italy

Paolo Machi Department of Neuroradiology Unit, CHU Gui de Chauliac,

Montpellier University Hospital, Montpellier, France

Matylda H. Machnowska Division of Neuroradiology, Department of

Medical Imaging, School of Medicine, Sunnybrook Health Sciences Centre,
University of Toronto, Toronto, ON, Canada
Contributors xxiii

Nathan Manning Department of Neurological Surgery, Columbia Univer-

sity Medical Centre, New York, NY, USA
Jean-Louis Mas Service de Neurologie, Faculté de Médecine, Centre
Hospitalier Sainte-Anne, Université Paris Descartes, Sorbonne Paris Cité,
Paris, France
Tomasz Matys Department of Radiology, School of Clinical Medicine,
University of Cambridge, Cambridge Biomedical Campus, Cambridge, UK
Michael Mayich Department of Radiology, University of Chicago, Chicago,
IL, USA
Jean Francois Meder Service de Neuroradiologie, Centre Hospitalier
Sainte-Anne, INSERM UMR894, Université Paris Descartes, Sorbonne
Paris Cité, Paris, France
Sarah Milla Department of Radiology, Emory University School of Medi-
cine Egleston Hospital, Children’s Healthcare of Atlanta, Atlanta, GA, USA
Timothy R. Miller Department of Diagnostic Radiology, Neuroradiology,
University of Maryland Medical Center, Baltimore, MD, USA
Maxim Mokin Department of Neurosurgery, School of Medicine and
Biomedical Sciences, University at Buffalo, State University of New York,
Buffalo, NY, USA
Department of Neurosurgery, Gates Vascular Institute/Kaleida Health,
Buffalo, NY, USA
Roberto Montisci Department of Vascular Surgery, AOU of Cagliari,
Monserrato, Cagliari, Italy
Saul F. Morales-Valero Department of Neurologic Surgery, Mayo Clinic,
Rochester, MN, USA
Svein Harald Mørkve Department of Neurosurgery, Haukeland University
Hospital, Bergen, Norway
Olivier Naggara Service de Neuroradiologie, Centre Hospitalier Sainte-
Anne, INSERM UMR894, Université Paris Descartes, Sorbonne Paris Cité,
Paris, France
Service de Radiologie pédiatrique, Faculté de Médecine, Hôpital Necker
Enfants Malades, Université Paris Descartes, Sorbonne Paris Cité, Paris,
France
Aurélien Nouet Department of Neurosurgery, Pitié-Salpêtrière Hospital,
Paris VI University, Paris, France
Tomohisa Okada Department of Diagnostic Imaging and Nuclear Medicine,
Graduate School of Medicine, Kyoto University, Sakyo-ku, Kyoto, Japan
Catherine Oppenheim Service de Neuroradiologie, Centre Hospitalier
Sainte-Anne, INSERM UMR894, Université Paris Descartes, Sorbonne
Paris Cité, Paris, France
xxiv Contributors

Darren B. Orbach Neurointerventional Radiology, Boston Children’s

Hospital, Harvard Medical School, Boston, MA, USA

Sachin K. Pandey Neurointerventional Radiology, Brigham and Women’s

Hospital, Boston, MA, USA
Srinivasan Paramasivam Mount Sinai Roosevelt Hospital, Hyman New-
man Institute for Neurology and Neurosurgery, Centre for Endovascular
Surgery, New York, NY, USA

Athos Patsalides Department of Neurological Surgery, Weill Cornell Medi-

cal Center, New York, NY, USA

Cintia Elias Pires Department of Radiology, Clinics CDPI/DASA, National

Institute of Cardiology, Federal University of Rio de Janeiro, Rio de Janeiro,
Brazil
Michele Porcu Department of Radiology, AOU of Cagliari, Monserrato,
Cagliari, Italy

Emanuele Pravatà Neuroradiology Department, Neurocenter of Italian

Switzerland, Lugano, Switzerland
Luca Quilici Ospedale Niguarda – Ca’ Granda, Milan, Italy

Stephen Quinet Department of Radiology, School of Medicine and

Public Health, Clinical Science Center, University of Wisconsin, Madison,
WI, USA

Eytan Raz Neurointerventional Radiology Section, Department of

Radiology, NYU Langone Medical Center, New York, NY, USA

Christine Rodriguez-Regent Service de Neuroradiologie, Centre

Hospitalier Sainte-Anne, INSERM UMR894, Université Paris Descartes,
Sorbonne Paris Cité, Paris, France

Jessica Rotman PGY-3 Diagnostic Radiology Resident, New York Presby-

terian Hospital, Weill Cornell Medical Center, New York, NY, USA

Luca Saba Department of Radiology, AOU of Cagliari, Monserrato,

Cagliari, Italy

Beatrice Sacconi Department of Radiological, Oncological and Anatomo-

Pathological Sciences, Sapienza University of Rome, Rome, Italy

Daniel Sahlein Goodman Campbell Brain and Spine, Indianapolis, IN, USA

Naoko Saito Department of Diagnostic Radiology, Saitama Medical Univer-

sity International Medical Center, Saitama, Japan

Osamu Sakai Departments of Radiology, Boston University, Boston Medi-

cal Center, Boston, MA, USA
Department of Otolaryngology – Head and Neck Surgery, Boston University,
Boston Medical Center, Boston, MA, USA
Contributors xxv

Department of Radiation Oncology, Boston University, Boston Medical Cen-

ter, Boston, MA, USA
M. Sawicki Department of Diagnostic Imaging and Interventional Radiol-
ogy, Pomeranian Medical University, Szczecin, Poland
Pamela Whitney Schaefer Neuroradiology, Massachusetts General Hospital,
Boston, MA, USA
Lubdha M. Shah Department of Radiology, University of Utah Health
Sciences Center, Salt Lake City, UT, USA
Maksim Shapiro Neurointerventional Radiology Section, Department of
Radiology, NYU Langone Medical Center, New York, NY, USA
Nasim Sheikh-Bahaei Department of Radiology, School of Clinical
Medicine, University of Cambridge, Cambridge Biomedical Campus,
Cambridge, UK
Ravishankar Shivashankar Department of Diagnostic Radiology, Neurora-
diology, University of Maryland Medical Center, Baltimore, MD, USA
Eimad Shotar Department of Interventional Neuroradiology, Pitié-
Salpêtrière Hospital, Paris VI University, Paris, France
Manohar Shroff Department of Diagnostic Imaging, The Hospital for Sick
Children, University of Toronto, Toronto, ON, Canada
Adnan H. Siddiqui Department of Neurosurgery, School of Medicine and
Biomedical Sciences, University at Buffalo, State University of New York,
Buffalo, NY, USA
Department of Neurosurgery, Gates Vascular Institute/Kaleida Health,
Buffalo, NY, USA
Department of Radiology, School of Medicine and Biomedical Sciences,
University at Buffalo, State University of New York, Buffalo, NY, USA
Toshiba Stroke and Vascular Research Center, University at Buffalo, State
University of New York, Buffalo, NY, USA
Jacobs Institute, Buffalo, NY, USA
Nader-Antoine Sourour Department of Interventional Neuroradiology,
Pitié-Salpêtrière Hospital, Paris VI University, Paris, France
Christopher J. Stapleton Department of Neurosurgery, Massachusetts
General Hospital, Boston, MA, USA
Martin H. Steinberg Departments of Medicine, Pediatrics, and Pathology
and Laboratory Medicine, Boston University, Boston Medical Center, Boston,
MA, USA
Jie Sun Department of Radiology, University of Washington, Seattle,
WA, USA
Jun C. Takahashi Department of Neurosurgery, National Cerebral and
Cardiovascular Center, Suita City, Osaka, Japan
xxvi Contributors

Daniel Tibussek Children’s Stroke Program, Division of Neurology, The

Hospital for Sick Children, University of Toronto, Toronto, ON, Canada
Department of General Pediatrics, Neonatalogy and Pediatric Cardiology,
University Children’s Hospital, Heinrich-Heine University, D€usseldorf,
Germany
Denis Trystram Service de Neuroradiologie, Centre Hospitalier Sainte-
Anne, INSERM UMR894, Université Paris Descartes, Sorbonne Paris Cité,
Paris, France
Patrick Turski Department of Radiology, School of Medicine and Public
Health, Clinical Science Center, University of Wisconsin, Madison, WI, USA
Behroze Adi Vachha Neuroradiology, Massachusetts General Hospital,
Boston, MA, USA
Maria Isabel Vargas Department of Neuroradiology, Geneva University
Hospitals, Geneva, Switzerland
Jennifer Vaughn Department of Radiology, Boston Children’s Hospital,
Boston, MA, USA
Brian P. Walcott Department of Neurosurgery, Massachusetts General
Hospital, Boston, MA, USA
Richard H. Wiggins III Department of Radiology, University of Utah
Health Sciences Center, Salt Lake City, UT, USA
Max Wintermark Department of Radiology, School of Medicine, Univer-
sity of Virginia, Charlottesville, VA, USA
Joanna Wojczal Department of Neurology, Medical University of Lublin,
Lublin, Poland
Akira Yamamoto Department of Diagnostic Imaging and Nuclear Medicine,
Graduate School of Medicine, Kyoto University, Sakyo-ku, Kyoto, Japan
Chun Yuan Department of Radiology, University of Washington, Seattle,
WA, USA
Robert Zimmerman Diagnostic Radiology Department, New York Presby-
terian Hospital, Weill Cornell Medical Center, New York, NY, USA
Robert Zivadinov Buffalo Neuroimaging Analysis Center, Department
of Neurology, University at Buffalo, State University of New York, Buffalo,
NY, USA
MRI Clinical Translational Research Center, University at Buffalo, State
University of New York, Buffalo, NY, USA
 Part I
The Basics: Embryology, Pathological
Basis of Vascular Diseases, and Imaging
Techniques
 Cerebrovascular Development and
Evolution 1
Maksim Shapiro and Eytan Raz

Contents Abstract
Vascular Neurophylogeny . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 The comprehensive appreciation of the embryo-
The Spinal Cord . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 logical events is paramount to develop a good
The Brain: From Fishes to Humans . . . . . . . . . . . . . . . . . . 6 understanding of the cerebrovasculature in all its
Vascular Neuroembryology . . . . . . . . . . . . . . . . . . . . . . . . . 12 aspects, from the anatomy to the physiology and
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 hence the pathology. Vascularization of the brain
Stage I (Prechoroidal Stage): and spinal cord has certain fundamental similar-
Approximately 4 Weeks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Stage II (Choroidal Stage):
ities. The evolution of brain circulation attests to
Approximately 5 Weeks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 progressive recruitment of already existing vas-
Stage III: 6 Weeks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 cular networks to supply emergent cortical terri-
Stage IV: Approximately 7 Weeks . . . . . . . . . . . . . . . . . . . . 16 tories, rather than development of de novo
Stage V: Approximately 9 Weeks . . . . . . . . . . . . . . . . . . . . . 17
Stage VI: Approximately Adult Pattern . . . . . . . . . . . . . . 17
arterial solutions. This principle can be applied
to many concepts in evolutionary developmental
Practical Aspects of Neurophylogeny biology. And, since, in many aspects, “ontogeny
and Neuroembryology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Fetal Posterior Cerebral Artery . . . . . . . . . . . . . . . . . . . . . . . 18 recapitulates phylogeny” to quote Ernst Haeckel,
Accessory MCA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 father of the embryological parallelism law, the
Dominant Anterior Choroidal Artery . . . . . . . . . . . . . . . . . 19 phylogeny of the brain and spinal cord is a key
Persistent Trigeminal Artery . . . . . . . . . . . . . . . . . . . . . . . . . . 19 concept to understand in order to develop a mind
Persistent Hypoglossal Artery . . . . . . . . . . . . . . . . . . . . . . . . 19
Proatlantal Artery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 framework which becomes then useful for the
Variations in Fusion Patterns of the Basilar Artery . . . 20 approach of embryology.
Duplications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Anterior Cerebral Artery Variations . . . . . . . . . . . . . . . . . . 25 Part of the text, the concept, and the figures in this
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 chapter are used with permission from Maksim
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Shapiro, from the website www.neuroangio.org.

Vascular Neurophylogeny

M. Shapiro (*) • E. Raz The Spinal Cord

Neurointerventional Radiology Section, Department of
Radiology, NYU Langone Medical Center, New York, The neural tube and the spinal cord can be thought
NY, USA
e-mail: Maksim.Shapiro@nyumc.org; of, simplistically, as a cylinder (Fig. 1). After
eytan.raz@gmail.com; eytan.raz@nyumc.org closure of the neural tube, nutritional support
# Springer Science+Business Media New York 2016 3
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_15
4 M. Shapiro and E. Raz

can be provided by simple diffusion, but, at a coming from the two dorsal aortae (Fig. 2). These
certain point of body growth, dedicated vascular vessels are also called metameric: in biology,
supply to the spinal cord and adjacent tissues is metamerism is the phenomenon of having a linear
needed and is established using segmental vessels series of similar body segments; each segment
may be called somite or metamer. Each
metameric vessel supports its own block of tissue,
including the endodermal, ectodermal, and meso-
dermal tissue (Fig. 3). As diagrammatically shown
in Fig. 4, between segmental vessels longitudinal
anastomoses extending craniocaudally along the
length of the cord are then formed (Fig. 4),
establishing a primordial matrix of the adult vascu-
Spinal Cord lature. Once the longitudinal connection between
the metamers is established, there is no need for a
segmental vessel at each single level, and most
segmental vessels shrink to supply only the region
of the nerve root of the same level (Figs. 5 and 6).
The spinal cord is now supplied by an anterior
spinal artery and posteriorly by two spinal vessels
that will become the posterior spinal arteries. The
Fig. 1
segmental vessels that supply flow to the anterior
spinal artery are called radiculomedullary arteries,
while the ones that supply only the nerve root are
called radicular branches. Sometimes, one segmen-
Dorsal tal vessel supplies two nerve roots via longitudinal
Aorta anastomoses (Fig. 7). As the spinal cord continues
Folded Neural Tube to enlarge, sulco-commissural branches are
formed, which can be matched with the brain per-
forator arteries (Fig. 8).
In Figs. 9 and 10, the spinal angiographic cor-
relation can be visualized in vivo with subtracted
and unsubtracted views which demonstrate a
radiculomedullary vessel giving rise to an anterior
Fig. 2 spinal artery.

Fig. 3
To ribs,
muscle,
Dorsal etc.
Aorta

Segmental (a.k.a Metameric)

Vessels
1 Cerebrovascular Development and Evolution 5

Fig. 4
Devloping To ribs,
Longitudinal muscle,
Anastomoses etc.

Segmental (a.k.a. Metamaric)

Vessels

Fig. 5
To ribs,
muscle,
etc.
Vertebral
artery-like
vessel

Future Anterior Spinal Artery

Segmental Vessels

Fig. 6

Longitudinal vessel, future

Anterior Spinal artery

Surface pial vessel

Most segmental vessels regress and remain

purely radicular; several dominant
radiculomedullary ones remain
6 M. Shapiro and E. Raz

Fig. 7
Obviously, only To ribs,
Radiculomedullary
one aorta muscle,
artery
remains in the Anterior etc.
adult, from Spinal
which left and Artery
right segmental
vessels emerge.
Two are shown
here for
diagrammatic
purpose.

Radicular artery
Dominant radicular artery
supplying two levels

variations. The fact that newly formed vessels

Anterior are the exception, while existing pathways are
Spinal usually utilized and enlarged to accommodate
Artery
tissue needs, was recognized since the early twen-
tieth century.

Sulco-comissural Arteries
Fig. 8
The Brain: From Fishes to Humans
When thinking about the brain neuroembryology
and neurophylogeny, it is useful to conceptualize
its structure as a complex spinal cord, supplied by
segmental vessels. Once, together with higher
organism complexity, more complicated anatom-
ical structures are added on, we will see a more
sophisticated vasculature pattern that is better
understood keeping in mind the very simple
basic segmental arrangement of the spinal cord.
We can think of the developing brain as a progres-
sively enlarging segment of the cord, requiring
larger surface and deep perforator vessels as key
to understanding evolution and vascular
Fishes
The earliest vertebrates still existing today are the
jawless fishes, cyclostomes, and hagfish. The
brains of these animals possess all of the features
of vertebrates, including five divisions of the
brain: medulla, pons, midbrain, diencephalon,
and telencephalon (Fig. 11). Even though some
earlier fishes lack a cerebellum (lampreys and
hagfish), its presence becomes a regular factor in
later fish evolution. Other than a primitive cere-
bellum, in fishes we see a small but effective
olfactory lobe and a primitive three-layer cortex,
the forerunner of the hippocampus. The “higher
center” of movement is controlled by what will
eventually become the present-day basal ganglia
(Fig. 11). The spinal cord will remain essentially
unchanged, supplied by segmental radiculo-
medullary arteries. The vertebral arteries form as
a result of longitudinal anastomoses between
metameric branches of the cervical somites
(Fig. 12).
In the brain two large longitudinal vessels – the
carotids – travel up to the neck and enter the
cranial cavity, in a way similar to that of segmental
radiculomedullary vessels in the cord (Fig. 12).
Inside the skull, the internal carotid artery gives
rise to paired longitudinal arteries: one that goes
1 Cerebrovascular Development and Evolution 7

Fig. 9 Frontal view from a spinal angiogram, subtracted spinal artery (black arrow). In this particular case, the
and unsubtracted views, demonstrating a radiculo- radiculomedullary vessel arises from the T9 intercostal
medullary vessel (blue arrow) giving rise to an anterior artery (white arrow)

Fig. 10 Frontal view from

a spinal angiogram,
subtracted view. Anterior
spinal artery (red arrows)
originating from a
radiculomedullary branch
(orange arrow) arising from
the costocervical trunk
(purple arrow). This
conformation attests to the
metameric nature of the
system, where the spinal
artery can arise from
longitudinal vessels other
than the vertebral artery. A
tumor blush (yellow arrow)
from T1 hepatoma
metastasis is present, along
with several coil masses
(black arrows) deposited
after embolization of the
left-sided tumor soft tissue
component
8 M. Shapiro and E. Raz

up, the “cranial ramus,” and one extending cau- the forerunner of the PComm and of a small part
dally, the “caudal ramus” (Fig. 12). The “cranial of the basilar artery, which, at this point, is already
ramus” is the forerunner of the anterior cerebral fused at midline. At this point of development, a
and middle cerebral arteries. The caudal ramus is functionally significant anastomosis between the
caudal ramus and the posterior circulation system
is not activated yet, and the entire intracranial
CNS (including the brainstem and the cerebellum)
is supplied by the carotids.
Olfactory
Lobe
The cranial ramus has two notable branches:
the medial olfactory branch, which is the future
Primitive ACA, and the lateral olfactory branch, which is
Cortex the forerunner of the Heubner and of the anterior
(mostly choroidal artery. The caudal ramus gives off a
forerunner of tectal artery which supplies the posterosuperior
hippocampus) aspect of the midbrain; from this vessel comes
off a cerebellar artery supplying the primitive
cerebellum – forerunner of the superior cerebellar
Forebrain artery.
Segmental anastomoses also exist, usually in
Midbrain very small form, between the carotid and vertebral
systems. In the human embryo, these pathways
are quite functional at early stages while they
Cerebellum regress as the vertebrobasilar system. Occasional
persistence of these channels leads to the presence
of anomalous vessels such as the persistent tri-
Cord geminal artery, the persistent hypoglossal artery,
or the proatlantal arteries, depending on the seg-
ment that failed to develop.
Fig. 11

Fig. 12 Medical Olfactory

Lateral Olfactory

Cranial Ramus

Caudal Ramus
Carotid Tectal
Cerebellar
Fused Caudal Ramus (Basilar)

Segmental Inferior Cerebellar (PICA)

Carotido- Vertebral
vertebral
Anastomoses
1 Cerebrovascular Development and Evolution 9

Fishes Amphibians
Medial Olfactory
Lateral Olfactory
Lateral Striate
Posterior Telencephalic
(future Ant. Choroidal)
Cranial Ramus
Caudal Ramus
Posterior Choroidal
Tectal
Cerebellar
Fused Caudal Ramus (Basilar)
Inferior Cerebellar (PICA)

Vertebral
Segmental Carotido -
vertebral Anastomoses

Carotid

Fig. 13

Amphibians cranial ramus arise multiple “perforators” supply-

The changes seen with the shift to amphibians are
schematized in Fig. 13. The brain continues to
grow, with a larger primitive hippocampus and
larger basal ganglia. This larger size of the
brain leads to enlargement of the supplying,
already existing branches and to the fusion of
smaller feeders into larger arteries (Fig. 13).
The lateral olfactory branch has enlarged with
additional named vessels and now consists of
two main branches – the lateral striate artery,
forerunner of the Heubner, of the Middle cerebral
artery (MCA), and of the MCA perforators, and
the “posterior telencephalic branch,” forerunner
of the anterior choroidal artery (Fig. 13). On the
caudal ramus, the tectal/cerebellar complex
assumes supply of a developing choroid plexus
with a posterior choroidal branch. The anastomo-
sis with the posterior circulation is still absent in
these animals.
Reptiles
The changes seen with the shift from amphibians
to reptiles are schematized in Fig. 14. From the
lateral striate (Heubner forerunner) branch of the
ing the cerebral hemispheres (Fig. 14). These
multiple perforators will eventually become the
MCA. We understand now the concept that the
very large human MCA is a relatively late
phylogenetical acquisition and represents a
hypertrophied perforator branch of the anterior
cerebral artery (Fig. 14). From another perspec-
tive we can also understand how the MCA can be
embryologically and phylogenetically considered
a branch of the ACA.
The “posterior telencephalic branch” of the
ICA, forerunner of the anterior choroidal, supplies
the posterior telencephalic structures and is
connected anteriorly with the olfactory artery;
this function will continue to be fulfilled until
very late in mammals’ evolution, when the poste-
rior brain territory will be annexed by the “poste-
rior cerebral artery.” Other important steps in
reptiles are the formation of an anterior anastomo-
sis between the medial olfactory arteries, which is
the future Acomm, and the enlargement of the
cerebellum, particularly of the more caudal ver-
mis, which is supplied by the distal “basilar”
system.
10 M. Shapiro and E. Raz

Fig. 14 Amphibians Reptiles

Medial Olfactory
ACOM
Lateral Olfactory
Lateral Striate
Lat. Striate perforators (MCA)

Posterior Telencephalic
(future Ant. Choroidal)
Posterior Choroidal
Tectal
Cerebellar

Inferior Cerebellar (PICA)

Vertebral
Segmental Carotido-
vertebral Anastomoses
Carotid

Fig. 15 Reptiles Birds

Medial Olfactory (ACA)
Lateral Olfactory
(Heubner)

Lat. Striate perforators

MCA
Posterior Telencephalic
(future Ant. Choroidal)

Posterior Choroidal
Tectal
Cerebellar

Inferior Cerebellar (PICA)

Vertebral
Segmental Carotido -
vertebral Anastomoses

Carotid

Birds of its territory (Fig. 15). This large telencephalic

In the birds a discrete MCA emerges as a domi-
nant perforator from among a collection of lateral
striate vessels, driven by rapidly enlarging cortical
volume (Fig. 15). The posterior cerebral artery
(PCA) of birds, which is in fact the anterior cho-
roidal, continues as the dominant parieto-occipital
supply source, while the tectal artery (which is the
real PCA) extends its coverage and forms a func-
tional anastomosis with the anterior choroidal
system, setting the ground for future annexation
territory of the anterior choroidal will be lost at the
early stages of embryogenesis. In the birds, the
vertebral system remains confined to the cord or in
some cases lower medulla/vermian region
(Fig. 15).
Mammals
The biggest change in mammals regards the
development of the caudal ramus, since the
enlargement of the cerebellum and of the posterior
1 Cerebrovascular Development and Evolution 11

Fig. 16 Birds Mammals

ACA
(Heubner)
Lat. Striate perforators

MCA
Ant. Choroidal
Posterior Choroidal

Tectal (PCA)
Cerebellar
AICA
Inferior Cerebellar (PICA)

Segmental Carotido -
vertebral Anastomoses

Fig. 17 Mammals Human

ACA
(Heubner)
Lat. Striate perforators
MCA
Ant. Choroidal
Posterior Choroidal
Tectal (PCA)
Superior Cerebellar

AICA
Inferior Cerebellar (PICA)

Segmental Carotido-
vertebral Anastomoses

telencephalon place increasing demand on the
carotid system. Notwithstanding, also in some
mammals such as the sheep, the vertebral system
does not contribute to brain supply, and the basilar
artery is continuing to demonstrate craniopetal
flow from the internal cerebral system. However,
in higher mammals the carotid system appears to
reach its functional limit, leading to progressive
annexation of the brainstem and cerebellar terri-
tory by the vertebrobasilar system (Fig. 16). At
first, this is confined to the PICA region. Progres-
sively, however, the vertebral artery annexes
portions of basilar territory, resulting in reversal
of craniopetal basilar flow to the craniofugal form
observed in monkeys, apes, and humans (Figs. 16
and 17).
The phylogenetically novel participation of the
vertebral system in brain supply is maximized in
the human, where the posterior cerebral artery,
previously known as the tectal artery in the
“lower” species, is now functionally acquired by
the posterior circulation, even though develop-
mentally it belongs to the carotids (Fig. 17).
Even as such, frequent occurrence of “fetal”
12
Pcomm disposition attests to this relatively recent
phylogenetic acquisition (see embryology sec-
tion). Recent territory shifts in this region
are borne out by marked variability in PICA-
AICA dominance, whereas the phylogenetically
older superior cerebellar arteries are much less
variable.
Vascular Neuroembryology
Introduction
At about 4 weeks of gestation, the neural plate has
just closed, and the brain is a tube with a central
canal of fluid, similar to the spinal cord. A series
of small vessels wrap around the brain and nourish
it by simple diffusion in a similar way to that of the
spinal cord (Fig. 18).
Fig. 18
M. Shapiro and E. Raz
The way the vessels are arranged, i.e., trans-
verse vessels connected by longitudinal ones
(Fig. 18), is called metameric, because they ulti-
mately relate to the metamers, somites, segments,
levels, etc., that are the basic building blocks of
our anatomy, similar to the arrangement found in
the lower species.
The early arrangement centers on the six aortic
arches, which relate to the branchial arches
(Fig. 19). The aortic arches consist of the dorsal
and ventral aortae, connected by longitudinal
anastomoses – the arches themselves. There are
a few adult arrangements which are almost
constant:
• The left-sided fourth aortic arch will become
the adult aortic arch.
• The third arch will become the proximal inter-
nal carotid artery.
1 Cerebrovascular Development and Evolution
Fig. 19
• The dorsal aorta between the third and fourth
arch regresses, while homologous ventral aorta
becomes the common carotid.
• The ventral aorta rostral to the third arch will
become the external carotid artery.
Other than the dorsal and ventral aortae, there
is an additional longitudinal system, called the
longitudinal neural system, related more
directly to the developing brain and spinal cord
(Fig. 20), which consists of a series of metameric
vessels with longitudinal connections. This longi-
tudinal neural system will eventually mature
into the vertebrobasilar system. We will go
through six stages: the prechoroidal stage
(stage 1), the choroidal stage (stage 2), the telen-
cephalic primitive stage (stage 3), the telence-
phalic intermediate stage (stage 4), the
telencephalic final stage (stage 5), and the adult
disposition (stage 6).
13
Stage I (Prechoroidal Stage):
Approximately 4 Weeks
There is a large vessel arising from the fourth
aortic arch (Fig. 21). Keeping in mind that this is
a simplification, we will call it the common
carotid. The place of origin of the internal carotid
is shown, but again the external carotid is not
displayed. The carotid system appears to feed
the entire brain, which is in fact true of the early
human embryology; the posterior circulation does
not exist yet. From this carotid vessel, a series of
small arteries wrap around the brain and nourish it
by simple diffusion (Fig. 21).
At this point, the carotid divides in two rami –
the cranial and the caudal. The cranial ramus is
simply a larger feeding vessel among the many
smaller vessels that wrap around the brain. It is the
forerunner of the anterior cerebral artery, which is
the phylogenetically oldest telencephalic vessel.
14
Fig. 20
The caudal ramus feeds the midbrain and hind-
brain and is the forerunner of the future Pcomm,
P1, and part of the basilar. These adult segments
therefore all belong embryologically to the ante-
rior circulation; this is a quite important concept to
bear in mind in order to understand the vascular
variation and pathology. Blood flow in the caudal
ramus at this point is craniocaudal.
Four transverse vessels are known to anasto-
mose the caudal ramus with the ICA,
corresponding to the segmental arrangement men-
tioned above. These segmental vessels are named
trigeminal, hypoglossal, and proatlantal type
1 and 2 arteries (Fig. 22). They are very similar
to the abovementioned transverse metameric
arteries; they receive special attention because of
their relationship with branchial arteries and
nerves. For example, the trigeminal artery sup-
plies the trigeminal nerve, which is the nerve of
the first arch.
M. Shapiro and E. Raz
These transverse arteries are in hemodynamic
balance with the caudal ramus of the carotid artery
in supply of the brainstem. As the longitudinal
neural system matures, the future vertebral artery
begins to meet to demands of the brainstem and
cerebellum, and, as a consequence, the anasto-
motic transverse arteries typically regress, though
never fully disappear. Rarely, one of these vessels
remains dominant and thus continues to supply
the brainstem, for example, in the so-called per-
sistent trigeminal artery. As was mentioned, the
posterior circulation does not exist as such in
embryological terms. Now look again at the
small vessels surrounding the brain. From
among many of these feeders, a dominant vessel
eventually arises to wrap around the brain and
sprout its own small branches. The concept of
selection for a dominant vessel among many is
very important, because it explains most “vari-
ants” seen in the adult. From this arrangement it
1 Cerebrovascular Development and Evolution
Fig. 21
is clear that if two of these small vessels continue
to be in hemodynamic balance instead of one
vessel being dominant, an apparent duplication
may ensue.
Stage II (Choroidal Stage):
Approximately 5 Weeks
The brain continues to be supplied entirely by the
anterior circulation, and the choroid (purple stuff
over the blue tube in Fig. 23) develops as an
invagination of the ependymal cells into the cen-
tral canal from the “roof” of the brain and helps
supply nutrients to the brain from the inside of the
neural tube. The choroid plexus is quite metabol-
ically active and promotes development of its own
vascular supply. At this stage, a vessel is seen
coming off the cranial ramus to feed the plexus
of the telencephalon – this is the future anterior
choroidal artery (Fig. 23). It is a hemodynamic
balance with the distal cranial ramus (the ACA)
and is the earliest feeder of the posterior
15
telencephalon. The choroid plexus of the dien-
cephalon and midbrain is supplied by the artery
called “posterior choroidal” (Fig. 24). The con-
nection between telencephalic and diencephalic
central tube – where in the future the lateral and
third ventricles communicate through the foramen
of Monro – is important as the area of hemody-
namic anterior choroidal/posterior choroidal bal-
ance. The choroid of the medulla (future 4th
ventricle) is being supplied by a dominant feeder
that will become the PICA. Variations of posterior
fossa supply are common, but the 4th ventricle
choroid plexus is almost always supplied by a
PICA-like vessel even when an AICA-PICA var-
iant is present.
The early cerebellum is fed by a ramus that
belongs to the anterior circulation, the future supe-
rior cerebellar artery. As a rule, the older the
vessel, the more constant its presence. And this
is particularly true for the SCA, which is essen-
tially always present, while variations involving
the AICA and PICA are quite common (Fig. 25).
When a case of agenesis of these early vessels is
16
Fig. 22
considered carefully, the vessel is in fact found to
exist under an incorrect name. For example, ante-
rior choroidal agenesis in fact represents an
enlarged anterior choroidal that supplies the
PCA territory, while the latter is hypoplastic.
This is confirmed by the observation that in
most mammals the flow in the vertebrobasilar
system is directed caudally and the reversal of
flow of the posterior circulation is a very recent
phenomenon. The important changes taking place
in the longitudinal neural system (LNS) are
responsible for this and will be illustrated in the
next stage.
Stage III: 6 Weeks
What happens is that the discontinuous LNS
plexus coalesces to form actual distinct longitudi-
nal vessels (Fig. 20). The one next to the cord is
known as the “longitudinal spinal artery” and is
M. Shapiro and E. Raz
the homologue of the future anterior spinal artery.
The one more laterally related to the vertebral
bodies will become the future vertebral artery.
In the anterior circulation, the growing cere-
bellum is going to annex the brainstem forerun-
ners of the AICA and PICA, and the branches of
the cranial ramus develop branches of their own,
some of which will become important adult struc-
tures, such as the Heubner or the MCA, but at this
point the ACA and the anterior choroidal continue
to dominate the supply of the telencephalon.
Stage IV: Approximately 7 Weeks
The LNS continues to “mature” in an interesting
way: the two longitudinal neural arteries that are
the forerunners of the anterior spinal artery and the
basilar artery fuse into one vessel for most of the
length of the cord and the brainstem. The whole
vasculature is starting to resemble the adult
1 Cerebrovascular Development and Evolution
Fig. 23
pattern. Still, at this point the flow in the new
basilar remains caudally directed. The PICA con-
tinues to expand to cover the cerebellum and
acquires its two basic divisions – choroidal and
parenchymal. The trigeminal, hypoglossal, and
proatlantal vessels continue to regress as the ver-
tebral axis matures.
As regards the anterior circulation, the most
important change concerns the development of
the future “PCA,” which takes over the territory
previously supplied by the anterior choroidal
artery (Fig. 26). The growth of the PCA parallels
the development of the vertebrobasilar system, by
which it will be annexed. But it is not known
whether it takes place after the flow in the basilar
has been reversed or before.
Stage V: Approximately 9 Weeks
The posterior circulation has assumed a more or
less adult pattern (Fig. 27). At some point about
this time, the flow becomes craniocaudal. The
17
change in flow direction seems to trigger some
of the final events of embryogenesis. The relief of
the anterior circulation from the responsibility to
supply the posterior brain and the posterior
fossa allows for continued growth of the PCA
and for development of the true MCA from sev-
eral ACA perforators. Indeed, conceptually and
embryologically, the MCA is simply a very large
perforator with cortical territory, similar to the
Heubner. For example, the accessory MCA vari-
ant is simply a case of two perforators being in
relative balance and thus developing into
MCA-like vessels.
Stage VI: Approximately Adult Pattern
One of the particular curiosities of the late vascu-
lar embryogenesis is the seemingly sudden
appearance of the Acomm (Fig. 28). Conceptu-
ally, it is easy to think of it as a short segment
fusion of the ACA in a fashion homologous to the
basilar.
18
Fig. 24
To conclude, it is noteworthy to state how this
section only illustrates the basic framework of the
cerebral circulation, but the development of
smaller vessels, countless branches of branches,
and of the penetrating vasculature continues well
after birth, as the brain matures and this level of
complexity is quite beyond our scope.
Practical Aspects of Neurophylogeny
and Neuroembryology
Fetal Posterior Cerebral Artery
This configuration is the most common “variant”
of the cerebral circulation. The fetal Pcomm, or
fetal origin of the posterior cerebral artery, occurs
when the Pcomm is larger than the P1 segment
and supplies the bulk of the posterior cerebral
artery territory. This reflects the effective persis-
tence in carotid supply of the PCA, a reminder of
the phylogeny and embryology of this vessel.
M. Shapiro and E. Raz
A unilateral fetal posterior cerebral artery is seen
in 25 % of cases, bilateral in 5 % cases (Figs. 29,
30, and 31). In Figs. 30 and 32 examples of fetal
Pcomm strokes are demonstrated. The existence
of a “duplicated Pcomm” is not possible – as may
be confused on occasion with a larger anterior
choroidal artery.
Accessory MCA
The MCA (purple in Fig. 33) arises as a dominant
vessel from a number of lateral striate perforators
of the anterior cerebral artery in the reptile. It may
happen in the case where two perforators remain
dominant, resulting in the “accessory MCA”
(Fig. 33). The accessory vessel is, in fact, either
a hypertrophied recurrent artery of Heubner – a
medial ACA perforator – or another perforator-
like vessel. To qualify as an MCA, the vessel must
have cortical territory. In the cartoon in Fig. 33,
the schematic on the left side shows the accessory
1 Cerebrovascular Development and Evolution
Fig. 25
MCA configuration where both branches appear
to originate proximal to the A1 complex (which is
here defined as segment past the more “distal”
MCA branch). These are known as “Manelfe
type 1” or “type 2” – depending on which branch
is larger. The important feature however is to note
from which vessel the perforators originate and
whether they are medial or lateral. The schematic
on the right side shows the Heubner-type aMCA,
known as “Manelfe type 3.” Check out the exam-
ples in Figs. 34 and 35.
Dominant Anterior Choroidal Artery
The anterior choroidal artery, until late in evolu-
tion, supplies a substantial portion of cortical ter-
ritory, which in the human is annexed by the PCA.
This annexation is determined by the proximity of
the two vessels in the subarachnoid space around
the midbrain. On occasion, the degree of annexa-
tion is incomplete, with large portions of the
19
cortex remaining under anterior choroidal control
while PCA contribution is correspondingly
reduced. This is sometimes erroneously consid-
ered a “duplicated PCA” (Fig. 36). Mistaking the
choroidal artery for the Pcomm can have disas-
trous surgical consequences.
Persistent Trigeminal Artery
This most common embryonic carotid-
vertebrobasilar anastomosis at the mid-basilar
level represents persistence of segmental
ICA-basilar connection along the trigeminal
nerve (Fig. 37). The posterior communicating
artery is usually small in these cases.
Persistent Hypoglossal Artery
More caudal than the persistent trigeminal artery,
the persistent hypoglossal artery enters the cranial
20
Fig. 26
vault through the hypoglossal canal, connecting
the distal vertebral and the proximal internal
carotid arteries. It is, like the trigeminal and
proatlantal vessels, a segmental artery. Examples
in Figs. 38 and 39.
Proatlantal Artery
An embryonic carotido-vertebral anastomosis,
the proatlantal artery is, in fact, the
occipital artery. A C1 segmental connection
between the occipital and vertebral artery is the
proatlantal type I; C2 level connection is
proatlantal type II (Fig. 40). Usually, the vertebral
artery proximal to the proatlantal segment is
hypoplastic.
M. Shapiro and E. Raz
Variations in Fusion Patterns
of the Basilar Artery
The basilar artery is formed via coalescence of
multiple channels belonging to the longitudinal
neural system, following development of the ante-
rior circulation. This process sets the stage for
multiple basilar artery variations, which can be
considered from the standpoint of (1) extent of
fusion and (2) completeness of fusion (i.e.,
fenestrations).
It is conceptually helpful to think of the basilar
artery as being “zipped” in the middle, with ver-
tebral and PCA segments being “unzipped.”
Thus, both length and the integrity of the zipper
determine the final configuration of the artery. As
such, the artery may be:
1 Cerebrovascular Development and Evolution
Fig. 27
1. “Short” – corresponding to relative lack of
fusion of the caudal segment
2. Unfused at the top – corresponding to
“unzipping” at the basilar tip, less common
but much more confusing
3. Fenestrated – broken zipper in the middle
The first one, short basilar, is in fact a contin-
uum in terms of fusion extent (Fig. 41). There is
no “number” to guide what short or long is, but
some basilar arteries are clearly quite
foreshortened (the tortuous basilar of the aged
hypertensive is a different matter). The second
variation, involving lack of basilar tip fusion,
can generate a lot of confusion (Fig. 42). Effec-
tively, the top of the basilar is split in two, so that
one or both superior cerebellar arteries originate
from the P1 segment. This variant is not, there-
fore, a primary superior cerebellar artery
21
aberration but instead a “deficiency” in basilar
fusion (Fig. 42).
P1 origin of the superior cerebellar artery. The
SCA (blue) on the right arises from the P1 seg-
ment. Further “zipping” up would elongate the
basilar trunk, foreshorten both P1 segments, and
shift the SCA onto the basilar (Fig. 43). When
especially prominent, basilar tip “unzipping” can
lead to some confusion, as in this patient reported
to have a basilar tip aneurysm on CTA. Catheter
angiogram shows an unfused basilar tip with tor-
tuous P1 vessels simulating aneurysm, further
complicated by the presence of a fetal Pcomm
on the right.
The third, basilar fenestration (Fig. 44), is quite
common and usually of little clinical significance,
except when it is so short as to mimic a dissection
(Fig. 45). Occasional species, including some
monkeys, do not have a basilar artery at all. The
22 M. Shapiro and E. Raz

Fig. 28

Fig. 29 Bilateral fetal Pcomm (blue arrows) which reflect effective persistence in carotid supply of the PCA. The P1
segment of the PCA (brown) is hypoplastic
1 Cerebrovascular Development and Evolution 23

Fig. 30 CTA reconstruction demonstrating a right fetal related to a clot in both the fetal Pcomm and the MCA with
Pcomm (black arrows). This disposition was particularly a very large infarct. The AChoA territory was preserved
unfavorable in this patient when he/she developed a stroke

Fig. 31 Fetal Pcomm on a cerebral angiogram marked in Notice other unusual arrangements in this case, including a
red on left ICA injection lateral projection. AP vertebral somewhat low origin of the left superior cerebellar artery
artery injection shows a small left P1 segment (orange), as (yellow) and a basilar fenestration (brown)
compared with normal course of the right PCA (green).
24
Fig. 32 PCA territory infarction in a case of “fetal
Pcomm.” Time to Peak and rCBV maps (top) demonstrate
a completed infarction in right PCA territory. The embolus
(right lateral ICA injection on bottom left) originated from
caudal rami are unfused throughout, mimicking
the anterior cerebral disposition.
Duplications
As in the case of accessory MCA, essentially any
vessel on occasion is gifted with a twin, not an
M. Shapiro and E. Raz
the right ICA. Typical diminutive appearance of the basilar
artery in a patient with bilateral fetal Pcomms is seen on
bottom right T2 image
identical one, since these branches actually
supply separate territories. Understanding that
arteries in essence are dominant collectors arising
from a number of potential homologous candi-
dates clarifies this occurrence. The most com-
monly duplicated vessel is the superior
cerebellar artery. Duplicated AICA is quite rare
(Fig. 46).
1 Cerebrovascular Development and Evolution
Fig. 33
Fig. 34 Catheter angiogram of a patient with type I/II
accessory MCA. Notice how well MCA perforators can
be seen, originating from the smaller and more distal of the
two MCA vessels
25
Anterior Cerebral Artery Variations
The number of potential ACA variations may
seem mind-boggling. Phylogeny and embryology,
however, help make sense of frustration and con-
fusion. The appearance of the ACA system can be
conceptualized in terms of two factors – extent of
fusion and pattern of branch takeoff.
On the fusion side, you can expect anything
from no fusion – that is, no Acomm, various short
segment patterns of fusion (duplicated Acomm),
or elongated segments of fusion (unpaired or azy-
gos arrangements). The unpaired type is one
where a long segment of fusion finally breaks up
into two branches, each of which gives rise to
individual cortical branches to its respective hemi-
sphere. The azygos type is where the ACAs are
fused for the entire length, with individual
branches to both hemispheres arising from a com-
mon trunk. Thinking of unpaired/azygos arrange-
ments as a kind of “basilar” is helpful and may not
be conceptually incorrect. Many species, such as
dogs, have a “normal” azygos arrangement with
“variants” of unpaired or “human-type” disposi-
tion. In the human, unpaired cases have a much
higher incidence of aneurysm formation at the
eventual branch point, – not a favorable evolu-
tionary development.
On the takeoff pattern side, you can have any
takeoff pattern imaginable, but of particular
importance is the arrangement at the Acomm
region. The frontopolar branch may arise from
the Acomm complex. More importantly, the
callosomarginal-pericallosal split may occur any-
where along the length of ACA from Acomm to
the callosum genu. Particularly when this split
occurs at the Acomm region, a triplicated appear-
ance may result with pericallosal and
callosomarginal trunks arising from one side and
regular A2 from the other, for a total of three
arteries (Fig. 47). Ultimately, ACA analysis in
terms of some combination of fusion extent and
branch takeoff can clarify any ACA arrangement.
26
Fig. 35 Type III (Heubner-
like) aMCA originating
from the Acomm complex
Fig. 36 The anterior choroidal artery (black), until late in
evolution, supplies a substantial portion of cortical terri-
tory, which in the human is annexed by the PCA (brown).
On occasion, the degree of annexation is incomplete, with
Conclusion
The deep understanding of the phylogenesis and
embryogenesis of the brain and spine vasculature
is paramount to develop a good perception of
M. Shapiro and E. Raz
large portions of the cortex remaining under anterior cho-
roidal control (blue), while PCA contribution is corre-
spondingly reduced (green). On MRA images, the
Pcomm is labeled in red and choroidal in yellow
anatomy, the physiology, and the pathology of
the central nervous system. Countless are the
cases in which this knowledge comes in handy
in the real life of a neuro doctor, radiologist,
clinician, or surgeon.
1 Cerebrovascular Development and Evolution 27

Fig. 37 This diagram

demonstrates the
persistence of the trigeminal
artery. The angiographic
images illustrate a persistent
trigeminal, an accessory
MCA, and an associated
aneurysm treated with coils.
Notice the classic tau shape
of the trigeminal artery in
the lateral view

Fig. 38 More caudal than

the persistent trigeminal
artery (green arrow on the
diagram), the persistent
hypoglossal artery enters
the cranial vault through the
hypoglossal canal,
connecting the distal
vertebral and the proximal
internal carotid arteries. It
is, like the trigeminal and
the proatlantal vessels, a
segmental artery
28 M. Shapiro and E. Raz

Fig. 39 Different patient. Source and reconstructed MRA vertebrals and Pcomms are hypoplastic. Notice the flow
demonstrating a large vessel traversing through the hypo- void on the axial T2-weighted image
glossal canal and becoming the basilar artery. Both

Fig. 40 Stereo AP and lateral views of left ECA injection, (red). This is the proatlantal type I; the proximal occipital artery
opacifying the vertebrobasilar system (yellow) via the C1 (red) is the proatlantal. Occipital artery distal to the vertebral
segmental artery (purple) connection to the occipital artery anastomosis is blue, and internal maxillary artery is black
1 Cerebrovascular Development and Evolution 29

Fig. 41

Fig. 42

Fig. 43 P1 origin of the

superior cerebellar artery.
The SCA (blue) on the right
arises from the P1 segment.
Further “zipping” up would
elongate the basilar trunk,
foreshorten both P1
segments, and shift the SCA
onto the basilar
30 M. Shapiro and E. Raz

Fig. 44

Fig. 45

Fig. 46 Rare case of

duplicated AICA (brown) is
shown on diagram and
frontal view of left vertebral
artery injection
1 Cerebrovascular Development and Evolution
Fig. 47 Lateral projection angiogram with early
callosomarginal takeoff (purple) and pericallosal in red.
Corresponding MRA image with left A2 shown in yellow
and Acomm in dark blue. This is the appearance of a
“triplicated ACA” which is discussed above; it is a
References
1. Abanou A, Lasjaunias P, Manelfe C, Lopez-Ibor L
(1984) The accessory middle cerebral artery
(AMCA). Diagnostic and therapeutic consequences.
Anat Clin 6(4):305–309
2. Butler AB, Hodos W (2005) Comparative vertebrate neu-
roanatomy: evolution and adaptation. Wiley, Hoboken
3. Campos C, Churojana A, Rodesch G, Alvarez H,
Lasjaunias P (1998) Basilar tip aneurysms and basilar
tip anatomy. Interv Neuroradiol 4(2):121–125
4. Dimmick SJ, Faulder KC (2009) Normal variants of
the cerebral circulation at multidetector CT angiogra-
phy. Radiographics 29(4):1027–1043
5. Gore AV, Monzo K, Cha YR, Pan W, Weinstein BM
(2012) Vascular development in the zebrafish. Cold
Spring Harb Perspect Med 2(5):a006684
6. Krings T, Baccin CE, Alvarez H, Ozanne A, Stracke P,
Lasjaunias PL (2007) Segmental unfused basilar artery
with kissing aneurysms: report of three cases and liter-
ature review. Acta Neurochir (Wien) 149(6):567–574;
discussion 574
7. Lasjaunias P, Braun JP, Hasso AN, Moret J, Manelfe C
(1980) True and false fenestration of the vertebral
artery. J Neuroradiol 7(3):157–166
31
misleading term since not three but two ACA arteries are
present. It is the A2 segment on the right that is absent
because of early bifurcation into callosomarginal and
pericallosal
8. Lasjaunias P, Manelfe C (1979) Arterial supply for the
upper cervical nerves and the cervicocarotid anasto-
motic channels: systematization of radiological anat-
omy. Neuroradiology 18(3):125–131
9. Lasjaunias P, Moret J, Manelfe C, Théron J, Hasso T,
Seeger J (1977) Arterial anomalies at the base of the
skull. Neuroradiology 13(5):267–272
10. Lasjaunias P, Santoyo-Vazquez A (1984) Segmental
agenesis of the internal carotid artery: angiographic
aspects with embryological discussion. Anat Clin
6(2):133–141
11. Lasjaunias P, ter Brugge KG, Berenstein A (2007)
Surgical neuroangiography: vol. 3: clinical and inter-
ventional aspects in children, vol 3. Springer Science &
Business Media, Berlin
12. Meckel S, Spittau B, McAuliffe W (2013) The persis-
tent trigeminal artery: development, imaging anatomy,
variants, and associated vascular pathologies. Neuro-
radiology 55(1):5–16
13. Siqueira M, Piske R, Ono M, Marino Júnior R (1993)
Cerebellar arteries originating from the internal
carotid artery. AJNR Am J Neuroradiol
14(5):1229–1235
14. Vignaud J, Hasso AN, Lasjaunias P, Clay C (1974)
Orbital vascular anatomy and embryology. Radiology
111(3):617–626
 Atherosclerosis: Biological and
Pathological Basis 2
Michele Porcu, Eytan Raz, and Luca Saba

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Atherosclerosis is a extremely complicated and
still well not perfectly understood pathologic
Arteries: Classification and Anatomical
Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
process that involves the arteries of the human
body. It is a multi-step process that starts with
Normal Physiological Aspects . . . . . . . . . . . . . . . . . . . . . . . 34 the endothelial damage and can be influenced
Pathogenesis of Atherosclerosis . . . . . . . . . . . . . . . . . . . . . 34 by many factors, modifiable or not. The defi-
Endothelial Injury and Dysfunction . . . . . . . . . . . . . . . . . . 35 nition of the grade of stenosis determined by an
Plaque Growth and Evolution . . . . . . . . . . . . . . . . . . . . . . . . 35
Types of Plaque and Possible Complications . . . . . . . . . 36
atherosclerotic plaque and the identification of
vulnerable plaques are the main goals for diag-
Imaging Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
nosis, staging and therapies.
Therapeutic Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Keywords
Pathogenesis • Risk factors • Nitric Oxide
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
(NO) • Vulnerable plaque

Introduction

Atherosclerosis can be defined as a multifactorial

chronic inflammatory disease of the large (elastic)
and medium (muscular) arteries, characterized by
endothelial dysfunction, vascular inflammation,
and deposition of fibro-fatty particles in the con-
text of the vessel wall [1, 2]. This pathological
M. Porcu (*) • L. Saba
Department of Radiology, AOU of Cagliari, Monserrato, process is on the bases of a huge amount of com-
Cagliari, Italy mon clinical conditions, including myocardial
e-mail: micheleporcu87@gmail.com; lucasabamd@gmail. ischemia and infarction and cerebrovascular and
com; lucasaba@tiscali.it
peripheral vascular disease.
E. Raz In this chapter, we will briefly analyze the
Neurointerventional Radiology Section, Department of
biological and pathological basis of atherosclero-
Radiology, NYU Langone Medical Center, New York,
NY, USA sis, focalizing our attention in particular on the
e-mail: eytan.raz@gmail.com formation process of the atherosclerotic plaques.
# Springer Science+Business Media New York 2016 33
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_48
34
Arteries: Classification and Anatomical
Structure
The arteries are those tubular shape structures of
the human in which the blood flows from the heart
to the periphery.
According to the caliber of the vessel, we can
recognize large arteries (caliber between 30 and
7 mm) and medium and small (caliber between
7 and 0.2 mm) arteries [3]. Because of the differ-
ent characteristics of the tunica media (see below),
the large arteries are called even elastic arteries,
whereas those of medium and small caliber are
called muscular arteries.
The vessel wall can be divided into three dif-
ferent layers, which are from the inner to the outer
(1, 2, 3, 4):
(a) Tunica intima: it is the innermost, and it
consists of a single layer of endothelial cells
(endothelium) that lie on their basement
membrane. The endothelial cells have an
important role in the production and secretion
of several molecules that play a key role in the
regulation of the blood flux, of the hemostasis,
and of the inflammatory process.
(b) Tunica media: this layer is the thickest of the
three, and it is composed of vascular smooth
muscle cells (VSMC), elastin fibers, and
extracellular matrix. The elastin fibers are
more represented in the tunica media of large
caliber arteries, whereas the tunica media of
medium and small arteries consists mainly of
VSMC. The VSMC are arranged on a circular
or spiral way.
The function of this layer in the elastic
arteries is to accumulate the kinetic energy
generated during the systolic phase of the
cardiac cycle, releasing it during the diastolic
phase in order to facilitate the blood flow to
the periphery; in the muscular arteries, it rules
the tone of the arteries and so the amount of
blood which reaches the organs.
(c) Tunica adventitia: it is the outermost layer
and consists of connective tissue in continuity
with the perivascular connective tissues. In
the context of this layer, the vasa vasorum
M. Porcu et al.
Table 1 Principal mediator produced by the endothelium
Molecule Function
Nitric oxide (NO) Vasodilator
Prostacyclin (PGI2) Vasodilator
Endothelin-1 (ET-1) Vasoconstrictor
Endothelium-derived contracting Vasoconstrictor
factors (EDCF)
are present: these small vessels are fundamen-
tal for the nutrition of the tunica media and
could be an important site for
neoangiogenesis and inflammatory
processes [1].
Normal Physiological Aspects
An artery is not a simple “conduct” for the blood;
by the opposite, it has to be considered a proper
organ, with proper physiological functions
(in particular the endothelium).
The endothelial cells can produce and secrete
different molecules after mechanical or chemical
stimulation or the interaction with molecules
transported in the blood, such as the mediators
of inflammation [1, 2, 4]: some of the most impor-
tant molecules produced by the endothelium are
reported in Table 1.
The activity of the VSMC can be regulated
directly by the sympathetic and parasympathetic
system, by the circulating catecholamines (adren-
aline and noradrenaline) and other molecules such
as angiotensin II, antidiuretic hormone (ADH),
and corticosteroid [4].
Pathogenesis of Atherosclerosis
The pathogenesis of atherosclerosis is a compli-
cated and still not well-known multistep process.
Most of these processes are chained one to each
other, and a huge number of factors are involved.
In this paragraph, we will try to explain the
main elements and mechanisms which lead to the
formation of the atherosclerotic plaque.
2 Atherosclerosis: Biological and Pathological Basis
Endothelial Injury and Dysfunction
In paragraph two, we have seen how the endo-
thelium has to be considered not just an internal
smooth cover of the artery but an active tissue
important for many functions. On the surface of
the endothelium, a great number of receptors are
present, for example, insulin, serotonin, hista-
mine, serum lipoproteins, and adhesion mole-
cules more expressed during the inflammatory
status.
The endothelium is also able to produce many
chemical substances, including nitric oxide
(NO) via the endothelial isoform of the nitric
oxide synthases (eNOS) [1, 2, 5]. This chemical
mediator has a very simple chemical structure; it is
unstable, with a very short life (very low seconds),
and once it is produced, it acts on the VSMC
inducing vasodilation and slowing their prolifera-
tion and reduces the expression of adhesion mol-
ecules for macrophages on the endothelial surface
(see below) [6, 7].
The first step of the pathogenesis of the athero-
sclerosis is the endothelium damage [1, 2, 7].
This process occurs more frequently in those arte-
rial regions in which the blood flow is non-laminar
or turbulent (e.g., the bifurcations or the origin of
collateral vessels) [1], and it is favored by “risk
factors” such as hypertension, smoking, and
hypercholesterolemia through mechanisms like
free radical oxidation and the activation of multi-
ple genetic pathways [2, 7].
The endothelial damage is characterized by the
expression of pro-inflammatory receptors on their
surface, the reduction of the production of the NO,
and the rearrangement of the junction and cyto-
skeletal proteins [2, 8, 9]. The final results are an
increased endothelial cell turnover, an increased
permeability of the endothelium with the recruit-
ment of inflammatory cells (mainly scavenger
macrophages), and the accumulation of lipopro-
teins (in particular LDL; see below) beneath the
tunica intima [2]. This process tends to have a
positive feedback: the more the permeability, the
more the inflammation, the more the lipid accu-
mulation, the more the reduction of the caliber of
the lumen of the vessel, and the more the
35
Table 2 Principal modifiable risk factors of
atherosclerosis
Principal modifiable risk factors
Hypertension
Dyslipidemia and hypercholesterolemia
Obesity
Diabetes
Smoking
endothelial damage induced by the blood flow
(together with the other risk factors we have seen
before).
Plaque Growth and Evolution
Once the endothelial damage has started, the pro-
cess continues with the formation of the athero-
sclerotic plaque.
The first lesions evident in the arteries are the
so-called fatty streaks, which are linear lesions
well evident in particular in younger patients.
These kinds of lesions tend to modify during the
years, according to various factors:
• Non-modifiable factors (genetic, familiarity)
• Modifiable factors
The modifiable factors include the “risk fac-
tors” mentioned before, i.e., hypertension,
dyslipidemia and hypercholesterolemia, obesity,
diabetes, and smoke (Table 2): the presence of
these factors accelerates the process and forma-
tion of the atherosclerotic plaques, increasing the
probabilities of minor and major cardiovascular
and cerebrovascular events and pathologies of the
peripheral vascular system [10].
At a microscopic level during this phase, the
macrophages, the T cells, and the VSMC are
the most important cellular elements involved;
they operate inside the growing lesion in an
inflammatory context, interacting in particular
with the low-density lipoproteins (LDLs).
LDLs are lipoproteins with a very high rich
content in cholesterol. These lipoproteins circu-
late free in the blood. When the endothelium is
36
damaged, the LDLs can infiltrate beneath the
tunica intima and accumulate into the growing
plaque. At this level, they are susceptible to the
oxidative stress of various enzymes such as
myeloperoxidases and lipoxygenases (produced
and released normally during an inflammatory
process by the immunity cells). The oxidized
LDLs (oxLDLs) thanks to their cytotoxic action
promote the recruitment of monocytes from the
blood circulation upregulating the expression of
adhesion molecules on the surface of the endothe-
lial cells [2, 11]. Once they have penetrated inside
the plaque, they differentiate into resident macro-
phages [1]. OxLDLs phagocytized by the macro-
phages are more resistant to degradation of their
lysosome, and the macrophages continue to accu-
mulate cholesterol, becoming foam cells [2]. The
death of the foam cells led to the formation of the
necrotic lipidic core inside the plaque [1].
The macrophages continue to feed this inflam-
matory reaction producing chemokines such as
monocyte chemoattractant protein-1 (MCP-1),
interferon-γ (INF-γ), and macrophage colony-
stimulating factor (MCFS), which increase
recruitment of other inflammatory cells [2].
Some of these chemokines attract in the grow-
ing plaque and stimulate the proliferation of the
VSMC [1]. These cells produce extracellular pro-
teins such as collagen and elastin, producing a
fibrous cap over the plaque and contributing to
the stability of itself; other molecules produced,
the proteoglycans and fibronectin, create a reticu-
lar web that contribute to trap LDLs inside the
lesion [12]. Other molecules produced are the
matrix metalloproteinases (MMPs), whose role
in the degradation of the extracellular matrix is
very important in the macrophage migration and
destabilization of the plaque [13, 14].
At a microscopic level, it has been seen that as
the plaques increase in complexity, mineralization
processes can develop inside them, leading to the
formation of calcifications [2]. In particular medi-
ators of the bone metabolism have been evidenced
inside the plaque, close to the calcification
centers [15].
Of course because of the involvement of the
smooth musculature of the media layer in the
atherosclerotic process and the subsequent
M. Porcu et al.
stiffening of the arterial walls, the physiological
mechanisms of narrowing and dilation of the arte-
rial walls in response to vasoactive molecules and
neurovegetative regulation are damaged. Even
this condition further promotes the atherosclerotic
plaque growing process.
The continuous growth of the plaque can
extend inside the lumen of the vessel (negative
remodeling) or on the external side (positive
remodeling).
The negative remodeling is important because
it is responsible for the reduction of the blood flow
distally to the plaque, especially during stress
conditions. For example, a plaque that obstructs
the lumen of the coronary artery limits the flux of
blood to the territories supplied by that vessel: in
the early stages, when the obstruction is not crit-
ical (up to 70 % of reduction of the vessel caliber),
this condition could be compensated during rest,
but during stress condition such as intensive phys-
ical activity, the presence of the plaque does not
allow an adequate dilation of the coronary artery
and an adequate blood flow for the heart segments
supplied by the vessel, inducing myocardial
ischemia which manifests with pain and
neurovegetative symptoms.
On the other hand, the positive remodeling is
responsible for the dilation of the vessel involved
in the atherosclerotic process and can lead to the
formation of aneurysm. The evolution of the
aneurismatic lesions is highly variable, even if
the most terrible is the rupture of the vessel and
the consequent hemorrhage, often fatal.
Types of Plaque and Possible
Complications
The natural evolution of the atherosclerotic plaque
is highly variable, according to many factors,
modifiable or not (see before). In particular, we
can recognize two types of plaques, non-vulner-
able and vulnerable.
The non-vulnerable plaques are characterized
by the presence of a small necrotic lipidic core and
a thick fibrous cap and are less prone to rupture.
The vulnerable plaques are those more prone to
rupture and thrombosis [16]. From a histological
2 Atherosclerosis: Biological and Pathological Basis
point of view, the vulnerable plaques are usually
characterized by the presence of a large fatty
necrotic core, covered by a very thin fibrous cap
[17]. The cap consists of molecules (in particular
collagen, elastin, and proteoglycans) produced
largely by the VSMC and protects the core of
the plaque from the direct contact with the blood
[16]. The necrotic lipid core is rich in
prothrombotic molecules; after the fissuring of
the plaque, the necrotic lipid core is exposed to
the blood flow, unleashing the coagulation cas-
cade and the formation of thrombus; the thrombus
can completely occlude the vessel determining
ischemia and then infarction of the tissues sup-
plied by the vessels and/or can shatter and
embolize peripherally. A vulnerable atheroscle-
rotic lesion could even fissure after an acute
intra-plaque hemorrhage because of the rupture
of the thin-walled vasa vasorum [18].
Imaging Evaluation
Nowadays, we have many diagnostic imaging
methods to evaluate the gravity of the atheroscle-
rotic disease (in particular trying to identify the
vulnerable plaques, more prone to rupture)
suspected on the clinical and laboratory data. In
this paragraph, we will briefly expose the main
characteristics of these examinations.
Ultrasound (US) usually is the first imaging
technique used; it is safe, repeatable, and very pre-
cise in particular for the evaluation of the superficial
vessels, such as supra-aortic trunks located in the
neck, and the arteries of the upper and lower limbs.
Multidetector computed tomography (MDCT)
is more accurate for the evaluation of atheroscle-
rotic plaque; it uses ionizing radiation, and an
intravenous iodinate contrast medium administra-
tion is required in order to enhance the vessels and
to evaluate the composition of the plaque and the
grade of stenosis of the vessel. For example, the
evaluation of coronary arteries with a coronary
computed tomography angiography (CTA) is not
an invasive method for the study of patients with
low to intermediate risk of coronary artery disease
and allows not only to evaluate the grade of
37
stenosis of the vessels but even to evaluate the
composition of the plaques [19].
The new techniques introduced in magnetic
resonance imaging (MRI) are opening the way
for the characterization of the vulnerable
plaque [20].
Among the invasive methods of imaging, ves-
sel angiography is still the gold standard method
to evaluate the grade of occlusion of the vessel in
some fields (e.g., the study of coronary arteries)
and allows even to operate directly at the level of
the vessel-operating angioplasty, using stent in
order to make again patent the vessel or injecting
thrombolytic drugs intra-arteriously in case of
recent thrombosis. The intravascular ultrasound
(IVUS) is another promising technique that
allows assessing the composition of the plaque
with a very small US probe placed at the tip of
an angiographic catheter [21].
Therapeutic Options
The most important concept to remember is that
atherosclerosis is not a local disease, but it must be
considered a systemic disease of the body; a local-
ization of disease in the carotid arteries could be
associated often with the presence of plaques at
the level of the coronary artery and vice versa.
According to the localization, the characteris-
tics, and the dimensions of the atherosclerotic
plaques, we have different therapeutic options,
surgical (open surgery or endovascular treat-
ments) and nonsurgical.
Among the nonsurgical options, the most
important are the removal and the treatment of
the modifiable risk factors; in particular, in case
of high levels of LDL, the use of statins could be
indicated. These drugs block the cholesterol syn-
thesis acting on the 3-hydroxy-3-methylglutaryl
coenzyme A reductase at the hepatic level, but it
has also been evidenced that they act directly at
the level of the plaque inhibiting some inflamma-
tory processes, diminishing the recruitment of
inflammatory cells inside the lesion, and increas-
ing the production of NO [22].
The surgical options are reserved for those
cases in which the other therapies have been
38
resulted insufficient; it is not the goal of this
paragraph to expose in detail the huge number of
surgical procedures at our disposal, and we invite
the gentle readers to deepen in the next chapters of
this book and in other specialist texts.
Summary
In these pages we have seen how much compli-
cated and variable is the pathogenesis of athero-
sclerosis, in which a central role is played by the
endothelium damage and its subsequent dysfunc-
tion. Many factors, modifiable or not, influence its
evolution. The imaging techniques are fundamen-
tal for the evaluation of the plaque and of the
vessel stenosis, leading the clinicians and the sur-
geons to the correct therapeutic approach for the
patient.
References
1. Douglas G et al (2010) The pathogenesis of atheroscle-
rosis. Medicine 42:9; 38(8):397–402
2. Tao W et al (2012) Atherosclerosis: pathogenesis and
pathology. Diagn Histopathol 18(11):461–492
3. Balboni GC et al (2004) Anatomia Umana (terza edizione,
tomo I, chapter 3). Edi.Ermes s.r.l., Milano
4. Rindi G et al (2005) Fisiologia Umana (IX edizione).
UTET S.p.A. Divisione scienze mediche, Corso
Raffaello, Torino
5. Hopkins PN (2013) Molecular biology of atheroscle-
rosis. Physiol Rev 93:1317–1542. doi:10.1152/
physrev.00004.2012
6. Ignarro LJ et al (1987) Endothelium-derived relaxing
factor produced and released from artery and vein is nitric
oxide. Proc Natl Acad Sci U S A 84(24):9265–9269
7. Vanhoutte PM (2009) Endothelial dysfunction; the first
step toward coronary arteriosclerosis. Circ J 73:595e601
8. Deanfield JE, Halcox JP, Rabelink TJ (2007) Endothe-
lial function and dysfunction: testing and clinical rele-
vance. Circulation 115:1285e95
M. Porcu et al.
9. Sima AV, Stancu CS, Simionescu M (2009) Vascular
endothelium in atherosclerosis. Cell Tissue Res
335:191e203
10. Whayne TF Jr (2011) Atherosclerosis: current status of
prevention and treatment. Int J Angiol 20(4):213–222.
doi:10.1055/s-0031-1295520
11. Tsimikas S, Miller YI (2011) Oxidative modification of
lipoproteins: mechanisms, role in inflammation and
potential clinical applications in cardiovascular
disease. Curr Pharm Des 17:27e37
12. Doran AC, Meller N, McNamara CA (2008) Role of
smooth muscle cells in the initiation and early progres-
sion of atherosclerosis. Arterioscler Thromb Vasc Biol
28:812e9
13. Heissig B, Hattori K, Friedrich M, Rafii S, Werb Z
(2003) Angiogenesis: vascular remodeling of the extra-
cellular matrix involves metalloproteinases. Curr Opin
Hematol 10:136e41
14. Visse R, Nagase H (2003) Matrix metalloproteinases
and tissue inhibitors of metalloproteinases:
structure, function, and biochemistry. Circ Res
92:827e39
15. Dhore CR, Cleutjens JP, Lutgens E et al (2001) Differ-
ential expression of bone matrix regulatory proteins in
human atherosclerotic plaques. Arterioscler Thromb
Vasc Biol 21:1998e2003
16. Mishra TK et al (2013) An approach to the classifica-
tion, diagnosis and management of vulnerable plaque.
J Indian College Cardiol 3:57–66
17. Virmani R, Burke AP, Kolodgie FD, Farb A (2002)
Vulnerable plaque: the pathology of unstable coronary
lesions. J Interv Cardiol 15:439e446
18. Depre C, Havaux X, Wijns W (1996) Neovascu-
larization in human coronary atherosclerotic lesions.
Cathet Cardiovasc Diagn 39:215e220
19. Sun Z et al (2012) Coronary CT angiography: current
status and continuing challenges. Br J Radiol
85:495–510
20. Yang J et al (2013) MR and targeted molecular MRI of
vulnerable plaques. Interv Neurol 1(3–4):124–131.
doi:10.1159/000346767
21. Kovárník T et al (2014) Current status of intravascular
ultrasound in interventional cardiology. Vnitr Lek
60(12):1062–1067
22. Biasucci LM, Biasillo G, Stefanelli A (2010) Inflam-
matory markers, cholesterol and statins: pathophysio-
logical role and clinical importance. Clin Chem Lab
Med 48:1685e91
 Nonatherosclerotic Vascular Disease:
Biological and Pathological Basis 3
Pierleone Lucatelli, Beatrice Sacconi, and Carlo Catalano

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Atherosclerosis is the main cause of vascular
disease in most cases; in about 10 % of cases,
Takayasu Arteritis (TA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
carotid artery disease is related to
Giant Cell Arteritis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 nonatherosclerotic causes, including several
Fibromuscular Dysplasia (FMD) . . . . . . . . . . . . . . . . . . . 42 unfrequent pathologies such as Takayasu arter-
itis, giant cell arteritis, fibromuscular disease,
Moyamoya Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
moyamoya syndrome, arterial dissection and
Extracranial Internal Carotid Artery extracranial carotid aneurysm. These entities
Aneurysms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
are discussed in the present chapter, with a
Craniocervical Arterial Dissection . . . . . . . . . . . . . . . . . 44 special focus on pathogenesis. Indeed, the
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 aetiology of these diseases is in most cases
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
not completely known, since related to several
factors (genetic, immune and infectious). Early
diagnosis, usually leading to a good patient’s
outcome, is usually achieved after clinical
examination and imaging tests.

Keywords
Arterial dissection • Carotid artery disease •
Extracranial carotid aneurysm • Fibromuscular
dysplasia • Giant cell arteritis • Moyamoya
syndrome • Nonatherosclerotic disease •
Takayasu arteritis

Introduction

P. Lucatelli (*) • B. Sacconi • C. Catalano Atherosclerosis is identified as the main cause of

Department of Radiological, Oncological and Anatomo- vascular disease in around 90 % of cases; in the
Pathological Sciences, Sapienza University of Rome,
remaining 10 % of cases, nonatherosclerotic
Rome, Italy
e-mail: pierleone.lucatelli@gmail.com; beatrice. causes include several less common entities,
sacconi@fastwebnet.it; carlo.catalano@uniroma1.it such as Takayasu arteritis, giant cell arteritis,
# Springer Science+Business Media New York 2016 39
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_49
40
fibromuscular disease, moyamoya syndrome,
arterial dissection, and extracranial carotid aneu-
rysm. The pathogenesis of these diseases is in
most cases unclear or even related to several fac-
tors, such as genetic, immune, and infectious,
often associated with triggering events. The cor-
rect diagnosis is currently achieved after clinical
examination and imaging tests; although US, CT,
and MRA are useful, catheter angiography repre-
sents the gold standard in diagnosing most of
these diseases. The majority of patients have a
good outcome if the specific disease is diagnosed
early.
Takayasu Arteritis (TA)
Takayasu arteritis (TA) is a granulomatous arteri-
tis affecting the aorta and its branches [1]. The first
case was reported by Takayasu in 1905 [2]; lately,
the disease was more comprehensively described
as “pulseless disease” by Shimizu and Sano in
1951 [3]. It generally affects women in the first
fourth decades of life (nine females/one male),
with a general incidence of 2.6 cases per million
per year in the USA and major prevalence in
patients of Asiatic origin [4].
The etiology of TA is still unclear; the under-
lying pathologic process is inflammatory, with
several etiologic factors, either infective or auto-
immune, having been proposed. The most likely
hypothesis is that an unknown stimulus triggers
the expression of the 65 kDa heat-shock protein in
the aortic tissue which induces the major histo-
compatibility class I chain-related A (MICA) on
vascular cells. The T cells and NK cells recognize
MICA on vascular smooth muscle cells and
release perforin, resulting in acute vascular
inflammation; pro-inflammatory cytokines are
also released and increase the recruitment of
mononuclear cells within the vascular wall. Th1
lymphocytes drive the formation of giant cells
through the production of interferon-γ and acti-
vate macrophages with release of VEGF resulting
in increased neovascularization and PDGF,
resulting in smooth muscle migration and intimal
proliferation. The inflammatory cellular infiltrate
P. Lucatelli et al.
mainly involves the media and adventitia, usually
characterized by three stages: [5, 6].
1. A systemic stage, characterized by signs and
symptoms of an acute inflammatory condition,
such as fever, arthralgia, anemia, and increased
erythrocyte sedimentation rate
2. A vascular inflammatory stage, when vascular
stenosis and less frequently aneurysms occur,
with corresponding signs and symptoms
(stroke, transitory ischemic attacks, hypoten-
sive ischemic retinopathy with visual symp-
toms, vertebrobasilar ischemia, hypertensive
encephalopathy)
3. A burned-out stage, when fibrosis sets, usually
associated with remission According to the
American College of Rheumatology (ACR)
[7], the criteria for assessing the diagnosis are:
• Age at disease onset <40 years
• Claudication of extremities
• Decreased brachial artery pulse
• BP difference >10 mmHg
• Bruit over subclavian arteries or aorta
• Arteriogram abnormality, represented by arte-
riographic narrowing or occlusion of the entire
aorta, its primary branches, or large arteries in
the proximal upper or lower extremities, not
due to arteriosclerosis, fibromuscular dyspla-
sia, or similar causes (changes usually focal or
segmental)
A patient shall be said to have TA if at least
three of these six criteria are present.
According to the classification proposed by
Hata et al. [8], TA can be divided into six types
based on angiographic involvement:
• Type I – branches of the aortic arch
• Type IIa – ascending aorta, aortic arch, and its
branches
• Type IIb – Type IIa region plus thoracic
descending aorta
• Type III – thoracic descending aorta, abdomi-
nal aorta, renal arteries, or a combination
• Type IV – abdominal aorta, renal arteries, or both
• Type V – entire aorta and its branches
3 Nonatherosclerotic Vascular Disease: Biological and Pathological Basis
Hence, only types I, IIa, IIb, and V usually
involve the carotid arteries, whereas the other
types more commonly involve the thoracic
descending and abdominal aorta and the renal
arteries.
Angiography is the gold standard for diagno-
sis, even though color Doppler imaging can be
useful in the first part of the diagnostic process,
showing the mural thickening of the common
carotid arteries (hypoechoic in the early stages
and then hyperechoic after the development of
fibrosis). CT, and especially MRA due to the
young age of these patients, is also extremely
helpful, even more during the follow-up [9]. If
possible, whole-body MRI technique should be
employed in order to exclude other localizations
of disease [10].
TA is a chronic relapsing and remitting disor-
der. The overall 10-year survival rate is approxi-
mately 90 %, this rate being reduced in case of
major complications, such as stroke, intracranial
hemorrhage, and graft stenosis and/or occlusion.
For this reason, strict management of traditional
cardiovascular risk factors is mandatory in order
to minimize secondary cardiovascular complica-
tions. Approximately 20 % of patients have a
monophasic and self-limited disease; in the
remaining 80 % of patients, TA requires immuno-
suppressive treatment, resulting in remission in
around 60 % of cases [11].
Giant Cell Arteritis
Giant cell arteritis (GCA), also known as temporal
arteritis, is the most common form of vasculitis
occurring in adults, with a prevalence of persons
with active or remitted GCA of 200 cases per
100,000 population aged 50 years or older. The
female-to-male ratio is 3.7:1 [12]. It is a granulo-
matous arteritis affecting large- or medium-sized
vessels, usually the terminal branches of carotid
arteries (more frequently the temporal and oph-
thalmic arteries). The clinical manifestations of
giant cell arteritis result from two different pro-
cesses. Inflammatory cells infiltrate the arterial
wall and cause structural damage, eventually
41
leading to vascular complications. In the majority
of patients, a systemic inflammatory syndrome is
also present. The systemic and the vascular com-
ponents of giant cell arteritis seem to have differ-
ent underlying pathogenetic mechanisms: the
vascular inflammation results from abnormal
adaptive immune responses, whereas the systemic
inflammation likely depends on an excessively
activated innate immune system [13]. Despite
increased understanding of the inflammatory cas-
cade responsible for the disease process, the initial
event triggering the cascade remains uncertain;
the adventitia is more likely considered as the
site of initial immunologic injury [14].
According to the American College of Rheu-
matology [15], the criteria for assessing the diag-
nosis are:
• Age at disease onset >=50 years
• New headache
• Temporal artery abnormality (temporal artery
tenderness to palpation or decreased pulsation,
unrelated to arteriosclerosis of cervical
arteries)
• Elevated erythrocyte sedimentation rate
• Abnormal artery biopsy (biopsy specimen with
artery showing vasculitis characterized by a
predominance of mononuclear cell infiltration
or granulomatous inflammation, usually with
multinucleated giant cells)
A patient shall be said to have GCA if at least
three of these five criteria are present.
The disease usually has an acute or subacute
start, characterized by symptoms related to the
acute inflammatory status (fever, night sweats)
and headache, jaw pain, and blurred or double
vision; if the disease is not early diagnosed, com-
plications may occur, such as blindness and less
frequently stroke [9].
The diagnostic process starts with clinical
examination and imaging studies. US findings
can show swelling of the vessel’s wall as a
hypoechoic dark halo around the color-coded
flow in the artery; the disease is segmental; there-
fore, its visualization is suitable for localization of
the biopsy [16]. CT and MR imaging can show an
42
abnormal wall enhancement of the affected arter-
ies after contrast media injection in the acute
phase, whereas vascular stenosis and aneurysm
can be observed in the subacute-chronic phase.
In the chronic phase, the differential diagnosis
between GCA and atherosclerotic disease may
not be easy, but the different localization of the
abnormalities (temporal and ophthalmic arteries
for GCA, ubiquitarious for atherosclerosis) can
lead to the correct diagnosis. Standard test for
definitive diagnosis is biopsy of the temporal
artery, being more samples needed because the
inflammation usually does not involve all parts
of the artery [10].
The prognosis for patients with untreated
GCA is extremely poor; these patients may suffer
blindness or death from myocardial infarction,
stroke, or dissecting aortic aneurysm. On the
contrary, with prompt and adequate therapy, full
recovery is the rule. Symptoms from temporal
arteritis usually improve within days of treatment
with corticosteroids, except for those symptoms
related to an effective vision damage that
occurred before initiation of therapy, which are
often irreversible. The mean duration of treat-
ment is 2 years [17].
Fibromuscular Dysplasia (FMD)
Fibromuscular dysplasia (FMD) is an angiopathy
affecting medium-sized arteries, more frequently
in young women of childbearing age. Among
patients with identified FMD, renal involvement
occurs in 60–75 % and cerebrovascular involve-
ment in 25–30 %; involvement of visceral arteries
and arteries of the limbs is less commonly
observed (about 9 % and 5 %, respectively); in
the case of cerebrovascular localization, the inter-
nal carotid artery is more frequently affected
(C2 segment) [9, 18].
The etiology of FMD is not known, even
though several factors have been considered as
involved in its pathogenesis. More in detail, hor-
monal factors such as estrogen have been pro-
posed; however, although in the US Registry
91 % of registrants were female, clear supporting
P. Lucatelli et al.
epidemiological evidence for the role of female
hormones beyond the sex and age distribution of
FMD has not been described yet. Although no
etiologic genes for FMD have been identified,
evidence supports a genetic basis for susceptibil-
ity to FMD; a few authors reported the potential
inheritance pattern to be autosomal dominant
with variable penetrance. However, several fac-
tors have limited the identification and character-
ization of genes contributing to FMD, such as
disease rarity, variable phenotype, and gene-
environment interactions [19]. FMD is usually
described in terms of the affected arterial layer
and the composition of the lesions; depending
on the type of FMD, the narrowing (stenosis) of
the artery is caused by an excess of either
the fibrous or muscular components of the arte-
rial wall [18].
Many people with FMD do not have any symp-
toms; symptoms can occur if the stenosis is severe
enough to restrict blood flow through the affected
artery or if dissection occurs. Symptoms of FMD
in the carotid artery include headaches, ringing or
“swishing” noise in the ears, or light-headedness;
advanced cases of FMD can cause stroke or a
transient ischemic attack [9].
FMD is often accidentally diagnosed when the
beaded appearance in the arteries is observed dur-
ing examinations performed for other reasons.
Noninvasive imaging studies such as duplex ultra-
sound, MRA, and CTA can be used to confirm the
diagnosis of FMD and determine the extent of the
lesions. In general, angiographic studies are
performed only when the diagnosis is not clear
or if the patient requires a therapeutic procedure
such as a balloon angioplasty [10].
The most common type of FMD is the medial
fibrodysplasia (75–80 % of cases), affecting the
tunica media; it is characterized by areas of vessel
stenosis alternating with areas of ectasia, resulting
in a classic “beads on a string” appearance on
angiograms. Intimal and perimedial fibroplasias
are less common (10 % of cases, respectively),
caused by collagen deposits in the intima and in
the outer portion of the tunica media; in the inti-
mal type, a concentric, smooth, narrowing (with-
out beads) appearance of arteries can be observed,
3 Nonatherosclerotic Vascular Disease: Biological and Pathological Basis
whereas the perimedial type has a “beads on a
string” appearance, but with “beads showing a
smaller diameter in comparison to the normal
vessel” [20].
When a patient is diagnosed with FMD in the
carotid arteries, additional imaging studies may be
obtained to evaluate the other blood vessels, espe-
cially vertebral arteries, Willis circle, and renal
arteries.
When FMD is present without any symptoms,
it usually does not require intervention; risk fac-
tors for vascular disease, such as high blood pres-
sure, diabetes, and high cholesterol, should be
evaluated and treated, and imaging studies should
be performed at regular intervals to evaluate the
disease progression. Angioplasty is recommended
for patients with FMD of the internal carotid
artery who experience TIAs or stroke related to
severe arterial narrowing. Stenting is rarely nec-
essary only, in the case of carotid or vertebral
artery dissection or carotid aneurysm. Surgery
depends upon the location and the extent of dis-
ease and consists in removing or bypassing the
affected portion of the artery to restore normal
blood flow; reconstructive surgery may be
recommended for patients with an aneurysm of
the internal carotid or vertebral arteries [19, 21].
Moyamoya Syndrome
Moyamoya is a disease causing intimal thickening
of the walls and stenosis of the terminal branches
of the internal carotid arteries bilaterally; the term
“moyamoya” (Japanese for “puff of smoke”)
refers to the angiographic appearance of abnormal
vascular collateral networks that develop adjacent
to the stenotic vessels [22, 23]. Moyamoya dis-
ease occurs primarily in Asians; the female-to-
male ratio of moyamoya disease is 1.8:1. The
disease has two peaks of incidence in the first
and fourth decades of life [8, 24]. The cause of
moyamoya disease is not clear. The disease is
believed to be hereditary; a few authors suggested
that the transmission may be autosomal dominant
with incomplete penetrance based on age and
genomic imprinting factors [25].
43
According to the Classification of the Japanese
Health Ministry, there are four clinical forms of
the moyamoya disease: ischemic, hemorrhagic,
epileptic, and “other.” Clinical presentation in
children is usually ischemic, related to occlusion
of internal carotid artery or one of the branches of
the Willis circle, resulting in paresis, sensory
impairment, involuntary movements, headaches,
dizziness, or seizures; mental retardation is often
present. In adults, the hemorrhagic form, espe-
cially with subarachnoid hemorrhage, is more fre-
quently observed as a result of hemorrhage of
fragile vessels [9, 26].
In the affected cerebral vessels, pathological
examinations do not show atherosclerotic or
inflammatory lesions. Currently, the major pro-
teins believed to be involved in the pathogenesis
are vascular endothelial growth factor (VEGF),
basic fibroblast growth factor (bFGF), hepatocyte
growth factor (HGF), transforming growth factor-
β (TGF-β), and granulocyte colony-stimulating
factor (G-CSF) [27]. The cause of stenosis is the
overgrowth of the smooth muscle layer, with
thrombotic changes. The disease leads to different
degrees of stenosis and occlusions of arteries of
the anterior part of the Willis circle and to the
development of the collateral vasculature. The
vessels of the collateral circulation are formed as
a result of the widening of the existing vessels
or development of new perforating arteries. There
are three main pathways of collateral circulation –
parenchymal, meningeal, and transdural. In the
vessels of the collateral circulation, there may
appear thrombotic changes, which are the cause
of ischemic symptoms. An increased blood flow
through thin collateral walls during stress, as well
as the presence of microaneurysms, is the proba-
ble cause of intracranial hemorrhages [28].
Before CTA or MRA were introduced to a wide
clinical practice, the final diagnosis of the vascular
changes was based on cerebral angiography.
Although CT examination is sufficient to diagnose
ischemic or hemorrhagic stroke in the course of the
disease, MRA can be considered as the first imaging
technique for definitive diagnosis, considering the
young age of the patients [29]. To support the diag-
nosis, the following findings should be observed:
44
• Stenosis or occlusion at the terminal portion of
the internal carotid artery or the proximal por-
tion of the anterior or middle cerebral arteries
• Abnormal vascular networks in the vicinity of
the occlusive or stenotic areas
• Bilaterality of the findings (although some
patients may initially present with unilateral
involvement)
Mortality rates from moyamoya disease are
around 10 % in adults and 4 % in children.
About 50–60 % of affected patients experience a
gradual deterioration of cognitive function, likely
due to recurrent strokes. The outcome of the dis-
ease depends on the severity of the hemorrhage,
the prognosis on recurrent attacks. Cases of mild
clinical course are normally treated conserva-
tively. In severe cases, surgery is indicated,
including direct anastomoses, indirect procedures,
and combined therapies [30].
Extracranial Internal Carotid Artery
Aneurysms
Aneurysms of the extracranial internal carotid artery
are extremely rare, being more commonly observed
in males (prevalence ranging from 0.1 % to 3.7 %);
most of them are of atherosclerotic origin. In case of
nonatherosclerotic etiology, they can be congenital
(due to collagen disorders, FMD), infectious (TBC,
HIV, mycotic), and posttraumatic. True aneurysms
must be differentiated from pseudoaneurysms,
which are usually iatrogenic (due to carotid punc-
ture or endarterectomy). For a definitive diagnosis
of aneurysm, the diameter of the ectatic artery must
increase of at least 50 %, if compared to the normal
vessel diameter [9, 10].
Most of these aneurysms are asymptomatic,
whereas TIA or stroke is the most common
cause of hospital admission, related to emboliza-
tion from the aneurysmal contents. Spontaneous
rupture or bleeding is very rare, but fatal compli-
cations are seen particularly in cases with mycotic
aneurysms [31].
The mainstay of treatment of extracranial
carotid artery aneurysms is surgical repair; it con-
sists in resecting the portion of the carotid artery
P. Lucatelli et al.
involved with the aneurysm and performing a
bypass from the normal artery below the aneu-
rysm to the normal artery above the aneurysm.
Stroke and cranial nerve injuries (particularly XII
pair) are potential risks for surgical treatment of
aneurysms with a diameter of more than 3 cm
and/or extending to the skull base; endovascular
treatment may be preferred in appropriate cases as
an alternative to surgical therapy; since there is no
risk of cranial nerve injury, it allows to treat
lesions that are hard to reach surgically, and
there is no need for general anesthesia [31].
Craniocervical Arterial Dissection
Craniocervical artery dissection (CCAD) is one of
the major causes of ischemic symptoms in young
adults. It has a prevalence of 2.6 cases per
100, 000 persons per year; vertebral artery dissec-
tions are less common than carotid artery dissec-
tions. Genetic factors such as constitutional
weakness of the arterial wall (such as in FMD or
in monogenic connective tissue disease, mainly
Ehlers-Danlos syndrome or Marfan’s syndrome)
might have a role in the pathophysiology of
CCAD; environmental factors such as minor
trauma act as a trigger [32]. Arterial dissections
begin with a tear in the intima or media resulting
in bleeding in the arterial wall; expansion of the
wall by intramural blood causes compression and
narrowing of the lumen, contributing to the for-
mation of an intraluminal thrombus. The intramu-
ral hematoma can create a false lumen that might
reconnect with the true lumen and forms parallel
flow; alternatively, wall rupture through the
adventitia causes extraluminal bleeding. CCAD
can be asymptomatic and discovered during rou-
tine examinations. The most common symptom is
pain in the head and neck and in the region of
dissections, usually following a minor trauma. If
the dissection compromises the arterial lumen or
causes thrombus formation, clinical symptoms are
related to ischemia; the closer the dissection to the
brain is, the higher the possibility of brain infarc-
tion is present. On the contrary, if the dissection is
more extracranial, the probability of local symp-
toms from space-occupying lesions is higher;
3 Nonatherosclerotic Vascular Disease: Biological and Pathological Basis
bleeding in the subadventitial wall results in com-
pression of the adjacent structures, such as lower
cranial nerves. Patients with subadventitial intra-
cranial dissections often present with subarach-
noid hemorrhage [32, 33].
CCAD can be noninvasively diagnosed by
performing MRA and CTA [34]; color Doppler
flow imaging showed good results in visualization
of dissections, even though the main limitation is
represented by intracranial dissection (which is
the most common site of localization). The tradi-
tional method for CCAD diagnosis is catheter
angiography that may show irregular luminal
narrowing occlusion, pseudoaneurysm, intimal
flap, double lumen, or distal branch
occlusion [35].
There is no general consensus regarding opti-
mal management of internal carotid artery dissec-
tion, but the choice among medical, endovascular,
and surgical options may depend on the type of
and anatomic location of dissection; while surgery
has a limited role, candidates for angioplasty and
stent placement include patients with persistent
ischemic symptoms despite adequate
anticoagulation, patients with a contraindication
to anticoagulant therapy, and patients with signif-
icantly compromised cerebral blood flow [36, 37].
Overall, the prognosis for spontaneous internal
carotid artery dissection is favorable, with about
75 % of patients making a good recovery; the
reported mortality is less than 5 %. The functional
outcome is also generally good, and recurrence of
cerebral ischemia and carotid artery dissection is
rare; the risk of recurrence is highest in the first
month and then remains around 1 % per year
for about a decade. Headache may persist for a
long time, in some cases for years, after the
dissection [38].
Summary
Atherosclerosis is the main cause of vascular dis-
ease, being responsible in around 90 % of cases of
carotid disease; in the remaining cases,
nonatherosclerotic causes such as Takayasu arter-
itis, giant cell arteritis, fibromuscular disease,
moyamoya syndrome, arterial dissection and
45
extracranial carotid aneurysm must be taken into
account. These less common diseases have some
similarities, such as an unclear pathogenesis and a
good outcome if the specific disease is diagnosed
early.
References
1. Jennette JC (1994) Nomenclature of systemic vasculi-
tis. Arthritis Rheum 37:187–192
2. Numano F et al (1996) Takayasu arteritis-five doctor in
the history of Takayasu arteritis. Int J Cardiol
54:S1–S10
3. Shimizu K, Sano K (1951) Pulseless disease.
J Neuropathol Clin Neurol 1.7–1.47
4. Hall S et al (1985) Takayasu arteritis. A study of
32 North American patients. Medicine 64:89–99
5. Arnaud L, Haroche J, Mathian A (2011) Pathogenesis
of Takayasu’s arteritis: a 2011 update. Autoimmun Rev
11(1):61–67
6. Matsunaga N, Hayashi K, Sakamoto I et al (1997)
Takayasu arteritis: protean radiologic manifestations
and diagnosis. Radiographics 17:579–594
7. Arend WP et al (1990) The American College of Rheu-
matology 1990 criteria for the classification of
Takayasu arteritis. Arthritis Rheum 33:1129–1134
8. Hata A, Noda M, Moriwaky R et al (1996) Angio-
graphic findings of Takayasu arteritis: new classifica-
tion. Int J Cardiol 54(suppl):S155–S163
9. Bartels E, Bartels S, Poppert H (eds) (2012) New trends
in neurosonology and cerebral hemodynamics-an
update. Perspect Med 1:244–249
10. Catalano C, Anzidei M, Napoli A (editors): (2013)
Cardiovascular CT and MR imaging. Springer-Verlag
Mailand Italia Srl; ISBN 978-88-470-2868-5.
11. Phillip R, Luqmani R (2008) Mortality in systemic
vasculitis: a systematic review. Clin Exp Rheumatol
26(5 Suppl 51):S94–S104
12. Salvarani C, Gabriel SE, O’Fallon WM, Hunder GG
(1995) The incidence of giant cell arteritis in Olmsted
County, Minnesota: apparent fluctuations in a cyclic
pattern. Ann Intern Med 123(3):192–194
13. Weyand CM, Goronzy JJ (2003) Giant-cell arteritis and
polymyalgia rheumatica. Ann Intern Med
139:505–515
14. Weyand CM, Goronzy JJ (2000) Pathogenic principles
in giant cell arteritis. Int J Cardiol 75(Suppl 1):S9–S15,
discussion S17–S9
15. Hunder GG, Bloch DA, Michel BA, Stevens MB,
Arend WP, Calabrese LH et al (1990) The American
College of Rheumatology 1990 criteria for the classi-
fication of giant cell arteritis. Arthritis Rheum
33:1122–1128
16. Schmidt WA, Kraft HE, Vorpahl K (1997) Color
duplex ultrasonography in the diagnosis of temporal
arteritis. N Engl J Med 337:1336–1342
46
17. Borchers AT, Gershwin ME (2012) Giant cell arteritis:
a review of classification, pathophysiology,
geoepidemiology and treatment. Autoimmun Rev
11(6–7):A544–A554
18. Luscher TF, Lie JT, Stanson AW et al (1987) Arterial
fibromuscular dysplasia. Mayo Clin Proc
62(10):931–952
19. Olin JW, Gornik HL, Bacharach JM et al (2014)
Fibromuscular dysplasia: state of the science and crit-
ical unanswered questions: a scientific statement from
the American Heart Association. Circulation
129(9):1048–1078
20. Begelman SM, Olin JW (2000) Fibromuscular dyspla-
sia. Curr Opin Rheumatol 12(1):41–47
21. Olin JW, Pierce M (2008) Contemporary management
of fibromuscular dysplasia. Curr Opin Cardiol
23(6):527–536
22. Kuroda S, Houkin K (2008) Moyamoya disease: cur-
rent concepts and future perspectives. Lancet Neurol
7(11):1056–1066
23. Scott R, Smith ER (2009) Moyamoya disease and
moyamoya syndrome. N Engl J Med 360:1226–1237
24. Uchino K, Johnson A, Claiborne S, Tirschwell DL
(2005) Moyamoya disease in Washington State and
California. Neurology 65:956–958
25. Mineharu Y, Takenaka K, Yamakawa H et al (2006)
Inheritance pattern of familial moyamoya disease:
autosomal dominant mode and genomic imprinting.
J Neurol Neurosurg Psychiatry 77(9):1025–1029
26. Shamim S, Kumar J, Jamalvi SW et al (2008) Moya
Moya disease in a child. J Coll Physicians Surg Pak
18(4):252–253
27. Weinberg DG, Arnaout OM, Rahme RJ et al (2011)
Moyamoya disease: a review of histopathology, bio-
chemistry, and genetics. Neurosurg Focus 30(6):E20
P. Lucatelli et al.
28. Lim M, Cheshier S, Steinberg G (2006) New vessel
formation in the central nervous system during tumor
growth, vascular malformations, and Moyamoya. Curr
Neurovasc Res 3(3):237–245
29. Tarasów E, Kułakowska A, Łukasiewicz A et al (2011)
Moyamoya disease: diagnostic imaging. Pol J Radiol
76(1):73–79
30. Burke GM, Burke AM, Sherma AK et al (2009)
Moyamoya disease: a summary. Neurosurg Focus
26(4):E11
31. Yetkin U, Y€ urekli, G€
urb€
uz A (2006) Extracranial inter-
nal carotid artery aneurysms. Internet J Thorac
Cardiovasc Surg 10(2)
32. Debette S, Leys D (2009) Cervical artery dissections:
predisposing factors, diagnosis and outcome. Lancet
Neurol 8:668–678
33. Caplan LR (2008) Dissections of brain-supplying
arteries. Nat Clin Pract Neurol 4:34–42
34. Debette S, Provenzale JM (2009) MRI and MRA for
evaluation of dissection of craniocerebral arteries: les-
sons from the medical literature. Emerg Radiol
16:185–193
35. Houser OW, Mokri B, Sundt TM et al (1984) Sponta-
neous cervical cephalic arterial dissection and its resid-
uum: angiographic spectrum. Am J Neuroradiol
5:27–34
36. Baumgartner RW (2010) Management of spontaneous
dissection of the cervical carotid artery. Acta Neurochir
Suppl 107:57–61
37. Xianjun H, Zhiming Z (2013) A systematic review of
endovascular management of internal carotid artery
dissections. Interv Neurol 1(3–4):164–170
38. Schievink WI (2001) Spontaneous dissection of the
carotid and vertebral arteries. N Engl J Med
344(12):898–906
 Essentials of Transcranial Doppler
Ultrasound 4
Jonathan D. Kirsch

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Transcranial Doppler (TCD) ultrasound imag-
ing is a noninvasive, portable imaging tech-
Anatomy and Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
nique that can be used to evaluate the
Physics of Doppler Ultrasound . . . . . . . . . . . . . . . . . . . . . 49 intracerebral arteries. TCD ultrasound has
Applications of Transcranial Doppler become an important modality in the monitor-
Ultrasound . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 ing and evaluation of vasospasm of the intra-
Vasospasm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 cerebral arteries in patients with subarachnoid
Sickle Cell Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Brain Death . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
hemorrhage due to aneurysmal rupture.
Internal Carotid Artery Occlusion . . . . . . . . . . . . . . . . . . . . 58 Advances in transcranial ultrasound imaging
Other Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 utilizing gray-scale, color Doppler flow, and
Ultrasound Biosafety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 spectral Doppler allow for direct visualization
and flow velocity measurements within the
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
vessels in real time. This is a significant
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 improvement over the older “blind” technique
which utilized a range-gated Doppler probe
that inferred the vessels being interrogated by
a complex set of parameters including vessel
depth, flow direction, and probe orientation.
TCD has also been found to be useful in the
evaluation of patients with sickle cell disease
who may be at the risk of stroke and determi-
nation of brain death, internal carotid artery
(ICA) occlusion and collateral pathways, ste-
nosis, arteriovenous malformations, and intra-
cardiac right-to-left shunts.

Introduction
J.D. Kirsch (*)
Transcranial Doppler (TCD) ultrasound has
Department of Diagnostic Radiology, Yale University
School of Medicine, New Haven, CT, USA emerged as an important technique for imaging
e-mail: jonathan.kirsch@yale.edu of the intracranial vasculature. Originally
# Springer Science+Business Media New York 2016 47
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_36
48
developed as an indirect “blind technique” involv-
ing range-gated spectral Doppler, the technique
has advanced in recent years with the ability to
image and visualize the vessels directly utilizing
gray-scale, color Doppler, and spectral Doppler
imaging. TCD has found its greatest application
and utility in the neurosurgical intensive care unit
in monitoring for vasospasm in patients with
intracranial bleeds due to ruptured aneurysms.
However, the technique has also found promising
and potential uses in the evaluation of stenosis,
internal carotid artery occlusion, brain death,
intracranial arteriovenous malformations, intra-
cardiac right-to-left shunts, and sickle cell
disease [1].
Although the intracranial vessels can be
imaged by other techniques such as conventional
angiography, CT angiography (CTA), or MR
angiography (MRA), these techniques require
transport of the patient to the diagnostic radiology
department. Conventional angiography is inva-
sive in nature with the potential risk of vessel
damage or stroke. All three techniques involve
the use of contrast agents which carry the risk of
anaphylactoid reaction for MR and CT contrast
agents, nephrotoxicity for CT contrast, and
nephrogenic systemic fibrosis with MR contrast.
CTA and conventional angiography, in addition,
require exposure to ionizing radiation.
Transcranial Doppler ultrasound, on the other
hand, is a noninvasive imaging modality that can
be performed portably, if necessary, in the ICU
setting. TCD does not require the use of contrast
or ionizing radiation and serial follow-up exami-
nations can be easily performed. This chapter will
review the relevant anatomy, physics, and tech-
niques of transcranial Doppler ultrasound and the
interpretation of the findings.
Anatomy and Technique
The circle of Willis is an anastomotic ring of
arteries that sits at the base of the brain encircling
the sella turcica and pituitary gland. It provides
communication between the internal carotid
arteries and the vertebrobasilar system. The ante-
rior portion of the circle is comprised of the
J.D. Kirsch
internal carotid arteries (ICA) which bifurcate
into the middle and anterior cerebral arteries
near the skull base (MCA, ACA). The anterior
cerebral arteries communicate via the anterior
communicating artery (AComm). Posteriorly,
the vertebral arteries join to form the basilar
artery which then divides into the posterior
cerebral arteries (PCA). The posterior communi-
cating arteries complete the circle joining the
internal carotid artery system anteriorly with
the posterior circulation.
Transcranial Doppler imaging can be utilized
to visualize the circle of Willis, the ophthalmic
artery, and the cephalad portions of the vertebral
and basilar arteries at the skull base.
Imaging of the circle of Willis can be achieved
by transcranial Doppler imaging. Before the
advent of high-resolution gray-scale and color
Doppler flow imaging, Aaslid et al. demonstrated
that cerebral arterial flow velocities could be mea-
sured using pulsed Doppler US through the tem-
poral bone or the foramen magnum [2]. The
technique was “blind” in that the vessels them-
selves were not directly visualized. The identity of
the vessel being interrogated was determined by
the depth of the vessel from the skull, its flow
direction, and the transducer orientation in rela-
tion to the skull [3]. Using this complex approach,
Aaslid was able to measure and determine normal
mean flow velocity (MFV) ranges of the intracra-
nial arteries (Table 1). This technique had obvious
limitations related to lack of visualization of the
vessels and the inability to angle correct the mea-
surements of flow velocity (see section “Physics
of Doppler Ultrasound”).
With advances in ultrasound equipment and
the utilization of gray-scale, color Doppler flow,
and spectral Doppler imaging, direct visualization
of the vessel and more accurate flow velocity
measurements are now possible. Imaging of the
circle of Willis is performed via a transtemporal
approach utilizing a 5–1 MHz sector-array trans-
ducer with most of the imaging being performed
at the lower frequencies (1–2 MHz) to allow pen-
etration of the temporal bone. The transducer is
placed on the temporal bone either above the
zygomatic arch and anterior to the external audi-
tory canal or slightly more cephalad and posterior
4 Essentials of Transcranial Doppler Ultrasound 49

Table 1 Parameters for evaluation of the intracranial vessels with range-gated spectral Doppler. Before the advent
of transcranial Doppler utilizing direct visualization of the vessels with gray-scale and color Doppler flow imaging, the
cerebral vessel being interrogated by spectral Doppler had to be inferred from multiple parameters as noted in the above
table. The vessel being measured was inferred from its depth from the skull, its flow direction with respect to the
transducer, and the transducer orientation to the vessel and skull. Mean flow velocity (MFV) ranges shown are those
determined by Aaslid et al. [2]
Flow Transducer
Artery Depth (mm) Window direction orientation MFV (cm/s)
Ophthalmic artery 40–50 Orbital Toward Slightly medial 16–26
Middle cerebral artery 35–60 Temporal Toward En face 46–86
(MCA)
Anterior cerebral artery 60–75 Temporal Away Anterior 41–76
(ACA)
Posterior cerebral artery 60–75 Temporal Toward Posterior 33–64
(PCA)
Vertebral artery 45–75 Transforaminal Away Superior and 27–55
oblique
Basilar artery 70–120 Transforaminal Away Superior 30–57

over the earlobe. Imaging is performed from the
left and right temporal bones with color Doppler
flow imaging utilized to visualize the vessels.
Flow velocity measurements (both MFV and
peak systolic velocity) are obtained. With the
transtemporal approach, visualization of the
MCA, ACA, PCA, and terminal ICA is possible.
At our institution, spectral Doppler waveforms
and flow velocities are obtained from the proxi-
mal, mid, and distal MCA and single measure-
ments are obtained in the visualized portion of the
ACA, PCA, and terminal ICA.
The basilar and distal vertebral arteries can be
visualized via a transforaminal approach. The ver-
tebral arteries enter the skull base through the
foramen magnum and then join to form the basilar
artery at the base of the medulla oblongata. The
basilar artery courses cephalad, anterior to the
medulla and pons, in the prepontine cistern. To
visualize these vessels utilizing the transforaminal
window, the transducer is placed in the midline
below the occiput and angled cephalad (Fig. 1).
The ophthalmic artery and carotid siphon of
the ICA can be visualized via the transorbital
window. The probe is lightly placed on the closed
eyelid with a small amount of gel applied to the
closed eyelid to aid in acoustic transmission
(Fig. 2).
The open fontanelles in infants provide an
excellent acoustic window for visualizing the
intracerebral vessels. High-frequency (5–10
MHz) linear, curved, or sector transducers can be
used depending on the size of the child. Utilizing
the anterior fontanelle, the ICA, MCA, ACA, and
anterior communicating artery (AComm) can be
visualized in the coronal plane (Fig. 3). Midline
sagittal imaging allows visualization and flow
velocity measurements from the ACA,
pericallosal, callosomarginal, and frontopolar
arteries in addition to the vein of Galen (Fig. 4).
Patency of the superior sagittal sinus can be dem-
onstrated through the anterior fontanelle or an
open sagittal suture in the coronal or sagittal
planes (Fig. 5). The posterior circulation can be
imaged through a transforaminal approach, as in
the adult, or through an open posterolateral fonta-
nelle just posterior to the mastoid process.
Physics of Doppler Ultrasound
An understanding of the basic underlying physical
principles related to spectral Doppler ultrasound is
essential for the interpretation of transcranial
Doppler measurements and results. All vascular
studies utilizing spectral Doppler make use of two
well-known physical principles when measuring
flow velocities and evaluating for stenosis and/or
vasospasm: the Doppler effect and Bernoulli’s
principle.
50
Fig. 1 Transtemporal and transforaminal acoustic win-
dows (a). The transducer is placed superior to the zygo-
matic arch and anterior to the external auditory canal or
more cephalad above the earlobe for the transtemporal
approach. For the transforaminal approach, the transducer
is placed in the midline below the occiput and angled
cephalad. Utilizing the transtemporal approach and color
The Doppler effect, first described in 1842 by
the physicist Christian Doppler, relates the change
in frequency of a sound wave observed (or other
periodic event) if the source of the wave is moving
relative to the observer. This effect is commonly
seen in everyday life when an automobile with a
siren passes by. The sound waves increase in
frequency as the car approaches the observer and
decreases in frequency as the car passes by
resulting in a frequency shift in the sound of the
siren heard by the observer. For flowing blood, a
similar shift in frequency is noted when interro-
gated by the sound beam from an ultrasound
transducer. The sound wave is emitted from the
J.D. Kirsch
Doppler flow imaging, the circle of Willis can be visualized
(b). The transforaminal approach allows visualization of
the basilar and vertebral arteries (c). ACA anterior cerebral
artery, ICA terminal internal carotid artery, MCA middle
cerebral artery, PCA posterior cerebral artery, PComm pos-
terior communicating artery
transducer at a certain frequency, fo, and is then
reflected back to the transducer from the moving
red blood cells at a different frequency, fr. This
frequency shift ( fd = fr
fo) and its relation to
velocity can be expressed by the following
equation:
c  fd
v¼
2 cos θ  fo
where v is the velocity of the flowing blood, c is
the speed of the sound wave, and θ is the angle
between the incident beam of the sound wave
arising from the transducer and the direction of
4 Essentials of Transcranial Doppler Ultrasound 51

Fig. 2 Transorbital acoustic window: The transducer is carotid siphon can be visualized by color Doppler flow
placed lightly on the closed eyelid (a) and utilizing the imaging (b). OA ophthalmic artery (Image used with per-
globe as an acoustic window, the ophthalmic artery and mission from Kirsch et al. [1])

Fig. 3 Coronal color Doppler flow imaging through the anterior cerebral artery (ACA), M1 segments of the middle
circle of Willis (a, b), utilizing the anterior fontanelle as an cerebral arteries (MCA), and the intracranial internal
acoustic window, demonstrates the A1 segments of the carotid artery (ICA)
52
blood flow within the vessel (Fig. 6). From the
equation, it becomes apparent that the most accu-
rate flow velocities would be obtained with the
transducer parallel to the direction of blood flow
as θ = 0 and cos θ = 1. The greatest error in
measurement and loss of signal would occur as
the angle approached 90 and cos θ approaches
zero. It is recommended that velocity measure-
ments be obtained with θ < 60 to obtain the
most accurate measurements.
The Doppler equation allows one to not only
calculate the velocity of the flowing blood but also
Fig. 4 Sagittal midline color Doppler flow imaging
through the anterior fontanelle demonstrates the
pericallosal artery, the callosomarginal artery (CM), the
anterior cerebral artery (ACA), the vein of Galen, and the
corpus callosum (CC)
Fig. 5 Coronal gray-scale (a) and sagittal color Doppler flow imaging (b) through the anterior fontanelle demonstrate the
superior sagittal sinus (arrows)
J.D. Kirsch
its direction. A shift to a higher frequency is a
positive Doppler shift and indicates flow toward
the transducer. A shift to a lower frequency is a
negative Doppler shift and indicates flow away
from the transducer. This frequency shift and
directionality can be color-coded to indicate flow
direction. Conventionally with color Doppler
flow imaging, red is used to indicate flow toward
the transducer and blue for away; however, this
color designation is arbitrary and can be switched
so attention must be paid to the color bar on the
image. The color on the top of the bar is toward
the transducer and the color below is away
(Fig. 7).
Bernoulli’s principle, as applied to flowing liq-
uids, can be utilized for the diagnosis of areas of
stenosis and vasospasm. According to the princi-
ple, as fluid flows through conduits of differing
diameters, the velocity and pressure associated
with the fluid will change. As fluid flows from a
conduit (or blood vessel) of greater diameter to
one of smaller diameter, the flow velocity must
increase to allow the same volume of blood to
flow through the narrower area (this is analogous
to one putting one’s thumb over the opening of a
garden hose as water is flowing out resulting in a
high-velocity jet of water being formed).
Bernoulli’s principle and the law of conservation
of energy also state that the pressure in the region
of increased velocity will decrease (Fig. 8). The
increase in velocity noted in areas of narrowing is
4 Essentials of Transcranial Doppler Ultrasound
Fig. 6 Doppler effect. The velocity of blood flow can be
calculated utilizing the Doppler effect. The ultrasound trans-
ducer emits a sound wave pulse at a certain frequency, fo,
which is reflected off the moving red blood cells back to the
transducer at a different frequency, fr. This resulting fre-
quency shift ( fd = fr
fo) can be used in the Doppler
equation to calculate the velocity of the flowing blood. The
angle, θ, is the angle between the incident sound wave beam
and the direction of the flowing blood. The most accurate
flow velocity measurements are obtained with θ < 60
(Image used with permission from Kirsch et al. [1])
Fig. 7 Color Doppler flow imaging allows one to visual-
ize vessels more easily and also encode flow direction by
the color map. The color bar in the upper right-hand corner
(thick arrow) denotes flow direction. Although the color
designation is arbitrary and can be reversed, the colors on
the top half of the bar codes are for flow toward the
transducer (red) and the colors on the bottom, away
(blue). In this example, the right MCA is flowing toward
the transducer and the left MCA and PCAs are
flowing away
53
a mainstay for diagnosing stenosis or vasospasm
in blood vessels. As described above, these flow
velocities can be easily calculated by the
Doppler equation. As elucidated by Spencer and
Reid [4], there is a fairly linear relationship
between flow velocity and the degree of stenosis
(up to a critical stenosis at which point flow veloc-
ities decline due to extreme narrowing of the
vessel). In general, the higher the velocity of the
blood flow, the greater the degree of stenosis or
vasospasm.
Spectral Doppler allows measurements of flow
velocity in a specified area of a vessel (the sample
volume) over time and displays these flow veloc-
ities as a time/velocity curve throughout the car-
diac cycle. From this waveform, peak systolic
velocity (PSV), the velocity of the flowing blood
at the peak of systole on the curve, and
end-diastolic velocity (EDV), the velocity just
before the next systolic upstroke, can be mea-
sured. The spectral waveform also indicates direc-
tionality of flow. Conventionally, a waveform
above the baseline (positive) indicates flow
toward the transducer and a waveform below the
baseline (negative) indicates flow away. Again,
close attention should be paid to the scale as this
can be inverted on the screen. The spectral line
and the area underneath it (the spectral envelope)
can indicate the type of flow present. A narrow
spectral line with a clear envelope beneath it indi-
cates fairly uniform (laminar) flow with a narrow
range of velocities. Filling in of the spectral win-
dow indicates a broader range of flow velocities
within the sample volume consistent with turbu-
lent flow seen with stenosis or narrowing (Fig. 9).
The amount of flow during the different portions
of the cardiac cycle provides information as to
whether the vessel is feeding a high-resistance
system such as a muscle bed or extremity or a
low-resistance system such as an end organ. High-
resistance waveforms will have little to no flow
during diastole, whereas low-resistance wave-
forms will have more abundant forward flow
during diastole.
The resistive index (RI) is a measure of
the resistance to the flow of blood into an
organ and is calculated utilizing the following
formula:
54
Fig. 8 Bernoulli’s principle. As fluid (blood) flows
through conduits of differing diameters, the velocity and
pressure of the fluid will change. As blood flows from a
vessel of greater to narrowed diameter (either due to vaso-
spasm or stenosis), the velocity of the blood will increase in
Fig. 9 Spectral Doppler waveform. The spectral Doppler
waveform is a time/velocity curve that demonstrates the
blood flow velocities over time within a defined sample
window in the vessel being evaluated. The area below the
curve or spectral line is called the spectral envelope or
window. With turbulent blood flow, this window will “fill
in” representing the myriad flow velocities associated with
turbulent flow (as seen in this example). The spectral
Doppler also can be used to measure peak systolic velocity
(thick arrow) and end-diastolic velocity (thin arrow) from
which the mean flow velocity and the resistive index can be
calculated (table to left of CDF image). A waveform above
the baseline (positive) indicates flow toward the transducer
and a waveform below indicates flow away from the
transducer
ðPSV
EDVÞ
RI ¼
PSV
In low-resistance waveforms, EDV will be fairly
high as there is forward flow of blood into the
organ during diastole and the ratio will, therefore,
be lower. In organs or systems that have a high
resistance to blood flow, EDV will fall toward
zero and the ratio will approach 1.
J.D. Kirsch
the area of narrowing and its pressure will decrease. This
principle, along with the Doppler effect, allows for ultra-
sonographic detection of areas of vasospasm or stenosis
(Image used with permission from Kirsch et al. [1])
Mean flow velocity (MFV) can be calculated
from the spectral Doppler waveform and is essen-
tially the average flow velocity over the time/
velocity waveform. Thresholds for PSV and
MFV have been determined that correlate with
varying degrees of vasospasm (see section
“Vasospasm”).
Applications of Transcranial Doppler
Ultrasound
Vasospasm
Transcranial Doppler ultrasound has found its
greatest application in the neurosurgical intensive
care unit screening patients for vasospasm sec-
ondary to subarachnoid hemorrhage (SAH) from
aneurysmal rupture. Vasospasm is used to
describe both the clinical picture of delayed-
onset neurological deficits following SAH and
the imaging findings of vessel narrowing. Vaso-
spasm from SAH has been associated with a
15–20 % risk of stroke or death [5]. About half
of those affected will demonstrate clinical signs
and symptoms. It has been reported that vaso-
spasm doubles the risk for mortality in SAH
[6–9].
Vasospasm following SAH is generally absent
in the first 72 h post bleed. Vasospasm generally
occurs by day 3, peaking between days 6 and
12, and then resolving 15–20 days after the
bleed [10]. The clinical findings associated with
vasospasm usually manifest themselves 4–5 days
4 Essentials of Transcranial Doppler Ultrasound 55

post hemorrhage. Insidious onset of confusion Table 2 MCA evaluation for vasospasm: Mean flow
and decreasing levels of consciousness are the velocity, peak systolic velocity, and Lindegaard ratio cut-
offs used at our institution for evaluation of varying
initial presenting signs which may progress to degrees of vasospasm in the middle cerebral arteries
focal neurologic defects, infarction, coma, and
Severity of MFV PSV Lindegaard
death. The severity and extent of vessel narrowing vasospasm (cm/s) (cm/s) ratio
has been seen to correlate with the clinical Mild 120–150 200–250 3.0–4.5
findings. vasospasm
The pathogenesis of cerebral vasospasm is Moderate 150–200 250–300 4.5–6.0
poorly understood but results primarily from vasospasm
prolonged smooth muscle contraction. Data sug- Severe >200 >300 >6.0
vasospasm
gests that calcium-dependent and calcium-
independent vasoconstriction is taking place.
Breakdown products from the blood in the sub-
arachnoid space may also be playing a direct or
indirect role. Other factors such as an imbalance Table 3 ACA and PCA evaluation for vasospasm:
between vasoconstrictors and vasodilators includ- Mean flow velocity and peak systolic velocity parameters
ing nitric oxide, prostaglandins, prostacyclin, free used at our institution for evaluation of the presence of
vasospasm in the anterior and posterior cerebral arteries.
radicals, and endothelin may also be involved in The Sloan ratio is calculated by dividing the MFV of the
the pathogenesis of vasospasm following ACA by the MFV of the distal extracranial ICA (NA not
SAH [11]. applicable)
Screening of patients at risk for vasospasm is MFV PSV Sloan
essential as patients cannot be placed prophylac- Vessel (cm/s) (cm/s) ratio
tically on treatment for potential vasospasm. Med- Anterior cerebral >80 >120 >4.0
artery
ical treatment for vasospasm consists of “triple-H
Posterior cerebral >85 >120 NA
therapy”: induced hypertension, hemodilution,
artery
and hypervolemia which are all aimed at improv-
ing cerebral perfusion. Calcium channel blockers,
such as nimodipine, have also been recommended the velocity, the greater is the degree of
but should be used cautiously to avoid the delete- vasospasm.
rious effects of hypotension [7]. Unfortunately, For intracranial vessels, cutoff values for PSV
these treatments carry their own inherent risks and MFV have been established that correlate
and potential complications including renewed with various degrees of vasospasm in the MCA
bleeding from the aneurysm site or into areas of and for the presence of vasospasm in the ACA or
infarction, increased cerebral edema, myocardial PCA (Tables 2 and 3). For the MCA, MFV in the
infarction, and congestive heart failure. If medical 120–150 cm/s range correlates with mild vaso-
therapy is ineffective, catheterization of the cere- spasm, whereas MFV in the 150–200 cm/s and
bral vessels can be performed with a vasodilating 200–250 cm/s ranges correlate with moderate and
agent injected. TCDI allows for quick and nonin- severe vasospasm, respectively [2, 4, 12].
vasive screening and monitoring of the patients in Torbey et al. have shown that the age of the
the ICU setting for vasospasm to determine if patient does play a role in the incidence and diag-
medical or interventional treatment is necessary. nosis of cerebral vasospasm. Older patients (>68
Vasospasm results in a narrowing of the vessel years) had a lower incidence of symptomatic cere-
lumen, and as noted above, the velocity of the bral vasospasm and a shorter window in which
blood flow within a vessel is inversely propor- vasospasm may occur (up to 10 days post bleed
tional to its diameter until a critical narrowing for older patients as compared to younger patients
(usually greater than 95 %) is reached at which that demonstrated symptomatic vasospasm as late
point blood flow through the extremely narrowed as 16 days following SAH). With age, cerebral
vessel actually decreases. In general, the higher blood flow velocities decrease and older patients
56
were seen to develop symptomatic vasospasm at
lower MFV than younger patients [13].
Physiologic processes such as autoregulation
and hyperemia and medically induced conditions
including hypertension and hypervolemia may also
result in elevated flow velocities within the intra-
cranial vessels. To help distinguish between these
different etiologies for elevated flow velocity in the
MCA, Lindegaard et al. developed a ratio utilizing
the MFV of the MCA and the MFV of the distal
extracranial ipsilateral ICA (VMCA/VICA).
Lindegaard ratios in the range of 3.0–6.0 are indic-
ative of mild to moderate vasospasm and ratios
>6.0 are indicative of severe stenosis. Elevated
flow velocities with a Lindegaard ratio <3.0 is
Fig. 10 (a) Noncontrast CT demonstrates an aneurysmal
SAH. Left frontal craniectomy has been performed to help
relieve intracranial pressure (*). Transcranial Doppler
imaging (TCDI) demonstrates mild to moderately elevated
peak systolic velocity and mean flow velocities in the left
middle cerebral artery (280 cm/s and 150 cm/s,
J.D. Kirsch
more indicative of an etiology other than vaso-
spasm as the cause of the flow velocity elevation
[14, 15]. An accurate measurement of the extracra-
nial distal ICA MFV is important as slight varia-
tions in this measurement can result in over- or
underestimating the presence of vasospasm. The
velocity should be obtained as close to the skull
base as possible (Fig. 10).
The Sloan ratio (MFV of the ACA divided by
the MFV of the extracranial ipsilateral ICA) can
aid in the detection of vasospasm in the ACA. A
Sloan ratio greater than 4.0 is considered indica-
tive of vasospasm.
Baseline and serial examinations are important
to obtain as elevations of flow velocities within a
respectively). The Lindegaard ratio, however, was mark-
edly elevated (11.5) consistent with severe vasospasm. (b)
3D reconstruction from a CT angiography performed after
the TCD revealed severe vasospasm of the left MCA
(arrow, c) (Images used with permission from Kirsch
et al. [1])
4 Essentials of Transcranial Doppler Ultrasound
Fig. 11 (a) Initial transcranial Doppler of the right ante-
rior cerebral artery (ACA) in a patient with SAH demon-
strates peak systolic and mean flow velocities in the normal
range (67 cm/s and 32 cm/s, respectively). (b) 3D recon-
struction form a CT angiogram performed at this time
vessel between subsequent exams may be indica-
tive of vasospasm. Mean flow velocity increases
of greater than 50 % or absolute increases in
velocity of greater than 50 cm/s over a 24 h period
may indicate vasospasm in the vessel (Fig. 11).
Through the transforaminal window, the distal
vertebral arteries (VA) and basilar artery (BA) can
be visualized and evaluated for vasospasm involv-
ing the posterior circulation. Sviri et al. demon-
strated good sensitivity and specificity utilizing a
ratio of the MFV of the extracranial vertebral artery
and the basilar artery in conjunction with MFV of
the basilar artery. The ratio was calculated using the
average of the mean flow velocities of both
57
showed a normal caliber vessel (arrow). TCDI obtained
10 days later demonstrates the PSV and MFV to have more
than doubled in the same vessel. (c) Repeat CT angiogram
now demonstrates vasospasm of the right ACA (d, arrow)
(Images used with permission from Kirsch et al. [1])
vertebral arteries and dividing this value into the
highest MFV obtained in the basilar artery. A
BA/VA MFV ratio of 2.0–3.0 plus a basilar mean
flow velocity greater than 85 cm/s was found to be
indicative of mild to moderate vasospasm. A ratio
>3.0 was found to be highly accurate in diagnosing
severe vasospasm [16].
Sickle Cell Disease
Sickle cell disease is the result of a coding muta-
tion in the beta chain gene for hemoglobin. The
mutation results in substituting thymine for
58
adenine, GAG to GTG, in the sixth codon. The
resulting abnormal hemoglobin (HbS) under deoxy
conditions forms polymers and demonstrates
increased viscosity and decreased solubility. This
polymer formation in red blood cells (RBCs)
which are homozygous for HbS results in the clas-
sic sickled shape of the RBCs. Factors other than
local tissue hypoxia that can precipitate the sickling
process include dehydration, infection, acidosis,
and cold weather. Recurrent episodes of sickling
result in membrane damage and loss of the ability
of the RBCs to reattain their normal shape. These
properties of HbS and the resulting physiological
changes in RBCs result in the clinical and physical
findings of the disease including hemolytic anemia,
vaso-occlusive crisis, and multiple organ damage
due to microinfarcts [17].
In children, the most common cerebrovascular
manifestation of sickle cell disease is cerebral
infarction due to occlusive vasculopathy. An
approximately 11 % incidence of stroke is seen
in patients with sickle cell disease by the age of
20. Incidence of stroke is highest in the first
decade [18]. Fibrous proliferation of the intima,
usually within the intracranial ICA, proximal
MCA, and ACA, leads to vessel narrowing, occlu-
sion, stroke, and infarction. In stroke-free chil-
dren, those with high cerebral blood flow
velocities in the distal ICA or proximal MCA
had a significantly increased risk of stroke [19].
Children with sickle cell disease and mean
flow velocities below 170 cm/s were considered
at low risk for stroke in the 60 months post TCD.
Patients with MFV in the 171–199 cm/s range
were recommended to have more frequent TCD
testing. Adams et al. have shown that children
with mean flow velocities greater than 200 cm/s
have a significantly increased risk of stroke and
showed significant decrease in the risk of stroke
when treated with blood transfusions that
decreased the HbS concentration to less than
30 % of the total hemoglobin [20, 21].
Brain Death
Transcranial Doppler imaging can aid in the diag-
nosis of brain death. Traumatic brain injury and
J.D. Kirsch
subarachnoid hemorrhage are the chief causes of
brain death in adults, whereas in children, abuse,
motor vehicle accidents, and asphyxia are more
common etiologies [22, 23]. Clinically, the deter-
mination of brain death includes documentation
of coma, absence of brain stem reflexes, and
apnea. Confirmatory tests include cerebral angi-
ography which should demonstrate no filling of
the intracerebral vessels from their point of entry
into the skull, electroencephalography which
should demonstrate absence of electrical activity
and lack of reactivity to intense somatosensory or
audiovisual stimuli, and cerebral scintigraphy
which demonstrate absence of radioisotope within
the brain [12]. TCDI has been found to have a
high sensitivity and specificity for the evaluation
of brain death and can be utilized to evaluate the
intracranial vessels utilizing the transtemporal
approach [24].
Due to increased intracranial pressures,
transcranial Doppler waveforms will demon-
strate initially decreased forward diastolic flow.
As pressure continues to increase, reversed dia-
stolic flow will be noted resulting in a reverber-
ating or oscillating flow pattern. Reversed
diastolic flow, however, is not pathognomonic
for brain death as other entities such as mass-
occupying lesions, large strokes, edema, or her-
niation may also result in increased intracranial
pressure with resulting loss of diastolic flow.
Small, low-velocity systolic peaks or spikes
may be noted (Fig. 12). Care must be taken in
the interpretation of apparent lack or cessation of
flow in the vessels as technical factors such as
poor skull penetration by the ultrasound beam
may result in false-positive exams.
Internal Carotid Artery Occlusion
Atherosclerotic plaques consisting of necrotic
cells, lipids, and cholesterol crystals can result in
stenosis of the internal carotid artery which, in
conjunction with superimposed thrombosis, can
result in complete occlusion of the vessel. Emboli,
originating either from a cardiac source or ather-
oma from the aortic arch, are the other most fre-
quent cause of occlusion. The proximal 2 cm of
4 Essentials of Transcranial Doppler Ultrasound
Fig. 12 Brain death: (a–c) Transcranial Doppler demon-
strates reversed diastolic flow (thin arrows) in the right and
left middle cerebral arteries and low-velocity, systolic
spikes in the anterior cerebral artery (thick arrow), wave-
forms typical for brain death. (d) Noncontrast CT demon-
strates diffuse cerebral edema with loss of gray-white
matter differentiation. Nuclear scintigraphy demonstrates
59
lack of intracranial arterial flow with an abrupt cutoff of
activity at the skull base because increased intracranial
pressure exceeds cerebral perfusion pressure. Note also
the “hot nose” sign which is related to the decreased flow
within the internal carotid arteries compared to the
increased flow in the external carotid circulation (e)
60
the origin of the artery and the carotid siphon are
the two most common sites for occlusion. The
occlusion may be clinically silent due to a collat-
eral pathway that forms from the external carotid
artery to the ophthalmic artery which reverses its
flow to feed the intracranial ICA/MCA. Duplex
and transcranial Doppler imaging can demon-
strate this collateral pathway. Ultrasonogra-
phically, the occluded internal carotid artery will
be filled with echogenic material and will demon-
strate absence of flow by color Doppler flow
imaging and spectral Doppler. The common
carotid artery waveform, which has normally
low resistance with forward diastolic flow, will
frequently convert to a high-resistance waveform
due to the occlusion. If the collateral pathway
develops, the common carotid artery may main-
tain its normal low-resistance pattern. The ipsilat-
eral external carotid artery, which normally has a
high-resistance waveform (little forward diastolic
flow), will become “internalized” with its wave-
form changing from a high-resistance pattern to a
low-resistance pattern. The ipsilateral ophthalmic
artery can be visualized via the transorbital
approach and will demonstrate flow reversal
away from the transducer and globe (Fig. 13).
Other collateral pathways are possible including
feeding of the anterior circulation from the poste-
rior vertebrobasilar system via the posterior com-
municating arteries or filling of the ipsilateral
MCA via the contralateral MCA and anterior com-
municating artery.
Other Applications
TCDI has found a limited role in detection of
stenosis, aneurysms, arteriovenous malformations,
and intracardiac right-to- left shunts.
For stenosis of intracranial vessels, TCDI has
been found to be useful but not definitive for the
diagnosis. Confirmatory studies such as CTA or
conventional angiography are often necessary. An
elevation in flow velocity of 50–150 % (ratio of
1.5–2.5) at the site of stenosis as compared to the
flow velocity just proximal or distal to the stenosis
has been used by Roubec et al. as criteria for the
detection of stenoses less than 50 %. Criteria for
J.D. Kirsch
stenoses greater than 50 % were elevations in flow
velocities greater than 150 % at the site of stenosis
compared to proximal or distal flow velocities in
the vessel or flow velocity ratios >2.5. Detection
of turbulent blood flow, aliasing, visualization of
narrowing on color Doppler flow imaging, and
distal tardus-parvus waveforms have also aided
in detection of stenosis [25, 26].
Although not approved by the Food and Drug
Administration for clinical use in the United
States, ultrasound contrast-enhancing agents
such as Levovist (Schering AG) can help aid in
the detection of vessels and areas of stenoses
[17]. Intracranial vessels may be difficult to see
due to an insufficient temporal bone window, poor
insonation angle, or low flow velocities or vol-
umes. Imaging after intravenous injection of
Levovist, galactose-based microbubbles (<5 μm
in diameter) coupled with palmitic acid, has been
found to increase significantly the ability to detect
vessels and obtain spectral Doppler waveforms
compared to pre-contrast examinations. Detection
of collateral flow through the anterior and poste-
rior communicating arteries in patients with high-
grade ICA stenosis or occlusion was also signifi-
cantly improved with the use of an echo-contrast
agent [27, 28].
Transcranial power Doppler imaging has been
useful in the detection of intracranial aneurysm.
The detection of aneurysm by TCDI is dependent
on the size of the aneurysm, its location, and its
morphology. Detection rates of 47–73 % have
been reported in the literature [29, 30]. Aneurysms
in the anterior communicating and basilar arteries
had the highest detection rates (>50 %) as did
those greater than 10 mm in diameter. Aneurysms
in the posterior communicating and pericallosal
arteries had the lowest detection rate. MCA and
ICA aneurysms had detection rates of about 50 %.
Morphologically, spherical and multi-loculated
aneurysms were more easily detected than elon-
gated ones. The administration of IV echo-
contrast agent significantly increased the detec-
tion rate for aneurysms [31]. TCDI has also
shown promise in the long-term surveillance of
aneurysms treated by detachable coils. High sen-
sitivity and specificity was seen in the ability of
TCDI to detect and distinguish aneurysms that
4 Essentials of Transcranial Doppler Ultrasound 61

Fig. 13 (continued)
62 J.D. Kirsch

Fig. 13 (continued)
4 Essentials of Transcranial Doppler Ultrasound
Fig. 13 Internal carotid artery (ICA) occlusion: Gray-
scale ultrasound demonstrates echogenic material filling
the lumen of the right internal carotid artery (a, arrows).
Power and spectral Doppler demonstrates no flow within
the vessel consistent with occlusion (b, c). The common
carotid artery (CCA) has maintained its normal
low-resistance (d) pattern as a collateral pathway has
formed to the brain via the ipsilateral ophthalmic artery.
had been successfully treated demonstrating com-
plete occlusion from those with significant resid-
ual flow [32].
In patients with patent foramen ovale (PFO)
and right-to-left intracardiac shunting, TCDI has
shown high sensitivity and specificity for
establishing the diagnosis. Patent foramen ovale,
a remnant of the fetal circulation, is a flap-like
opening in the atrial septum primum and
secundum that persists into adulthood. It is esti-
mated to be present in 20–25 % of the general
population and can cause paradoxical embolic
events resulting in stroke, migraine or migraine-
like symptoms, systemic embolism, and neuro-
logic decompression sickness [33, 34]. The pres-
ence of a PFO and right-to-left shunt is usually
diagnosed by transesophageal echo (TEE)
[35]. TCDI offers a less-invasive option for diag-
nosing PFO by intravenously injecting 1 cm3 of
microbubbles and then evaluating for the presence
of bubbles in the MCA by TCDI via a
transtemporal window indicating a right-to-left
shunt is present. The technique is showing prom-
ise as being as accurate as TEE for the
diagnosis [36].
63
The ipsilateral external carotid artery normally has a high-
resistance waveform with little diastolic flow but has
become “internalized” and demonstrates increased forward
diastolic flow (arrow, e). The right ophthalmic demon-
strates flow reversal (f) away from the transducer (wave-
form below baseline), away from the globe, and toward the
brain. (g) Normal waveform and flow direction in the left
ophthalmic artery for comparison (OA ophthalmic artery)
Ultrasound Biosafety
Although generally considered a safe imaging
modality, diagnostic ultrasound does carry poten-
tial biosafety concerns. As such, diagnostic ultra-
sound imaging should be performed only by
personnel who have an understanding of the
potential biohazards related to ultrasound imaging
and who are fully trained in its safe and
proper use.
Ultrasound waves are mechanical compression
waves through a given medium. The waves prop-
agate in a longitudinal fashion with alternating
areas of compression and expansion. The term
“ultrasonic” refers to sound waves above the audi-
ble range and generally is considered to be sound
waves above a frequency of 20,000 Hz (0.02
MHz). Medical ultrasound utilizes frequencies of
sound generally in the range of 1–15 MHz (1–15
 106 Hz).
Ultrasonic sound waves and the resultant
imaging results in the deposition of energy into
the body that can potentially cause damage to the
tissues secondary to thermal and mechanical
64
(nonthermal) mechanisms. Attenuation of the
ultrasonic sound wave as it passes through the
soft tissues or bone results in loss of energy due
to absorption and scatter. Absorption of the sound
wave results in the wave’s energy being converted
to heat. Spectral Doppler has a greater heating
potential than other modes of ultrasound (gray-
scale and color Doppler flow imaging). Since
sound waves are mechanical energy being com-
posed of alternating areas of compression and
expansion, nonthermal injury can also result
from direct interaction of the sound wave with
the soft tissue. Rarefaction from the passing ultra-
sound wave can result in cavitation of the tissue.
The presence of contrast agents containing
microbubbles can increase the probability of cav-
itation and nonthermal effects [37, 38].
To limit the potential for biosafety effects, the
key principle in the safe use of ultrasound includes
the ALARA (as low as reasonably achievable)
principle. The principle, as applied to medical
diagnostic ultrasound exams, consists of using
the lowest output power and the shortest scan
time reasonably possible to obtain the necessary
diagnostic information to formulate an accurate
diagnosis. It is acknowledged that it may be nec-
essary to use longer scan times or higher outputs if
that is what is required to obtain accurate diagnos-
tic information [39].
As noted above, insonation of the soft tissue
with ultrasound can result in a rise in temperature
of the tissue due to energy deposition causing
potential thermal injury. The extent of damage
depends on the degree of temperature rise, the
duration of exposure to the elevated temperature,
the type of tissue exposed, its cellular proliferation
rate, and its potential for regeneration. Tempera-
tures of 39–43  C have been shown to cause
detrimental effects in vitro. Temperatures above
44  C can result in denaturation of proteins
[40]. Self-heating of the ultrasound transducers
(probes) can occur but is more an issue with
endo-cavitary transducers than for transducers
placed on the skin where heat dissipation is
greater. The presence of the bone (e.g., skull) in
the ultrasound beam increases the likelihood of
temperature rise due to direct absorption of the
beam by the bone and subsequent conduction of
J.D. Kirsch
heat from the bone to the soft tissues. Exposure to
the eye by ultrasound should also be kept to a
minimum due to low perfusion in the eye, espe-
cially the lens, which results in decreased capabil-
ity for heat dissipation [41].
Two indices have been devised, the thermal
index (TI) and the mechanical index (MI) which,
through an on-screen display, give an indication to
the sonographer of the potential for ultrasound-
induced bio-effects.
Three types of thermal indices have been
devised. The thermal index for the soft tissue
(TIS) is utilized when the ultrasound beam only
insonates the soft tissue; the TIB (thermal index
for the bone) is used when the ultrasound beam
insonates the bone at or near its focal zone; and the
TIC (thermal index for cranial bone) is used for
transcranial Doppler imaging when the probes are
adjacent to the bone (the skull) during scanning.
The TI is an indicator of the relative potential
for tissue temperature rise and biological effect. It
is defined as the ratio of the emitted acoustic
power to the power required to raise the tempera-
ture of the tissue by 1  C [41]. The TI is intended
to give an approximate guide to the likely maxi-
mum temperature rise that might be incurred after
a long exposure to ultrasound. A greater TI repre-
sents a higher heating potential and a higher risk
for tissue injury. The MI is a rough guide as to the
likelihood of cavitation occurring. In general, the
risk of cavitation increases with an increasing
MI. Various recommendations are present in the
literature for suggested thresholds for TI and MI
depending on the age of the patient, the body part
being scanned, and the duration of exposure to the
ultrasound beam [39–41].
Summary
Transcranial Doppler ultrasound is an extremely
useful imaging modality for evaluating the intra-
cerebral arterial system. It has found its greatest
utility in evaluating and monitoring patients in the
neurosurgical intensive care unit for vasospasm
following subarachnoid hemorrhage due to aneu-
rysmal rupture. The technology has also found
clinical utility in evaluating and screening
4 Essentials of Transcranial Doppler Ultrasound
pediatric patients with sickle cell disease who may
be at risk of stroke. In conjunction with clinical
findings, electroencephalography, and nuclear
medicine imaging, TCDI has also been shown to
be a useful adjunct in determining brain death.
Other possible uses for the imaging modality
include detection of stenosis, arteriovenous
malformations, determination of collateral path-
ways with occlusion of the extracranial internal
carotid artery, and right-to-left intracardiac shunts
due to patent foramen ovale.
Biosafety concerns are of paramount impor-
tance. Although ultrasound is considered a very
safe imaging modality, the practicing sonographer
should be aware and cognizant of the potential
injuries that could result to soft tissue and bone
from potential thermal and mechanical injury.
References
1. Kirsch JD, Mathur M, Johnson MH et al (2013)
Advances in transcranial Doppler US: imaging ahead.
Radiographics 33:E1–E14
2. Aaslid R, Markwalder TM, Nornes H (1982) Noninva-
sive transcranial Doppler ultrasound recording of flow
velocity in basal cerebral arteries. J Neurosurg
57(6):769–774
3. Lupetin AR, Davis DA, Beckman I et al (1995)
Transcranial Doppler sonography. Part 1. Principles,
technique, and normal appearances. Radiographics
15:179–191
4. Spencer MP, Reid JM (1979) Quantitation of carotid
stenosis with continuous-wave (C-W) Doppler ultra-
sound. Stroke 10(3):326–330
5. Marshall SA, Nyquist P, Ziai WC (2010) The role of
transcranial Doppler ultrasonography in the diagnosis
and management of vasospasm after aneurysmal sub-
arachnoid hemorrhage. Neurosurg Clin N Am
21(2):291–303
6. Keyrouz SG, Diringer MN (2007) Clinical review:
prevention and therapy of vasospasm in subarachnoid
hemorrhage. Crit Care 11(4):220
7. Condette-Auliac S, Bracard S, Anxionnat E et al (2001)
Vasospasm after subarachnoid hemorrhage: interest in
diffusion-weighted MR imaging. Stroke
32:1818–1824
8. Awasthi D (1997) Cerebral Vasospasm: current think-
ing and future trends. www.medschool.lsuhsc.edu/neu
rosurgery/nervecenter/spasm.html
9. Bederson JB, Connolly ES Jr, Batjer HH et al (2009)
Guidelines for the management of aneurysmal sub-
arachnoid hemorrhage: a statement for healthcare pro-
fessionals from a special writing group of the Stroke
65
Council, American Heart Association. Stroke
40(3):994–1025
10. Rasulo FA, De Peri E, Lavinio A (2008) Transcranial
Doppler ultrasound in intensive care. Eur J
Anaesthesiol Suppl 42:167–173
11. Kolias AG, Sen J, Belli A (2009) Pathogenesis of
cerebral vasospasm following aneurysmal subarach-
noid hemorrhage: putative mechanisms and novel
approaches. J Neurosci Res 87:1–11
12. Wozniak MA, Sloan MA, Rothman MI et al (1996)
Detection of vasospasm by transcranial
Doppler sonography: the challenges of the anterior
and posterior cerebral arteries. J Neuroimaging
6(2):87–93
13. Torbey MT, Till-Karsten H, Bhardwaj A et al (2001)
Effect of age on cerebral blood flow velocity and inci-
dence of vasospasm after aneurysmal subarachnoid
hemorrhage. Stroke 32:2005–2011
14. Lindegaard KF, Nornes H, Bakke SJ et al (1989) Cere-
bral vasospasm diagnosis by means of angiography
and blood velocity measurements. Acta Neurochir
100:12–24
15. White H, Venkatesh B (2006) Intensive applications of
transcranial Doppler in the ICU: a review. Intensive
Care Med 32(7):981–994
16. Sviri GE, Britz GW, Douville CM et al (2006)
Transcranial Doppler grading criteria for basilar artery
vasospasm. Neurosurgery 59(2):360–366
17. Maakaon JE, Taher AT, Besa EC (2014) Sickle cell
anemia. Medscape reference: emedicine.medscape.
com/article/205926-overview
18. Frempong KO (1991) Stroke in sickle cell disease:
demographic, clinical and therapeutic considerations.
Semin Hematol 28:213–219
19. Adams RJ, MCvie V, Nichols FT et al (1992) The use
of transcranial ultrasonography to predict stroke in
sickle cell disease. NEJM 326:605–610
20. Adams RJ, McKie VC, Carl EM et al (1997) Long
term stroke risk in children with sickle cell disease
screened with transcranial Doppler. Ann Neurol
42:699–704
21. Adams RJ, McKie VC, Hsu L et al (1998) Prevention
of a first stroke by transfusions in children with sickle
cell anemia and abnormal results on transcranial Dopp-
ler ultrasonography. N Engl J Med 339:5–11
22. Wijdicks EFM (2001) The diagnosis of brain death.
N Engl J Med 344(16):1215–1221
23. Ropper AH, Kehne SM, Wechsler L (1987)
Transcranial Doppler in brain death. Neurology
37:1733–1735
24. Petty GW, Mohr JP, Pedley TA et al (1990) The role of
transcranial Doppler in confirming brain death: sensi-
tivity, specificity, and suggestions for performance and
interpretation. Neurology 40:300–303
25. Roubec M, Kuliha M, Jonszta T et al (2011) Detection
of intracranial arterial stenosis using transcranial color-
coded duplex sonography, computed tomographic
angiography, and digital subtraction angiography.
J Ultrasound Med 30:1069–1075
66
26. Feldman E, Wilterdink JL, Kosinski A et al (2007) The
stroke outcomes and neuroimaging on intracranial
atherosclerosis trial. Neurology 68:2099–2116
27. Hansberg T, Wong KS, Droste DW et al (2002) Effects
of the ultrasound contrast-enhancing agent Levovist on
the detection of intracranial arteries and stenoses in
Chinese by transcranial Doppler ultrasound.
Cerebrovasc Dis 14(2):105–108
28. Droste DW, Jurgens R, Weber S et al (2000) Benefit of
echocontrast-enhanced transcranial color-coded
duplex ultrasound in the assessment of intracranial
collateral pathways. Stroke 31(4):920–923
29. Becker G, Greiner K, Kaune B et al (1991) Diagnosis
and monitoring of subarachnoid hemorrhage by
transcranial color-coded real-time sonography. Neuro-
surgery 28:814–820
30. Klotzsch Z, Nahser HC, Fischer B et al (1996) Visual-
ization of intracranial aneurysms by transcranial
duplex sonography. Neuroradiology 38:555–559
31. Turner CL, Kirkpatrick PJ (2000) Detection of intra-
cranial aneurysms with unenhanced and echo contrast
enhanced transcranial power Doppler. J Neurol
Neurosurg Psychiatry 68:489–495
32. Turner CL, Higgins JN, Kirkpatrick PJ (2003) Assess-
ment of transcranial color-coded duplex sonography
for the surveillance of intracranial aneurysms treated
with Guglielmi detachable coils. Neurosurgery
53(4):866–871
33. Hagen PT, Scholz DG, Edwards WD (1984) Incidence
and size of patent foramen ovale during the first
10 decades of life: an autopsy study of 965 normal
hearts. Mayo Clin Proc 59:17–20
J.D. Kirsch
34. Schwerzman M, Nedeltchev K et al (2005) Prevalence
in size of directly detected patent foramen ovale in
migraine with aura. Neurology 65:1415–1418
35. Schneider B, Zienkiewicz T, Jansen V et al (1996)
Diagnosis of patent foramen ovale by transesophageal
echocardiography and correlation with autopsy find-
ings. Am J Cardiol 77:1202–1209
36. Jauss M, Kaps M, Keberle M et al (1994) A compari-
son of transesophageal echocardiography and
transcranial Doppler sonography with contrast medium
for detection of patent foramen ovale. Stroke
25:1265–1267
37. Miller DL, Averkiou MA, Brayman AA et al (2008)
Bioeffects considerations for diagnostic ultrasound
contrast agents. J Ultrasound Med 27(special
issue):611–632
38. Bioeffects Committee of the American Institute of
Ultrasound in Medicine (2008) American Institute of
Ultrasound in Medicine consensus report on potential
bioeffects of diagnostic ultrasound. J Ultrasound Med
27:503–515
39. The British Medical Ultrasound Society (2009) Guide-
lines for the safe use of diagnostic medical ultrasound
equipment
40. O’Brien WD, Deng CX, Harris GR et al (2008) The
risk of exposure to diagnostic ultrasound in postnatal
subjects: thermal effects. J Ultrasound Med
27:517–535
41. Nelson TR, Fowlkes B, Abramowicz JS et al (2009)
Ultrasound biosafety considerations for the practicing
sonographer and sonologist. J Ultrasound Med
28:139–150
 DSA Imaging Protocol and Technique
5
Eytan Raz

Contents Abstract
Preprocedural Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . 68 Currently, digital subtraction cervico-cerebral
angiography (DSA) is seldom the first-line
Patient Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
study to evaluate the vasculature of the neck
Tray . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 or of the head and is usually performed after
Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 duplex ultrasound, CT angiography, or MR
angiography, in cases where additional details
Vascular Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
or assessment is needed. Many times, DSA is
Catheter and Wire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 obtained in the same setting just before the
Cervico-Cerebral Angiography . . . . . . . . . . . . . . . . . . . . . 70 interventional endovascular treatment.
Projections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Special Situations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Keywords
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Cerebral angiogram • DSA • catheter • wire •
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 vascular access

Obviously, this approach is justified when the

same information can be obtained with a nonin-
vasive approach, but it is the belief and the expe-
rience of the author that cervico-cerebral
angiography may add many details and informa-
tion compared to other techniques and has to be
kept in the armamentarium to use in all cases in
which vascular questions remains open after non-
invasive studies are performed. Multiple such
examples will be presented in the current chapter.
Another introductory point that needs to be
pointed out is that DSA is an operator-dependent
technique, and as such, a wide spectrum of quality
of exam and spectrum of complications incidence
E. Raz (*) will result: in good hands, a DSA can have as low
Neurointerventional Radiology Section, Department of
as 0.03 % incidence of stroke.
Radiology, NYU Langone Medical Center, New York,
NY, USA
e-mail: eytan.raz@gmail.com; eytan.raz@nyumc.org

# Springer Science+Business Media New York 2016 67

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_50
68
Preprocedural Assessment
Visiting the patient before the angiogram is man-
datory. This evaluation should focus on the clini-
cal history, previous surgical history, history of
medications, and allergies. Many important
changes in the procedure are usually tailored
based on all the findings of the preprocedural
assessment.
For elective angiograms, a basic thorough
evaluation, other than medical clearance, includes
blood work (CBC with differential, PT, PTT,
BMP, and, for female patients, hCG), urinalysis,
chest x-ray, and EKG. The reason to obtain these
studies is to evaluate the general medical risk, to
evaluate for renal insufficiency or renal failure
(worsened by use of iodine contrast), and to eval-
uate for ongoing infections and for bleeding diath-
esis. If the patient is pregnant – and the procedure
can be postponed – that is usually the best option.
The authors perform angiograms on
anticoagulated patients knowing that a micro-
puncture kit and a groin closure device will prob-
ably be used.
During the preprocedural assessment, the
neurointerventionalist will educate the patient
about the exam. Focusing on the sensations that
the patient will experience during the procedure
will ultimately improve the collaboration and,
consequently, improve the quality of the study
and decrease the rate of complications. The most
important points to discuss with the patient are:
the sensations in the groin when the local anes-
thetic is injected, the warming sensation when the
contrast is injected, and the possible dizziness and
appearance of flashes behind the eyes. Likewise,
also exemplifying the usual commands that are
given prior to the injection is needed. A written
informed consent needs to be signed by the patient
before the procedure.
Patient Preparation
In elective angiograms, the autors request the
patient to be NPO since the midnight before.
The only exception is usually a sip of water in
E. Raz
order to take the routine morning medications.
Once the patient arrives in the department, a
20–22 G IV is placed by the nursing team,
connected to an intravenous line that will serve
as a means to administer fluids and medications.
This peripheral IV line is also a lifeline, in the
event of an emergency. After the tray is set up, the
patient is brought to the angiography suite, appro-
priately positioned on the table, and connected to
continuous EKG, oxygen saturation monitoring,
and intermittent blood pressure monitoring. Pubic
hairs close to the groin are removed, after which
the groins can be prepped and the patient draped.
The patient’s head is held with a tape and aligned
carefully along the long axis of the detector.
Tray
The basic tray should have the following items:
• Femoral drape
• Skin prep tray
• Needle box
• Sterile labels
• Telfa
• 18 G 22 G–25 G needles
• Disposable scalpel no. 11
• Five 12 cc syringes
• Light handle cover
• 3-way and 1-way stopcocks
• Torque device
• Small J wire 0.038
• Terumo 0.038/150 cm
• Heparin 4,000 units/1 L bag
• Lidocaine 1 %
• 4 F or 5 F micropuncture kit
• 4 F or 5 F vertebral catheter
Procedure
Before starting the procedure, a time-out should
be conducted before the incision. The time-out
should involve the entire operative team, use
5 DSA Imaging Protocol and Technique
active communication, be briefly documented,
and use a checklist which should include: (1) cor-
rect patient identity using at least name and date of
birth, (2) allergies, pregnancy, or any other patient
specific issue, and (3) agreement on the procedure
to be done.
Vascular Access
A successful angiography begins with the groin
puncture. Before puncturing, localization of the
femoral head with x-ray is suggested, especially
for obese patients, in which the fat folds may
masquerade the anatomic landmarks. The right
common femoral artery, which is in the vast
majority of cases the aimed vessel for the access,
is in fact located in the medial upper quadrant of
the femoral head. Local anesthesia is performed
using 1 % lidocaine using two different needles:
first superficial injection of 2 cc using a 25 G
needle, wait 30 s and then injection of approxi-
mately 8 cc deep to the artery with a 22 G needle,
after careful evaluation of no blood return to avoid
injection in the bloodstream. Because of its pH, as
the lidocaine is being injected, it can be painful.
Prior to puncturing, a small skin incision (few
millimeters long) is made using a no. 11 scalpel
parallel to the skin fold and dissected using a
hemostat in order to facilitate the following pas-
sage of needle and catheter and to prevent the
subcutaneous forming of hematoma, creating a
tract with the surface. For discussion of the vari-
ous techniques to obtain access in the femoral
artery, please report to excellent discussions else-
where such as in Harrigan and Deveikis book.
Once the access is obtained, two main options
exist: pushing the catheter over the wire without a
sheath in place, or placing a sheath in order to
facilitate the possible catheter exchange. Both
pros and cons can be detailed for either technique.
The main advantage of not placing the sheath is
that a smaller hole is present in the artery.
The catheter is then advanced under fluoro-
scopic guidance in the descending aorta. Another
option to face at this point is if a double flushing or
69
continuous flushing technique is preferable; both
pros and cons can be found in other excellent
reviews. Double flushing consists of aspiration
of the contents of the catheter with one 10 ml
syringe followed by partial aspiration and irriga-
tion with a second syringe, while continuous
flushing technique consists in connecting the cath-
eter hub with a Y-connector plus three-way stop-
cock connected to a heparinized pressurized
saline bag. On average, a catheter should be
flushed every 90–120 s and after each wire place-
ment. The tip of the syringe should be oriented
inferiorly so that the air will rise upward toward
the plunger and away from the catheter. As a
syringe is connected to the catheter or stopcock,
it is best to allow a meniscus of blood to form by
allowing slight backflow into the hub of the cath-
eter: this meniscus is then brought together
against the syringe meniscus in a meniscus-to-
meniscus technique which reduces the likelihood
of air being trapped between the tip of the syringe
and the hub. After the syringe has been removed
from the end of the catheter, the hub should be
flushed and the best way to do it is using a
so-called Jabbie in angio suite jargon, which
essentially is a syringe with a dilator attached to
the tip. The Jabbie fits easily into the hub and
allows a jet of flush to be injected for cleaning
the hub.
Catheter and Wire
The kind of catheter chosen is a personal decision.
In the authors’ experience, they start each angio-
gram with a vertebral catheter leaving the option
to exchange for another shape in case of difficult
access. The vertebral catheter is a simple curve
catheter, meaning that it has only a primary curve
and does not need to be formed. The catheter is
advanced over the wire in the ascending aorta, and
the wire is then pulled out. In a standard angio-
gram, the first vessel to be catheterized is the right
common carotid artery. In order to select this
vessel, once it is in the ascending aorta, the cath-
eter needs to be pulled and rotated
70
counterclockwise. Once in the ostium of the ves-
sel, additional selective catheterization can be
obtained getting a new roadmap, reinserting the
wire and readvancing the catheter over the wire.
Every time the wire is removed and the catheter is
in position for injection, a flow check injecting
1–2 cc of contrast under fluoroscopy should be
performed in order to confirm the appropriate
catheter positioning and exclude iatrogenic dis-
section during the catheter advancement.
A bifurcation run in both frontal and lateral
projections may be obtained if necessary; other-
wise, a roadmap to access the ICA may be
enough. Advancing the catheter in the ICA with
attention to have the vertebral catheter curve par-
allel to the bloodstream and not against the vessel
wall is important. This is usually obtained by
rotating the catheter slightly counterclockwise in
the right ICA and clockwise in the left ICA.
If difficulties are encountered to access a cer-
vical vessels, other techniques can be considered,
such as the breath-holding technique or exchang-
ing the catheter for a double curved one such as
the Simmons – please consult other excellent
resources to good use of these catheters and how
to form and unform. Keep in mind that the more
complex the vasculature and the more catheters
are used, the higher the incidence of
complications.
In some cases, an aortogram is needed, for
example, in case an arch study was not obtained
with MRA or CTA, in case of possible vasculitis
or in patients with trauma to exclude an aortic tear.
If the aortogram is needed, a groin sheath has to be
placed. The catheter used for aortogram is a pigtail
catheter which is placed in the proximal ascending
aorta, proximal to the innominate artery.
The contrast is diluted in a 2:1 solution with
heparinized saline. This ratio allows excellent
opacification of the bloodstream with a relatively
limited contrast load. The injection of contrast
(iopromide, which is a nonionic agent with
300 mg of iodine/ml) is different to obtain
roadmap or to obtain angiographic images. For
roadmaps smooth continuous injection is
performed, while for angiography, fast, strong
injection with appropriate handling is needed.
E. Raz
Table 1 Flow rate for different vessels
Injector
settings
Vessel injected ml/s ml
Aortic arch 20 25
Common carotid artery (CCA) 8 12
Subclavian artery 6 15
Internal carotid artery (ICA) 6 8
External carotid artery (ECA) 3 6
Vertebral artery 6 8
3D angiogram in internal carotid artery 3 20
If the automatic injection is used, the following
are the volumes and rates used – keep in mind
that, in jargon, the flow rate followed by the total
volume is indicated in the angio suite (e.g., 5 by
7 means 7 cc of contrast to be injected with a rate
of 5 cc/s) (Table 1).
Another feature to keep in mind is the rise: a
“rate rise” in cerebral angiography refers to a
progressive acceleration of contrast over the first
second of the injection. The linear rate rise feature
is especially useful when a catheter is near the
origin of a vessel and a sudden injection of con-
trast may result in recoil of the catheter proximal
outside of the desired vessel. The PSI setting of
the injector refers to the maximum pressure
(pounds per square inch) that the injector will
generate while injecting the contrast.
Cervico-Cerebral Angiography
Biplane angiography is nowadays the standard of
care for cervico-cerebral angiograms and allows
for two planes to be acquired simultaneously with
a single contrast injection, limiting the total
amount of time and contrast needed to complete
the study.
Projections
Standard projections are illustrated in the figures;
the x-ray detector has a limited space, and as such,
5 DSA Imaging Protocol and Technique
the space needs to be optimized. Standard
arch projection is a left anterior oblique that
opens the view of the aortic arch. To better
visualize the origin of the vertebral artery, a
Towne’s view is necessary, given the origin of
the artery in the superoposterior aspect of the
subclavian artery. The Towne’s view is obtained
angulating the detector 30–40 cranially aligning
the petrous ridges below the superior rim of the
orbits.
Standard neck arteries views are the straight
frontal and lateral projections (Fig. 1). If a single
plane is used, an ipsilateral oblique frontal view is
preferable.
Before obtaining the cranial images, careful
evaluation of the alignment of the head along the
long axis of the detector should be checked. In the
standard frontal (PA) view of the internal carotid
injection, the petrous ridges are aligned with the
lower third of the orbits (Figs. 2a–e and 3a–f). In
the lateral projection, careful evaluation of the
alignment of the orbital roof and of the external
auditory canals is needed (Figs. 2f–j and 3g–j).
While for a CCA injection, the outer skull has to
be included on the field of view, for an ICA
injection the positioning of the detector has to be
adjusted just to cover the inner skull. For vertebral
injection intracranial views, both a frontal and
Towne’s views are needed, while for the lateral
projection, the C1 vertebra coverage needs to be
checked (Figs. 4 and 5). Special views are
obtained in particular situations: for example, to
better see the ICA cavernous segment, the com-
municating segment, or the middle cerebral artery
branches, a Haughton view may be obtained; in
order to position for this view, the x-ray tube
needs to be tilted caudally towards the shoulder
of the side of interest. To better visualize the ICA
bifurcation or the anterior communicating artery
complex, an ipsilateral transorbital oblique is
recommended. To memorize all the useful pro-
jections for the different vessels can be cumber-
some; the best way to obtain the best views is to
create a 3D virtual map of the cerebral vasculature
in thier mind and use this virtual map to guide the
best projection based on the orientation of the
vessel of interest.
71
Most cerebral angiographic runs are obtained
with a rate of two frames per second (FPS). A
variable frame rate can also be obtained limiting
the rate for venous phase. Usually, an imaging
sequence lasts approximately 10–12 s in order to
visualize the arterial, capillary, and venous
phases.
Modern angiography makes use of three-
dimensional images in various occasions. In
order to obtain the source 3D images, only a
single plane is needed, which is rotated around
the head of the patient before and during the
injection of contrast in a CT-like manner
(Fig. 6). The volumetric images thus obtained
can be postprocessed in different ways and
reconstructed in the preferred views (Fig. 7). An
obvious advantage of the 3D images is that they
can further guide the best 2D projections useful
that can be selected based on the 3D reconstruc-
tion (Fig. 8).
After the angiogram, careful evaluation of the
images obtained is done before pulling the cathe-
ter out of the body. This suggestion has two
purposes: first, to do a thorough evaluation in
case additional images are needed and, second,
carefully evaluate that no evidence of complica-
tions such as dissection or thromboembolism is
present. Once these things are ascertained,
the management of arteriotomy can start,
preferably accomplished with manual compres-
sion, with pressure applied approximately
1–2 cm above the skin incision for 15–20 min,
or longer if oozing is still noted after that time.
In certain situations, a closure device can be
used for a mechanical closure and multiple
devices for this purpose are available in the
market.
After the angiogram, they request that the
patient lies flat on bed for 6 h with the accessed
leg extended and with head of bed at <30 . Vital
signs need to be checked q1h. Nursing
team should check the groin puncture as
follows: q15min for 1 h, q30min for 2 h, and
then q1h for 3 h. The patient may restart the
usual diet after the angiogram with encourage
water intake. The IV can be discontinued before
discharge.
72 E. Raz

Fig. 1 Standard angiographic views for the CCA bifurcation: lateral, subtracted and unsubtracted (a, b) and straight
frontal subtracted and unsubtracted (c, d)
5 DSA Imaging Protocol and Technique 73

Fig. 2 (continued)
74
Fig. 2 Right ICA injection. In the standard frontal view of
the internal carotid injection, petrous ridges are aligned
with the lower third of the orbits. (a–e) Frontal projection
of right ICA in early arterial – subtracted (a) and
unsubtracted (b), – capillary (c), early venous (d), and
Special Situations
For evaluation of atherosclerotic disease, an
aortogram is suggested as well as routine neck
views of both common carotid and vertebral
arteries.
E. Raz
late venous (e) views. (f–j) Lateral projection in early
arterial – subtracted (f) and unsubtracted (g) –, capillary
(h), late capillary/early venous (i), and early venous (j)
views
For evaluation of intracranial aneurysms, a
complete four-vessel angiogram should be
performed: in 15 % of patients with SAH, multi-
ple aneurysms are present. If surgery is an option,
an external carotid injection is also performed in
order to have a map of the vessels for possible
5 DSA Imaging Protocol and Technique 75

Fig. 3 (continued)
76
Fig. 3 Left ICA injection. In the standard frontal view of
the internal carotid injection, petrous ridges are aligned
with the lower third of the orbits. (a–e) Frontal projection
of left ICA in early arterial – subtracted (a) and
unsubtracted (b) – early capillary (c), late capillary (d),
intracranial arterial bypass (superficial temporal
artery, internal maxillary artery, etc.). When an
aneurysm is identified, the mere diagnosis is not
enough: for thorough planning, the anatomy of the
aneurysm has to be defined, especially trying to
E. Raz
early venous (e), and late venous (f) views. In the lateral
projection (g–j) the alignment follows the orbital roof and
the external auditory canals: early arterial – subtracted (g)
and unsubtracted (h) – capillary (i), and late venous (j)
views
obtain the best view to show the neck aneurysm.
Sometimes, cross-compression may be necessary
to visualize the anterior communicating artery by
compressing the contralateral carotid and
allowing crossflow through the Acomm. In a
5 DSA Imaging Protocol and Technique 77

Fig. 4 (continued)
78
Fig. 4 Left vertebral artery injection. In the standard
frontal view of the internal carotid injection, petrous ridges
are aligned with the lower third of the orbits. (a–e) Townes
projection of left vertebral artery in early arterial –
subtracted (a) and unsubtracted (b) – late arterial (c),
E. Raz
capillary (d), and late venous (e) views. In the lateral
projection (f–k), the alignment follows the orbital roof
and the external auditory canals: early arterial – subtracted
(f) and unsubtracted (g) – late arterial (h), capillary (i),
early venous (j), and late venous (k) views
5 DSA Imaging Protocol and Technique 79

Fig. 5 Additional vertebral artery injection views. Additional injection with frontal projection (a, b) and double oblique
(c) views

similar fashion, the carotid artery can be com-
pressed during vertebral artery injection to better
visualize the Pcomm.
In cases of arteriovenous malformations, high-
rate runs are obtained (3–5 FPS), which help to
understand the subjacent anatomy. Angiograms of
other branches are often the most important runs
in cases of a high-flow AVM because partial
opacification is visualized with higher details
(Fig. 9). Another important point when imaging a
80 E. Raz

Fig. 6 Source images from 3D injection. These are the are similar to tomographic images obtained from a CTA,
source images acquired during injection (in this case, left and can be reconstructed in a similar fashion
ICA injection) that are used to create the 3D views. They
5 DSA Imaging Protocol and Technique
Fig. 7 (a–c) The volumetric images obtained can be postprocessed in different ways and reconstructed in the preferred
views using different color algorithms
patient with AVM is that, in order to see the normal
brain venous drainage, a longer imaging sequence is
necessary. In AVM ECA injection should be
obtained in order to ascertain ECA feeders.
For evaluation of dural arteriovenous fistulas,
the ECA, some of its branches, more commonly
the ascending pharyngeal and the occipital, but
81
also the ascending and deep cervical arteries may
need to be separately imaged according to the
location.
What to look for on an angiogram:
The neuroradiologist should be familiar with
the evaluation of angiographic images. Multiple
notions should be kept in mind when looking at an
82
Fig. 8 After analysis of 3D images, the best view to
evaluate the neck of the aneurysm is selected and the 2D
acquisition is then performed. Compare with Fig. 7a
Fig. 9 (a, b) In cases of arteriovenous malformations,
angiograms of other branches are often the most important
runs in cases of high-flow AVM because partial
E. Raz
angiogram. Differently from a CTA, the contrast
is injected inside a single vessel and hemody-
namic consequences may be noted such as the
inflow of nonopacified blood from other
vessels (Fig. 10). The angiography is a dynamic
technique and all the phases have to be
evaluated, early arterial, late arterial, capillary,
early venous, and late venous. In order to
better understand the anatomy, or in order to iden-
tify some branches (especially the ECA
branches), the unsubtracted images are very help-
ful (Fig. 11).
In the arterial phase, the vessel contour and size
are evaluated. The arterial anatomy is visualized
with identification of anatomical variants or
nonvisualization of expected vessels; careful eval-
uation of early draining veins is accomplished
during this phase. On the capillary phase, the
blush of physiologic or pathologic structures is
opacification is visualized with higher details. In this
example of a right frontal AVM, better details of the
nidus are obtained during vertebral artery injection
5 DSA Imaging Protocol and Technique 83

identified: in order to better see it, hard Patency of the veins is visualized. Abnormal per-
windowing may be helpful. On the venous sistent arterial opacification in this phase should
phase, dominance of different veins is identified, prompt to think of possible thromboembolism.
which can help to better tailor an intervention.

Conclusion

Even though noninvasive methods to depict the

vessels of the neck and head are nowadays avail-
able, catheter-based angiography is still the gold
standard to visualize the vasculature. This proce-
dure needs to be performed following strict
criteria in order to decrease as much as possible
the incidence rate and increase as much as possi-
ble the diagnostic yield.

Fig. 10 Different from a CTA, the contrast is injected

inside a single vessel and hemodynamic consequences
may be noted such as the inflow of nonopacified blood
from other vessels

Fig. 11 (a, b) External carotid artery injection. In order to better understand the anatomy, or in order to identify some
branches (especially the ECA branches), the unsubtracted images are very helpful
84
References
1. Fifi JT, Meyers PM, Lavine SD, Cox V, Silverberg L,
Mangla S, Pile-Spellman J (2009) Complications of
modern diagnostic cerebral angiography in an academic
medical center. J Vasc Interv Radiol 20(4):442–447
2. Thiex R, Norbash AM, Frerichs KU (2010) The safety
of dedicated-team catheter-based diagnostic cerebral
angiography in the era of advanced noninvasive imag-
ing. AJNR Am J Neuroradiol 31(2):230–234
3. Willinsky RA, Taylor SM, TerBrugge K, Farb RI,
Tomlinson G, Montanera W (2003) Neurologic compli-
cations of cerebral angiography: prospective analysis of
2,899 procedures and review of the literature. Radiology
227(2):522–528
E. Raz
4. Björkesten G, Halonen V (1965) Incidence of intracra-
nial vascular lesions in patients with subarachnoid
hemorrhage investigated by four-vessel angiography.
J Neurosurg 23(1):29–32
5. Lin JP, Kricheff II (1972) Angiographic investigation
of cerebral aneurysms. Technical aspects. Radiology
105(1):69–76
6. Harrigan MR, Deveikis JP (2012) Handbook of cerebro-
vascular disease and neurointerventional technique.
Springer Science & Business Media. New York, NY,
USA, 850 p
7. Osborn AG (1999) Diagnostic cerebral angiography.
Lippincott Williams & Wilkins, Philadelphia, 462 p
8. Morris P (2013) Practical neuroangiography. Lippincott
Williams & Wilkins, Philadelphia, 528 p
 Part II
Carotid and Vertebral Artery: Anatomy,
Diseases, and Treatment
 Anatomy of the Neck Arteries
6
Nasim Sheikh-Bahaei, Tomasz Matys, and Jonathan H. Gillard

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 The anatomy of neck arteries, normal varia-
tions, and anastomoses between different arter-
Aortic Arch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
ies is discussed in this chapter. All major
Common Carotid Artery . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 arteries of the neck originate from the aortic
Cervical Internal Carotid Artery . . . . . . . . . . . . . . . . . . . 89 arch via three main vessels: the
brachiocephalic trunk, left common carotid
External Carotid Artery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
(CCA), and left subclavian arteries. The CCA
External Carotid Artery Anastomoses . . . . . . . . . . . . . 92 courses superiorly in the neck, anteromedial to
Vertebral Artery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 the jugular vein and alongside the vagus nerve.
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 CCA typically divides at the level of C3 or C4
vertebral body into internal and external
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
carotid arteries. The cervical segment of the
internal carotid artery (ICA) usually does not
have branches in the neck, unless there is rem-
nant of carotid-basilar anastomoses such as the
hypoglossal artery or proatlantal artery type I.
The external carotid artery (ECA) is the
smaller branch of the CCA and runs
anteromedial to ICA. It has six branches in
the neck before entering the parotid gland and
divides into two terminal branches: superficial
temporal artery and internal maxillary artery.
The ECA supplies most of the neck structures,
scalp, and meninges. The vertebral artery is
usually the first branch of the subclavian artery,
running superiorly in the transverse foramen of
C6 to C1 with no cervical branches. It then
N. Sheikh-Bahaei (*) • T. Matys • J.H. Gillard courses between the C1 and foramen magnum
Department of Radiology, School of Clinical Medicine, and enters intracranial space. There are exten-
University of Cambridge, Cambridge Biomedical Campus,
sive anastomoses between carotid and verte-
Cambridge, UK
e-mail: n.sh.bahaei@doctors.org.uk; ns548@cam.ac.uk; bral arteries, the knowledge of which is
tomasz.matys@me.com; jhg21@cam.ac.uk

# Springer Science+Business Media New York 2016 87

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_2
88
essential to prevent disastrous complications
during interventions.
Keywords
Aortic arch • Common carotid artery • Internal
carotid artery • External carotid artery • Verte-
bral artery • External-internal carotid artery
anastomoses • Embryonic carotid-basilar sys-
tem anastomoses • Superior thyroid artery •
Ascending pharyngeal artery • Lingual artery
• Facial artery • Occipital artery • Posterior
auricular artery • Superficial temporal artery •
Internal maxillary artery
Introduction
In this chapter, the anatomy of the aortic arch and
neck arteries is described. The aortic arch config-
uration and its common variations are reviewed as
well as the anatomy of CCA, ECA, cervical ICA,
and vertebral artery. The normal variations and
anastomoses between different arteries, the
knowledge of which is crucial for performing
safe catheter angiography, are also discussed.
The aim of this chapter is to illustrate the applica-
tion of anatomical knowledge in clinical practice.
Aortic Arch
All major arteries of the neck originate directly or
indirectly from the aortic arch. In the standard
configuration, the aortic arch gives rise to three
great vessels: the brachiocephalic trunk (also
termed brachiocephalic or innominate artery),
the left common carotid artery, and the left sub-
clavian artery. The brachiocephalic trunk then
bifurcates at the level of the right sternoclavicular
joint into the right common carotid and the right
subclavian arteries (Fig. 1a). This standard
branching pattern occurs in approximately 70 %
of patients [1, 2].
There are many possible anatomical configu-
rations of the aortic arch and its branches.
Description of all these variations is beyond the
N. Sheikh-Bahaei et al.
scope of this chapter; in particular, developmental
anomalies leading to the formation of right or
double aortic arches will be omitted here. The
most common anatomical variants, knowledge
of which is of paramount importance for
performing safe catheter angiography, deserve a
brief mention however.
The most frequent anatomical variants are the
common origin of the brachiocephalic trunk and
the left common carotid artery seen on average in
13 % of population and the origin of the left
common carotid artery from the brachiocephalic
trunk encountered in 9 % of patients [3, 4]. Both
of these variants are more common in blacks
(25 % and 10 %, respectively). Instead of arising
separately, the left common carotid artery shares
the aortic arch origin with the brachiocephalic
trunk (Fig. 1b) or arises from the brachiocephalic
trunk slightly more cranially (1–2.5 cm distal to its
ostium) (Fig. 1c). As a result, only two major
vessels ascend from the aortic arch. The common
trunk variant, or sometimes both of the above
variants, has been referred to as “bovine” arch
implying high frequency of such configuration in
cattle. In fact, the aortic arch in cows more often
gives rise to a single common vessel, and these
branching patterns are not specific on any species.
Thus, “bovine-type” aortic arch should be
avoided as a misnomer and replaced with an ana-
tomical terminology [3, 4].
The arch origin of the left vertebral artery as a
fourth separate arch vessel is seen in 0.5 % of
patients; in such cases the left vertebral artery
most commonly arises between the left common
carotid and the left subclavian artery (Fig. 1d) and
sometimes distal to the left subclavian artery. Rare
variants include the bi-innominate artery (sym-
metrical right and left innominate arteries each
bifurcating into the common carotid and subcla-
vian vessels, Fig. 1e) and bicarotid trunk (bifur-
cating into the right and left common carotid
vessels, with separate origins of subclavian arter-
ies, Fig. 1f). Different combinations in which the
neck vessels arise separately from the aortic arch
are also encountered, in extreme case giving rise
to six aortic arch branches (separate right and left
6 Anatomy of the Neck Arteries
Fig. 1 Branching patterns of the aortic arch. (a) Standard
configuration. (b) Common origin of the brachiocephalic
trunk and the left common carotid artery. (c) Origin of
the left common carotid artery from the brachiocephalic
trunk. (d) Arch origin of the left vertebral artery.
subclavian, common carotid, and vertebral arter-
ies, Fig. 1g) [1, 2].
The most frequent anomaly of the aortic arch
seen in 0.4–2 % of population is the aberrant right
subclavian artery (Fig. 1h) arising distal to the left
subclavian artery and passing to the right side of
the mediastinum behind the esophagus causing its
typical posterior indentation and dysphagia
lusoria [2].
Common Carotid Artery
The right and left common carotid arteries (CCA)
course superiorly on both sides of the neck lying
within the respective carotid space, anteromedial
to the internal jugular veins and accompanied by
89
(e) Bi-innominate artery. (f) Bicarotid trunk. (g) Separate
origins of subclavian, common carotid, and vertebral arter-
ies. (h) Aberrant right subclavian artery. R/L-SA right/left
subclavian artery, R/L-VA right/left vertebral artery, R/L-
CCA right/left common carotid artery
the vagus nerve. The common carotid artery typ-
ically divides at the level of C3 or C4 vertebral
body into the external carotid artery (ECA) and
the internal carotid artery (ICA); this corresponds
approximately to the superior border of the thy-
roid cartilage. Bifurcation can occur as low as T2
and as high as C2. Normally CCA does not have
any named cervical branches, but occasionally the
vertebral artery or proximal ECA branches can
arise from it (Fig. 2) [1].
Cervical Internal Carotid Artery
The cervical segment of the ICA (C1) extends
from the carotid bifurcation to the skull base
where the artery enters the carotid canal
90
Fig. 2 (a) Volume rendering demonstrating the cervical
arteries. Common carotid artery (CCA) bifurcates into the
internal (ICA) and external (ECA) carotid artery at the
level of C4 vertebral body. Note the variant origin of the
left common carotid artery from the brachiocephalic trunk.
transitioning into the petrous (C2) segment. At the
bifurcation, the ICA most frequently arises
posterolaterally to ECA; its medial origin from
CCA is seen in 10–15 % of the population. The
most proximal part of the ICA shows a focal
dilatation termed the carotid bulb or carotid
sinus. Reversal of flow and turbulent flow pattern
in this region predispose it to atherosclerotic
plaque formation. Ascending cranially, the ICA
passes medial to the ECA and lies medial to the
internal jugular vein and anterior to the transverse
processes of the upper cervical vertebrae. Usually
there are no major named branches of the C1
segment of the ICA. Important exceptions include
remnants of the carotid-basilar system anastomo-
ses such as persistent hypoglossal artery (arising
from the cervical ICA between C1 and C3 and
passing through the hypoglossal canal to join the
N. Sheikh-Bahaei et al.
The vertebral artery (VA) enters the transverse canal at C6
level. Note a dominant left vertebral artery with hypoplas-
tic right vertebral artery. (b) Posterolateral view. Distal to
the CCA bifurcation, the ICA lies posterior and lateral to
the ECA
basilar artery) and persistent proatlantal artery
type I (proatlantal intersegmental artery, joining
the vertebrobasilar system after passing through
the foramen magnum) [1].
External Carotid Artery
The ECA is the second branch of the CCA, which
unlike the ICA has multiple branches in the neck
and supplies most of the neck structures. The ECA
extends from the carotid bifurcation, usually at the
level of the upper border of the thyroid cartilage, and
curves anteriorly before inclining backward behind
the neck of the mandible. The ECA is anteromedial
to the ICA [1]. It gives six main branches in the neck
and, after entering the parotid gland, divides into its
two terminal branches. In comparison to ICA, it has
6 Anatomy of the Neck Arteries
smaller caliber and rapidly tapers along its course as
it gives multiple branches.
The following structures cross the ECA anteri-
orly: the hypoglossal nerve (CN XII), posterior
belly of digastric muscle, stylohyoid muscle and
ligament, and facial nerve (CN VII) within the
parotid gland, while the pharyngeal wall, superior
laryngeal branch of the vagus nerve (CN X), and
deep lobe of the parotid gland are posterior to
ECA. The pharyngeal branch of the vagus nerve
(CN X) and glossopharyngeal nerve (CN IX) are
between the ECA and ICA [5].
Branches of ECA
1. The superior thyroid artery is the first branch
of ECA, which arises anteriorly and descends
to the apex of the thyroid. It supplies the larynx
and superior thyroid and anastomoses with the
inferior thyroid artery which is a branch of the
thyrocervical trunk. The superior thyroid artery
may arise from CCA bifurcation (20 %) or
have common origin with the lingual artery as
thyrolingual trunk (2.5 %) or with both lingual
and facial arteries as the thyrolinguofacial
trunk (2.5 %). Its main branches are the hyoid
artery, sternocleidomastoid branches, superior
laryngeal artery, and cricothyroid branch [1, 5].
2. The ascending pharyngeal artery is the
smallest branch of the ECA which arises pos-
teriorly near the carotid bifurcation. It is a long
vessel that ascends superiorly between the ICA
and ECA deeply in the neck [5]. After a short
common trunk, it divides into two major
branches: anteriorly, the pharyngeal trunk,
which is extracranial, and, posteriorly, the
neuromeningeal trunk, which is intracranial.
The other branches are inferior tympanic and
musculospinal arteries. The pharyngeal trunk
supplies the nasopharynx, oropharynx, and
eustachian tube. The neuromeningeal branches
supply the dura, CN IX–XII, and first, second,
and third cervical roots. Inferior tympanic and
musculospinal branches supply the middle ear
and prevertebral muscles, respectively [6]. The
ascending pharyngeal artery has several impor-
tant and potentially dangerous anastomoses
with internal carotid, internal maxillary, verte-
bral, and occipital arteries [6, 7].
91
3. The lingual artery is the second anterior branch
of the ECA. Initially it runs obliquely upward
to the hyoid bone. Then it curves downward
and forward, forming a loop, which is crossed
by the hypoglossal nerve [1]. It passes beneath
the digastric and stylohyoid muscles and
ascends and courses forward underneath the
tongue. It supplies the tongue, oral cavity, and
submandibular gland and has four branches:
suprahyoid, dorsal lingual, deep lingual, and
sublingual arteries. The lingual artery may
have common origin with the facial artery
(10–20 %) [5].
4. The facial artery originates above the lingual
artery. It is sheltered by the ramus of the man-
dible and courses above the submandibular
gland, then curves upward over the body of
the mandible, and passes anterosuperiorly
across the cheek to the angle of the mouth
and nasolabial fold. It then runs toward the
inner canthus of the eye and terminates as the
angular artery. The facial artery is a very tortu-
ous vessel and supplies the face, lip, palate, and
superficial neck [1, 5]. Its cervical branches are
the ascending palatine artery, tonsillar,
submental, and glandular branches. Branches
in the face include inferior and superior labial
arteries, inferior alar artery, and lateral nasal
branch [8]. The facial artery has anastomoses
with the ophthalmic artery, branch of the ICA,
and other ECA branches [7, 8].
5. The occipital artery originates from the poste-
rior aspect of the ECA opposite to the facial
artery, at the level of the posterior belly of the
digastric muscle [1, 5]. It ascends posterosu-
periorly between the occiput and C1 in the
occipital groove toward the superior nuchal
line and then becomes subgaleal and extends
up to the vertex. It supplies the scalp, upper
cervical musculature, and posterior fossa
meninges. Branches of the occipital artery are
sternocleidomastoid, auricular, mastoid,
descending, and occipital [9]. The occipital
artery is a remnant of the embryologic type I
and type II proatlantal arteries. It corresponds
to C1 and C2 segmental arteries and, therefore,
retains extensive connection to the vertebral
artery in adulthood [10]. It is critical to
92
understand the type and pattern of anastomoses
between occipital and vertebral arteries to
avoid any disastrous complications during
embolization.
6. The posterior auricular artery arises from the
posterior ECA above the occipital artery and
opposite to the apex of the styloid process
[1]. It ascends posteriorly beneath the parotid
and along the styloid process toward the ear. It
supplies the pinna, scalp, external auditory
canal, and chorda tympani. Common origin
of occipital and posterior auricular arteries is
seen in 12.5 % [5].
7. The superficial temporal artery is the smaller
of the two terminal branches of the ECA. It
arises in the parotid gland behind the neck of
the mandible and runs superiorly between the
deep and superficial lobes of the parotid gland
[1]. The transverse facial artery comes off the
superficial temporal artery before it leaves the
parotid gland. Then the superficial temporal
artery passes superficially over the zygomatic
process and divides into two branches: frontal
and parietal. It supplies the scalp and some
parts of the face [11].
8. The internal maxillary artery arises within the
substance of the parotid gland behind the man-
dibular neck. It then runs forward between the
ramus of the mandible and sphenomandibular
ligament and courses along the pterygoid mus-
cle to the pterygopalatine fossa [5]. The inter-
nal maxillary artery is divided into three major
parts based on its relationship to the external
(lateral) pterygoid muscle. The first (mandibu-
lar) section is posterior to the external ptery-
goid muscle and courses along its lower border
giving rise to five branches: the deep auricular
artery, anterior tympanic artery, middle menin-
geal artery, accessory meningeal artery, and
inferior alveolar artery. The second (pterygoid
or muscular) section is within the pterygoid
muscle and runs obliquely forward and upward
as well as medially through the infratemporal
fossa. Branches arising from this section are
entirely muscular including: anterior and pos-
terior deep temporal branches, pterygoid
branches, masseteric artery, and buccinator
artery. The third section (pterygopalatine) is
N. Sheikh-Bahaei et al.
anterior to the external pterygoid muscle and
enters the pterygopalatine fossa. It terminates
by sending off six branches to deep facial
structures: the posterior superior alveolar
artery, infraorbital artery, artery of pterygoid
canal, pharyngeal artery, greater palatine
artery, and sphenopalatine artery [12].
There are several potential anastomoses
between the internal maxillary artery and ICA: in
the orbital region, potential collateral route from
both proximal and distal internal maxillary arteries
to the ophthalmic branch of ICA. In the cavernous-
petrous region, the major anastomotic route is
between the middle and accessory meningeal arter-
ies and inferolateral trunk of the cavernous ICA [7].
External Carotid Artery Anastomoses
There are three main regions of anastomoses
between the ECA and ICA and vertebral arteries:
1. Anterior or orbital region: ECA anastomoses
through internal maxillary and facial arteries
with ophthalmic branch of ICA.
2. Middle or petrocavernous: ECA connects with
petrocavernous branches of the ICA via inter-
nal maxillary and ascending pharyngeal
branches.
3. Posterior or cervical: anastomoses between the
ECA and vertebral artery via occipital or
ascending pharyngeal arteries [7]. Persistent
embryonic connection between the ECA and
vertebrobasilar system is called type II
proatlantal artery [1, 10].
Vertebral Artery
The vertebral artery usually arises as the first
branch of the subclavian artery, proximal to the
thyrocervical trunk. Less frequent anatomical var-
iations include the vertebral artery originating
together with the thyrocervical trunk, from the
thyrocervical trunk, from the subclavian artery
distal to the thyrocervical trunk, or from the com-
mon carotid artery. In its typical course, the
6 Anatomy of the Neck Arteries
vertebral artery courses superiorly and medially to
enter the transverse foramen of C6; this part of the
vertebral artery is termed an extraosseous segment
(V1). The foraminal (V2) segment of the vertebral
artery ascends through the transverse foramina of
C6 to C1 vertebrae. At the level of the axis, the
artery makes a sharp superolateral turn to pass
through its obliquely positioned transverse fora-
men. In the extraspinal (V3) segment between the
C1 transverse foramen and foramen magnum, the
artery courses medially and posteriorly lying
superior to the arch of C1. In the last intracranial
segment (V4), the artery pierces the dura and
courses superomedially over the clivus to unite
with its contralateral counterpart into the basilar
artery (Fig. 3) [1].
Fig. 3 Volume rendering demonstrating the typical course
of the vertebral artery. The segment proximal to the trans-
verse foramen of C6 is termed V1; the segment coursing
through the transverse canal from C6 to C1 is termed V2;
note a sharp superolateral turn at the C2 transverse fora-
men. Extraspinal V3 segment stretches from the transverse
foramen of C1 to the site at which the artery pierces the
dura and enters the cranium
93
The level at which the vertebral artery enters
the transverse foramen is variable and dependent
on which embryonic intersegmental artery per-
sists to form the vertebral artery during develop-
ment. Vertebral arteries often demonstrate
variation in size with unilateral (usually left
sided) dominance [1]. Cervical branches of the
vertebral artery include segmental muscular and
spinal branches, often with a single branch to
brachial widening of the cervical cord (which
can also arise from the thyrocervical or
costocervical trunk or superior thyroid artery). In
the presence of carotid-basilar anastomoses (per-
sistent hypoglossal artery or proatlantal artery
type I from the ICA or proatlantal artery type II
from the ECA), the vertebral artery proximal to
anastomosis may be hypoplastic or absent.
Summary
All major neck arteries originate from the aortic
arch. In the standard configuration the aortic arch
gives rise to three great vessels: the
brachiocephalic trunk, the left common carotid
artery, and the left subclavian artery. The most
frequent anatomical variants are the common ori-
gin of the brachiocephalic trunk and the left com-
mon carotid artery (13 %). The common carotid
arteries course superiorly on both sides of the neck
and divide at the level of C3 or C4 vertebral body
into the ECA and ICA. The cervical segment of
ICA (C1) extends from the carotid bifurcation to
the skull base. At the bifurcation, the ICA most
frequently arises posterolaterally to the ECA and
then passes medially. Usually there are no major
named branches of the C1 segment of the ICA.
Important exceptions are persistent hypoglossal
artery and proatlantal artery type I as the remnants
of the carotid-basilar system anastomoses. The
ECA is the smaller branch of the CCA which has
six branches in the neck before entering the
parotid gland and divides into two terminal
branches: superficial temporal artery and internal
maxillary artery. The ECA supplies most of the
neck structures, scalp, and meninges. The verte-
bral artery is usually the first branch of the sub-
clavian artery, running superiorly in the transverse
94
foramen of C6 to C1 with no cervical branches. It
then courses between the C1 and foramen mag-
num and enters intracranial space.
There are three main regions of anastomoses
between the ECA with ICA and vertebral artery:
(1) anterior or orbital region, (2) middle or
petrocavernous, and (3) posterior or cervical. Per-
sistent embryonic connection between the ECA
and vertebrobasilar system is called type II
proatlantal artery. It is very crucial to identify
these anastomoses or embryological remnants to
prevent disastrous complications during
interventions.
References
1. Harnsberger HR, Macdonald AJ (2006) Diagnostic and
surgical imaging anatomy, 1st edn, Brain, head and
neck, spine. Amirsys, Salt Lake City
2. Lippert H, Pabst R (1985) Arterial variations in man:
classification and frequency. J F Bergmann, Munchen
3. Katz JC, Chakravarti S, Ko HH, Lytrivi ID,
Srivastava S, Lai WW et al (2006) Common origin of
the innominate and carotid arteries: prevalence,
nomenclature, and surgical implications. J Am Soc
Echocardiogr 19(12):1446–1448
4. Layton KF, Kallmes DF, Cloft HJ, Lindell EP, Cox VS
(2006) Bovine aortic arch variant in humans:
N. Sheikh-Bahaei et al.
clarification of a common misnomer. AJNR Am J
Neuroradiol 27(7):1541–1542
5. Standring S (2011) Gray’s anatomy, 39th edn. Living-
stone, Churchill
6. Hacein-Bey L, Daniels DL, Ulmer JL, Mark LP et al
(2002) The ascending pharyngeal artery: branches,
anastomoses, and clinical significance. Am J
Neuroradiol 23:1246–1256
7. Geibprasert S, Pongpech S, Armstrong D, Krings T
(2009) Dangerous extracranial–intracranial anastomo-
ses and supply to the cranial nerves: vessels the
neurointerventionalist needs to know. Am J
Neuroradiol 30:1459–1468
8. Lohn JWG, Penn JW, Norton J, Butler PEM (2011)
The course and variation of the facial artery and vein,
implications for facial transplantation and facial sur-
gery. Ann Plast Surg 67:184–188
9. Alvernia JE, Fraser K, Lanzino G (2006) The occipital
artery: a microanatomical study. Neurosurgery
58(1 Suppl):114–122
10. Chen PR, Siddiqui AH, Chen PR (2011) Intracranial
arterial collateralization: relevance in neuro-
endovascular procedures In: Peres JFP (ed)
Neuroimaging for clinicians- combining research and
practice. InTech Europe, Rijeka, p 175–202
11. Pinar YA, Govsa F (2006) Anatomy of superficial
temporal artery and its branches: its importance for
surgery. Surg Radiol Anat 28:248–253
12. Kashiwagi N, Nakanishi K, Kozuka T, Sato Y
et al (2010) Vascular supply with angio-CT for
superselective intra-arterial chemotherapy in advanced
maxillary sinus cancer. Brit J Radiol 83:171–178
 Imaging of Carotid Atherosclerosis
7
Jie Sun, Niranjan Balu, and Chun Yuan

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Growing interest in the in vivo characterization
of plaque pathology has furthered the develop-
Pathology and Implications for Imaging . . . . . . . . . 96
Pathogenesis of Carotid Atherosclerosis . . . . . . . . . . . . 97
ment of vessel wall imaging approaches
Histopathological Characteristics . . . . . . . . . . . . . . . . . . . . 97 beyond traditional techniques of luminal imag-
Plaque Morphology and Clinical Presentation . . . . . . 98 ing. By leveraging histopathological informa-
MR Imaging Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 tion from carotid endarterectomy specimens,
Field Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 magnetic resonance imaging has been proven
Coils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 capable of providing qualitative and quantita-
Multi-slice Black-Blood MRI . . . . . . . . . . . . . . . . . . . . . . . 100 tive information on a number of morphological
3D Black-Blood MRI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Multi-Contrast MRI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 and pathological features of carotid atheroscle-
Lipid-Rich Necrotic Core . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 rosis, including fibrous cap and necrotic core,
Intraplaque Hemorrhage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 intraplaque hemorrhage, plaque neovascu-
Calcification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 larization, and inflammation. These technical
Loose Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Neovasculature and Inflammation . . . . . . . . . . . . . . . . . . . 105 advancements present new opportunities to
Measurement of Plaque Morphology expand the understanding of the pathophysiol-
and Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 ogy of ischemic stroke and devise more effi-
Current Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 cient diagnostic and prognostic tools to address
Imaging-Based Patient Stratification . . . . . . . . . . . . . . . . 106 contemporary clinical problems in the manage-
Carotid Artery Stenting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 ment of carotid atherosclerosis. From a clinical
Cryptogenic Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
In Vivo Insights on Pathobiological Mechanisms . . . 109
perspective, this article introduces the various
Efficacy Markers for Testing Therapeutic magnetic resonance techniques for carotid ath-
Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 erosclerosis imaging. The background and
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 pathological basis of direct plaque imaging in
carotid arteries are discussed, followed by
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
available solutions and key technical consider-
ations, as well as promising applications that
have arisen from these techniques.

J. Sun • N. Balu • C. Yuan (*) Keywords

Department of Radiology, University of Washington,
Carotid artery • Atherosclerosis • Magnetic
Seattle, WA, USA
e-mail: sunjie@u.washington.edu; ninja@u.washington. resonance imaging • Plaque rupture • Black-
edu; cyuan@u.washington.edu

# Springer Science+Business Media New York 2016 95

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_34
96
blood • Multi-contrast • Ischemic stroke •
Carotid endarterectomy • Carotid artery
stenting • Cryptogenic stroke
Introduction
Atherosclerosis is a systemic condition that
affects large- and medium-sized arteries, such as
the aorta and its branches. Carotid atherosclerosis,
especially atherosclerotic plaques at the carotid
bifurcation, is one of the most prevalent manifes-
tations of systemic atherosclerosis. It has been
estimated from pooled population-based data
that the prevalence of moderate carotid stenosis
(>=50 %) is 4.8 % and 2.2 % for men and women
less than 70 years old, respectively. The numbers
increase further still in those aged more than
70 years to 12.5 % and 6.9 % [1].
Although the majority of carotid atheroscle-
rotic plaques have a benign clinical course, a
portion of them result in ischemic stroke, a leading
cause of death and disability worldwide. A pri-
mary goal of clinical imaging of carotid athero-
sclerosis is therefore to identify lesions that will
lead to ischemic stroke despite control of risk
factors. Atherosclerotic plaques typically become
flow-limiting as they progress. Unfortunately, the
simple assessment of luminal narrowing has
proven to be of limited value in predicting the
risk of ischemic stroke. Expanded knowledge on
the relationship between atherosclerotic plaque
Fig. 1 Atherosclerosis in the carotid bulb from three
endarterectomy specimens showing morphological var-
iations. (a) A mature necrotic core with intraplaque hem-
orrhage (asterisk) is the main component of the plaque.
The overlying fibrous cap is an inhomogeneous thickness.
J. Sun et al.
morphology and clinical complications has
shifted the focus of imaging of carotid atheroscle-
rosis from luminal narrowing to plaque instability,
i.e., the risk of acute plaque rupture and thrombo-
sis. Magnetic resonance imaging (MR, MRI) has
emerged as a promising imaging modality for
in vivo assessment of plaque instability. Over the
years, many validation studies have leveraged
carotid endarterectomy specimens to correlate
in vivo imaging findings with histopathological
features. Carotid MRI has been shown capable of
characterizing a number of pathological features
and biological activities of atherosclerotic plaques
that are implicated in plaque rupture, including
fibrous cap and necrotic core, intraplaque hemor-
rhage, plaque neovascularization, and inflamma-
tion [2–5]. As such, direct MR imaging of carotid
atherosclerosis presents a unique opportunity to
develop more effective diagnostic and prognostic
approaches for clinical research and practice.
Pathology and Implications
for Imaging
Atherosclerotic plaques are small but complex
pathological structures with considerable mor-
phological variations (Fig. 1). For imaging scien-
tists and radiologists to develop effective imaging
protocols and properly interpret image findings, a
basic understanding of the pathology of carotid
atherosclerosis is essential.
(b) The plaque is mainly fibrotic with small areas of lipid
accumulation (white arrows) and calcification. (c) Pieces
of dense calcification comprise the main portion of the
plaque (hollow arrow)
7 Imaging of Carotid Atherosclerosis
Pathogenesis of Carotid
Atherosclerosis
Atherogenesis is now appreciated as a multifacto-
rial process in which the arterial wall interacts
with a number of systemic and local factors.
The disturbed flow conditions at carotid bifurca-
tion make the carotid bulb a predilection site of
extracranial carotid artery for atherosclerosis.
Nonetheless, more extensive disease, detectable
by improved imaging approaches, is not uncom-
mon, particularly in the common carotid artery.
Individual variations in the geometry of carotid
bulb and the unknown extent of outward
remodeling make it difficult to discern the true
plaque burden by angiography or similar
techniques.
The accumulation of extracellular lipids and
cell debris, the migration and proliferation of
smooth muscle cells, and extracellular matrix pro-
duction are recognized processes in the formation
of plaques with well-developed necrotic core
(s) and an overlying fibrous cap, or fibroatheroma
as described by Virmani et al. in coronary athero-
sclerosis [6]. Structural changes are paralleled by
the recruitment and activation of immune cells
that respond to local insult and modulate the
inflammatory process. A variety of immune cells
including monocytes, lymphocytes, and, more
recently, granulocytes have been shown to play a
role in the pathogenesis and prognosis of athero-
sclerosis. Furthermore, it has been suggested
through dedicated experimental and human
autopsy studies that immune cells can be highly
adaptive in response to microenvironmental sig-
nals. In accord, functionally indistinct monocytes
may differentiate into polarized macrophage sub-
populations in the atherogenic milieu, taking on
heterogeneous biological functions that may sta-
bilize or destabilize the plaque [7]. Therefore, the
biological complexity of plaque inflammation has
placed a higher demand on molecular imaging by
emphasizing the activities rather than the mere
presence of immune cells. Another morphological
change that initiates early in the disease process is
plaque neovascularization, resulting from the pen-
etration of the medial layer (i.e., tunica media) by
hyperplastic vasa vasorum (i.e., the nutritional
97
microvasculature present in large- and medium-
sized arteries, with the physiological function of
supplying oxygen and nutrients to the outer por-
tion of the medial layer). Initially considered as a
compensatory mechanism to alleviate arterial wall
hypoxia as wall thickness increases, accumulating
evidence from recent studies suggests that plaque
neovascularization has an active role in
atherothrombosis by contributing to the develop-
ment, progression, and destabilization of athero-
sclerotic plaques. Leaky microvessels may
facilitate the entry of inflammatory cells as well
as plasma proteins. Furthermore, it is thought that
episodic ruptures of plaque microvessels are
likely the origin of intraplaque hemorrhage, a
common finding in advanced human lesions.
Improvements in research approaches, for both
histopathology and imaging, have recently
enabled a better appreciation of the prominent
role of intraplaque hemorrhage in atherogenesis
as well as the pathophysiology of clinical mani-
festations. However, mechanistic insights about
intraplaque hemorrhage are still limited at this
moment, as the phenomenon is rarely seen in
prevailing animal models.
Histopathological Characteristics
To date, much of the understanding of carotid
atherosclerotic plaque morphology is based on
histopathological examination of carotid endarter-
ectomy specimens, which do not cover the full
spectrum of lesion severity as compared to human
autopsy. The peripheral sections of advanced
carotid specimens are presumed to give a glimpse
of early or intermediate lesions, which frequently
show proteoglycan accumulation, lipid retention,
or early stage necrotic cores. Based on these
observations, it appears sensible to describe
carotid plaque morphology with nomenclature
proposed for classifying coronary atherosclerosis.
Yet caution should be exercised when making an
analogy for a specific feature with regards to its
role in pathogenesis and prognosis, given the dis-
tinct differences in vessel diameter and flow con-
ditions between coronary circulation and
extracranial carotid circulation.
98
Few studies have attempted to compare plaque
morphologies between coronary and carotid ath-
erosclerosis, yet notable differences seem to be
present. The fibrous cap overlying necrotic core(s)
is at risk of rupture in both coronary and carotid
lesions. The high mechanical forces in the carotid
circulation may render the fibrous cap of carotid
atherosclerosis more vulnerable to rupture, as it has
been reported that ruptured carotid plaques have
much thicker fibrous cap and fewer macrophage
infiltration than ruptured coronary plaques [8].
Thrombotic activity typically ensues, yet the high
flow velocity is likely a limiting factor for the
propagation of mural thrombus and the healing of
fibrous cap rupture. In line with the hypothesis,
ulceration is reported more frequently in carotid
atherosclerosis than in coronary atherosclerosis,
whereas total thrombotic occlusion of carotid ste-
nosis is not as common as in coronary atheroscle-
rosis. Fibrous cap erosion and calcified nodules are
the other two types of surface disruption reported in
the coronary artery, which may be accountable for
over a third of atherothrombotic events in coronary
atherosclerosis [6]. However, in the carotid artery,
plaque rupture is predominant, while plaque ero-
sion does not seem to be a major cause for
atherothrombosis [9]. Although calcified nodules
are more frequently seen in carotid atherosclerosis
in comparison to coronary atherosclerosis, its role
in atherothrombosis in the carotid artery has yet to
be determined.
Plaque Morphology and Clinical
Presentation
Morphological and pathological features of
carotid atherosclerosis have been shown to be
associated with downstream neurological ische-
mic events. Despite the longtime focus of clinical
imaging on luminal stenosis, chronic stenosis or
occlusion from carotid atherosclerosis rarely leads
to ischemic stroke and is therefore not considered
the pathological substrate of ischemic stroke. In a
study of 269 carotid plaques obtained en bloc
from surgical endarterectomy (96 ischemic stroke,
91 transient ischemic attack, 82 asymptomatic),
Spagnoli et al. observed a high prevalence of fresh
J. Sun et al.
mural thrombus constituted of platelets or fibrin
on plaque surface, which was associated with
either plaque rupture (90.1 %) or erosion (9.9 %)
in patients with stroke [9]. Furthermore, fresh
thrombus was more frequently observed when
carotid endarterectomy was performed closer to
symptom onset. With carotid plaques collected
from participants of the North American Symp-
tomatic Carotid Endarterectomy Trial (NASCET)
and the Asymptomatic Carotid Atherosclerosis
Study (ACAS) trials (71 symptomatic, 170 asymp-
tomatic), Fisher et al. noted that thrombus was
associated with plaque ulceration, both of which
were more prevalent in symptomatic patients than
in asymptomatic patients [10]. Another study by
Redgrave et al. examined 570 consecutive symp-
tomatic carotid plaques and demonstrated a con-
tinued post-event decline of plaque vulnerability
features including cap rupture and plaque inflam-
mation [11]. These studies have provided impor-
tant insights onto the pathogenesis of ischemic
events associated with carotid atherosclerosis. A
typical plaque responsible for a recent ischemic
stroke can be described as one with active throm-
bosis on the luminal surface, triggered by plaque
rupture and exposure of thrombogenic materials
to the flowing blood. Subsequently, acute plaque
rupture and thrombosis may lead to artery-to-
artery embolism or thrombotic occlusion, which
is in line with clinical observations.
Mechanisms of plaque rupture are not yet fully
understood. In the study by Redgrave et al. [11],
two morphological features demonstrated inde-
pendent association with cap rupture – intraplaque
hemorrhage and inflammatory cell infiltration of
fibrous cap. Other studies based on data from
coronary atherosclerosis have suggested that
fibrous cap thickness and necrotic core size are
good discriminators of plaques at risk for rupture
from those that are stable [12]. Of note, the degree
of stenosis could fit in here as one of the risk
factors for plaque rupture, considering the
increased mechanical forces associated with
local narrowing. Some of these risk factors may
have a more dynamic natural course, such as cap
inflammation. It is plausible that in vivo charac-
terization of major risk factors of cap rupture may
enable accurate estimation of the short-term or
7 Imaging of Carotid Atherosclerosis
long-term risk of neurological ischemic events in
patients with carotid atherosclerosis.
MR Imaging Techniques
In contrast to luminal imaging using bright blood
MR angiography, vessel wall visualization and
the identification of components of atherosclero-
sis require suppression of signal from intraluminal
blood. Thus, MRI sequences commonly referred
to as black-blood MRI are essential for vessel wall
imaging. Normal extracranial carotids have thin
pulsating walls and are often situated in deep
tissue planes in patients with a large neck. Carotid
bifurcation, the most frequent site of carotid ath-
erosclerosis, has a complex flow pattern with
recirculating flow. Thus, complete flow suppres-
sion in carotid bifurcation requires special consid-
eration of the flow conditions at carotid
bifurcation.
In addition to blood suppression, consider-
ations of spatial resolution, signal-to-noise ratio
(SNR), motion suppression, etc., are also impor-
tant to obtaining diagnostic quality vessel wall
images. Atherosclerotic plaque components are
small but complicated structures that require
high spatial resolution for identification and quan-
tification. Of note, the requirement of high spatial
resolution must be achieved with good SNR
within a clinically acceptable scan time. Another
reason to avoid long scan times is to reduce
motion artifacts. Due to the unique location of
carotid arteries, patient swallowing has been one
of the major causes for degraded image quality of
carotid MRI, whereas it will be less of a problem if
the scanning time is short.
Contrast-to-noise ratio (CNR) is equally
important in order to characterize vessel wall
pathology. Delineation of the lumen requires
good CNR with adequate blood suppression,
while delineation of the outer wall requires good
fat suppression. This allows plaque boundaries to
be clearly identified and disease burden to be
accurately estimated. Furthermore, MRI
sequences with special contrast mechanisms are
required for identification of various plaque com-
ponents such as calcification, lipid-rich necrotic
99
core, intraplaque hemorrhage, etc. Indeed, a
unique advantage of MR plaque imaging is to
employ more than one contrast weightings to
help differentiate various atherosclerotic
components.
Thus, carotid vessel wall imaging places spe-
cial technical constraints on field strength, coils,
and pulse sequences, which are required for accu-
rate identification of pathology. Carotid MRI pro-
tocols need to consider all these aspects while
being tailored for each study. This section outlines
the established carotid MRI sequences and emerg-
ing trends in carotid atherosclerosis imaging.
Field Strength
Initial development of carotid vessel wall MRI
took place on clinical 1.5-T scanners. Moving to
higher field strengths can increase the available
SNR. Initial studies comparing 1.5 and 3-T carotid
MRI found an improvement of wall SNR and
lumen/wall CNR by 1.5-fold for T1-WI imaging
and 1.7–1.8-fold for proton-density-weighted/T2-
WI imaging [13]. With optimization of sequences
for change in field strength, plaque components
can be identified with similar repeatability at 3-T.
However, quantification of some components,
such as calcification and intraplaque hemorrhage,
may differ due to higher susceptibility effects at
3-T. Additionally, the specific absorption rate
(SAR) is higher at 3-T and should be accounted
for in protocol optimization. Carotid MRI at
higher field strengths is challenging due to inho-
mogeneous excitation, suboptimal fat suppres-
sion, and higher susceptibility. Despite these
disadvantages, carotid imaging is moving into
7-T due to the advantages of higher SNR.
Coils
At 1.5 and 3-T, body coil transmission is used for
homogenous excitation, which is also useful for
many blood suppression sequences that use
nonselective excitation. However, body coil
reception does not provide a high enough SNR
for carotid vessel wall MRI within clinically
100
acceptable scan times. Thus, phased array surface
coils with a small diameter and applied close to
the neck provide improved SNR and are often
used for vessel wall MRI. Phased array coils
with four, six, and eight coil elements have been
designed for carotid imaging. Compared to a four-
element design, an eight-element design was
shown to provide 1.7-fold higher SNR and larger
coverage along the axis of the artery [14]. Some of
these coil designs are available commercially on
all major scanner vendor platforms. MRI at 7-T
also requires custom coil design, but these are
currently available for research purposes only.
(c) Double inversion-recovery (DIR) and variants:
Multi-slice Black-Blood MRI
Traditionally, the carotid vessel wall is scanned
using 2 mm thick slices with higher in-plane res-
olution of 0.5–0.6 mm. A stack of 12–16 slices
centered on the carotid bifurcation covering a
3 cm region is used in a multi-slice black-blood
MRI protocol. Blood suppression can be achieved
by black-blood preparation of the magnetization
prior to gradient or spin-echo acquisition. The
sequences used for black-blood imaging can be
briefly described as below:
(a) Spin-echo acquisition: Direct spin-echo MRI
without any magnetization preparation can
provide black-blood contrast with proper
choice of echo time (TE). Only spins
experiencing both the excitation and
refocusing spins will produce signals. If after
experiencing the 90 pulse blood flows
through the slice during TE and does not
experience the refocusing pulse, it will pro-
duce a black-blood effect in the acquired echo.
Long TE spin-echo or fast spin-echo
sequences such as T2-WI images exhibit this
effect. However, slow flow and recirculating
flow are not suppressed by this method.
Increasing slice thickness will also lead to
more flow artifacts.
(b) Inflow saturation: In this method, the signal
from a thick slab proximal to the carotid bifur-
cation is completely dephased every repeti-
tion time (TR). The dephased blood then
J. Sun et al.
flows through the imaging slice and produces
a black-blood effect. This method can be used
for short TE sequences such as T1-WI con-
trast but is prone to flow artifacts depending
on the thickness of the saturation slab and its
distance from the imaging slice. A traveling
saturation band is used to ensure that the
distance to all slices remains the same and a
short delay time is used so that blood signal
does not recover before being imaged.
Advantages of inflow saturation include low
SAR, wide availability, and selectivity in
suppressing arteries and/or veins.
DIR preparation consists of two inversion
pulses followed by a delay before image acqui-
sition. A nonselective inversion pulse is
followed by a slice-selective re-inversion of
the imaging slice. Blood with inverted magne-
tization proximal and distal to the imaging slice
flows into the imaging slice while simulta-
neously undergoing T1 relaxation. The inver-
sion delay time (TI) is calculated such that
blood magnetization is zero during image
acquisition, thereby causing a black-blood
effect. Both arterial and venous blood are there-
fore nulled by DIR. The long TI time allows
sufficient time for blood inflow and ensures
good blood suppression. However, artifacts
may still occur due to slow or recirculating
flow since the non-inverted blood within the
imaging slice may not have time for outflow
from the imaging slice. Increasing imaging
slice or slab thickness tends to worsen flow
artifacts with DIR. DIR has better performance
than inflow suppression and is widely available
on all vendor scanner platforms. Accurate
lumen identification in vessel wall imaging
requires at least a DIR preparation. The pri-
mary disadvantage of DIR for carotid imaging
is that it is a single slice method. Multi-slice
DIR (MDIR) preparations have been proposed
to reduce scan time associated with DIR
[15]. The current most efficient method uses a
slab-selective re-inversion rather than a slice-
selective inversion, thus allowing up to eight
slices to be acquired with good flow suppres-
sion within a single TR [16].
7 Imaging of Carotid Atherosclerosis 101

(d) Motion sensitization: All the above methods 3D Black-Blood MRI

perform best with laminar blood flow with
artifacts arising in cases of slow flow, turbu-
lent flow, and recirculating flow. Flow in the
normal carotid bifurcation can be
recirculating or turbulent, causing plaque-
mimicking artifacts on black-blood MRI.
Use of motion-sensitized-driven equilibrium
preparation (MSDE) can effectively suppress
artifacts from recirculating or turbulent flow
[17]. MSDE with two refocusing pulses is less
affected by B1 inhomogeneity and eddy cur-
rent effects than MSDE with one refocusing
pulse [18]. MSDE is not dependent on inflow
of blood into the imaging slice. Hence, it can
be used for non-axial slice acquisition. Addi-
tionally, MSDE is also compatible with acqui-
sition of thick slabs and is therefore used for
three-dimensional (3D) black-blood MRI. An
extension of motion sensitization uses a train
of delay alternating with nutation for tailored
excitation (DANTE) subpulses with inter-
spersed gradient dephasers for suppression
of blood flow [19]. DANTE with high ampli-
tude dephasing gradients can also be used for
3D MRI.
MRI studies of atherosclerotic plaque have gen-
erally used 2–3 mm thick axial slices, since black-
blood techniques such as inflow suppression or
DIR rely on inflow or outflow for efficient blood
suppression. Compared to the traditional
two-dimensional (2D) sequences, 3D black-
blood MRI can improve the accuracy and the
repeatability of plaque measurements due to
higher spatial resolution in the z direction and
less partial volume averaging. However, this
requires blood suppression in a relatively thick
slab, and therefore use of DIR can result in
increased flow artifacts. Since MSDE flow sup-
pression does not depend on flow, 3D black-blood
MRI can be achieved with high isotropic spatial
resolution with this technique. A 3D MSDE pre-
pared rapid gradient echo (3D-MERGE) vessel
wall imaging technique can provide
0.7  0.7  0.7 mm3 isotropic resolution within
a 2-min scan time with good blood suppression
(Fig. 2) [20]. Since 3D-MERGE datasets are iso-
tropic, interactive reformatting of multiple image
planes can be used for plaque visualization. Spin-
echo-based sequences with variable flip angle

Fig. 2 3D isotropic high-

resolution
(0.7  0.7  0.7 mm3)
carotid artery images
acquired with
3D-MERGE. Two major
advantages offered by 3D
isotropic black-blood vessel
wall imaging are shortened
scan times for clinical utility
and flexible reformation
capability for improved
plaque visualization. Note
that vessel wall boundaries
are clearly visible on all
reformats (Reprinted with
permission from Balu
et al. [20])
102 J. Sun et al.

Table 1 Multi-contrast carotid MRI protocol for 3-T

Pre/post- Proton-
contrast Time-of- density 3D-MERGE
T1-WI flight T2-WI -weighted SNAP sequence sequence
Sequence TSE TFE TSE TSE PSIR TFE
Image mode 2D 3D 2D 2D 3D 3D
Scan plane Axial Axial Axial Axial Coronal Coronal
TR (ms) 800 20 4,800 4,800 13 10
TE (ms) 10 4.6 50 10 4 5
Flip angle ( ) 90 50 90 90 11 6
FOV (cm) 14  14 14  14 14  14 14  14 14  14  3.2
Resolution 0.55  0.55 0.55  0.55 0.55  0.55 0.55  0.55 0.7  0.7  0.7 0.7  0.7  0.7
(mm)
Slice 2 1 2 2 Not applicable Not applicable
thickness (interpolated)
(mm)
Blood QIR/MSDE Venous DIR/MSDE DIR/MSDE Not applicable MSDE
suppression inflow
suppression
Fat Yes No Yes Yes Yes Yes
suppression
3D-MERGE 3-dimensional MSDE prepared rapid gradient echo, DIR double inversion-recovery, FOV field of view,
MSDE motion-sensitized-driven equilibrium preparation, PSIR phase-sensitive inversion-recovery, QIR quadruple
inversion-recovery, SNAP simultaneous non-contrast angiography and intraplaque hemorrhage, TFE turbo field echo,
TSE turbo spin echo

long echo trains have also been suggested for
imaging plaque burden [21]. Further studies on
the use of 3D sequences for identification of
plaque components are required before 3D MRI
can be used in a similar way to current 2D MRI
protocols.
Multi-Contrast MRI
Atherosclerotic plaques may range from simple
wall thickening to complicated plaques with mul-
tiple plaque components. By varying pulse
sequence parameters, image contrast may be
manipulated such that specific plaque components
are visible in each contrast weighting. For exam-
ple, a highly T1-WI fat-suppressed sequence will
show intraplaque hemorrhage with high signal
intensity compared to fibrous tissue or muscle.
Thus, a multi-contrast MRI protocol where multi-
ple contrast weightings are used in a
complementary fashion is essential for compre-
hensive evaluation of atherosclerotic plaques.
Table 1 shows a typical carotid MRI protocol
with traditional 2D multi-contrast MRI in com-
parison to newer isotropic 3D MRI sequences.
Single contrast weightings may be used when
only a subset of plaque components is of interest.
Sequences of interest for detecting important
plaque components are briefly described. Table 2
shows the appearance of plaque components on
each of the different contrast weightings.
Lipid-Rich Necrotic Core
Lipid accumulation in plaques evolves from lipid
pools to necrotic cores containing cell debris with
progression of atherosclerosis. Fat suppression
used for plaque imaging suppresses signal from
triglycerides in subcutaneous fat but does not
suppress signal from necrotic core, where the
7 Imaging of Carotid Atherosclerosis
Table 2 Plaque component classification using multi-contrast MRI
Time-of-flight T1-WI
Lipid-rich necrotic core o o/+
Intraplaque hemorrhage + +
Calcification
Loose extracellular matrix o /o
Dense fibrous tissue o o
Signal intensities are relative to adjacent muscle
+ hyperintense, o isointense, hypointense
signal mainly originates from free cholesterols
and cholesteryl esters. Necrotic core is isointense
on T1-WI sequences, but can be seen on T2-WI
sequences as a hypointense region due to its short
T2 components. Thus, a non-contrast MRI proto-
col requires a T2-WI sequence for detection of
necrotic core.
Sensitivity and specificity for detection of
necrotic core can be improved by administration
of gadolinium contrast agent such as gadopentetic
acid (Gd-DTPA) [2]. Due to necrotic contents,
gadolinium uptake into necrotic core is negligible
compared to surrounding fibrous tissue. Post-
contrast difference between necrotic core and
fibrous tissue is maximal 5–10 min after contrast
injection. Therefore, necrotic core can be readily
identified by comparing pre-contrast and post-
contrast black-blood T1-WI images. However,
adequate blood-suppression post-contrast is chal-
lenging. The T1 of blood decreases sharply after
contrast administration, but the change in blood
relaxivity is not predictable since it is dependent
on distribution volume – an unknown patient-
dependent factor. Thus, while black-blood tech-
niques such as DIR provide good blood-
suppression pre-contrast, adjustment of TI of
DIR post-contrast is empirical and does not pro-
vide adequate blood suppression in all subjects. A
quadruple inversion-recovery (QIR) preparation
has been introduced for this purpose [22]. QIR
consists of two DIR modules in tandem and con-
sists of two TIs that are adjusted such that blood
suppression is achieved for the set repetition time
(TR) of the sequence. QIR blood suppression is
T1 insensitive, thus enabling the same sequence
parameters to be used for both pre- and post-
contrast T1-WI MRI.
103
Proton-density-weighted T2-WI Post-contrast T1-WI
o/+ /o
/o /o /+
+ + +
o +
Intraplaque Hemorrhage
Bleeding into the plaque may come from either
the lumen secondary to surface rupture or from the
vasa vasorum originating from the adventitia.
Hemoglobin in the red blood cells then passes
through several stages of degradation. One of the
intermediate stages is the formation of highly
paramagnetic methemoglobin. Methemoglobin is
visible as hyperintense regions on T1-WI images
such as T1-WI fast spin echo or time-of-flight
images. However, the sensitivity and specificity
for detection of intraplaque hemorrhage can be
improved by use of highly T1-WI sequences [23].
Magnetization-prepared rapid gradient echo
(MPRAGE) is often used for detection of
intraplaque hemorrhage [3]. The MPRAGE
sequence employs an inversion recovery preced-
ing a gradient echo acquisition. Fat suppression is
also employed to reduce signals from subcutane-
ous fat. MPRAGE has been shown to have higher
sensitivity and specificity for intraplaque hemor-
rhage detection than T1-WI fast spin echo or time-
of-flight. However, blood suppression in
MPRAGE can be challenging since TI has to be
optimized simultaneously for both blood suppres-
sion and visualization of intraplaque hemorrhage.
A better approach combines MPRAGE acquisi-
tion with phase-sensitive reconstruction (PSR).
While PSR requires the additional acquisition of
a reference image, it provides additional benefits
over blood suppression. PSR is used in the simul-
taneous non-contrast angiography and intraplaque
hemorrhage (SNAP) MRI sequence where the
nonselective inversion pulse of MPRAGE is
replaced with a slab-selective inversion [24]. The
TI of SNAP is adjusted such that blood flowing
104
Fig. 3 SNAP imaging and histopathology of a compli-
cated plaque with intraplaque hemorrhage and ulcera-
tion. Two image portions are simultaneously acquired. The
portion of non-contrast MR angiography (a) shows a high-
grade stenosis in the proximal external carotid artery
(arrowhead) and ulcerated plaque in the proximal internal
carotid artery (ICA) (arrow). The intraplaque hemorrhage
portion (b) can be fused with the angiogram portion (c).
into the imaging plane is inverted only a single
time compared to the blood in MPRAGE that may
be inverted multiple times. During image acquisi-
tion at TI, flowing blood which has a long T1 has
negative longitudinal magnetization, while short T1
components such as intraplaque hemorrhage have a
positive longitudinal magnetization. PSR of SNAP
takes advantage of this difference in sign of longi-
tudinal magnetization to provide two images from
the same acquisition: negative part provides an
angiogram, while the positive part provides signal
for intraplaque hemorrhage. Figure 3 shows an
ulcer on the SNAP angiogram with corresponding
intraplaque hemorrhage in the plaque. Since both
MR angiography and intraplaque hemorrhage
imaging are obtained from the same SNAP acqui-
sition, they are naturally registered together.
Calcification
Calcification is hypointense on all contrast
weightings due to its very short T2 relaxation
time. This forms the basis for identification of
J. Sun et al.
Axial reformats of the bifurcation are shown in (d) with an
ulcer (arrows) shown as a dark depression and intraplaque
hemorrhage as hyperintense signals. Imaging findings are
confirmed on matched histological cross sections
(Masson’s trichrome). Note that the good blood suppres-
sion in (d) allows clear distinction of intraplaque hemor-
rhage from lumen (Reprinted with permission from Wang
et al. [24])
calcification in multi-contrast plaque MRI. How-
ever, two issues need to be noted. First, it is
common to witness calcifications that are either
close to luminal surface or even in direct contact
with flowing blood in carotid atherosclerosis. In
such situations, it could be challenging to separate
calcification from lumen based on black-blood
images alone. As such, accurate identification of
calcification and therefore overall plaque size
depend on the combined use of bright-blood and
black-blood sequences. Most multi-contrast pro-
tocols used in previous clinical studies have
involved at least 3D time-of-flight imaging to
overcome the problem. Second, the quantification
of calcification should be mainly performed on
spin echo sequences since their size tends to be
overestimated on gradient echo sequences due to
increased susceptibility of calcification. Use of
very short TE sequences called ultrashort echo
time (UTE) sequences renders calcification bright
and is less affected by susceptibility. UTE
sequences require special hardware and spiral or
radial readouts in addition to dual echo acquisition
for the identification of calcifications.
7 Imaging of Carotid Atherosclerosis 105

Loose Matrix as gadolinium contrast is injected. Additionally, a

Loose matrix refers to the category of plaque
tissue rich in proteoglycans, which can be encoun-
tered frequently in carotid atherosclerosis of var-
ious stages. Loose matrix tissue is rich in water
and shows up as hyperintense signals on T2-WI
images. In addition, loose matrix may be perme-
able to gadolinium and show up as areas of high
contrast uptake on post-contrast black-blood
MRI. However, their clinical implications when
encountered in advanced lesions are uncertain.
Neovasculature and Inflammation
Unlike the previous components, neovasculature
and inflammation detection require specialized
modeling of gadolinium contrast uptake for detec-
tion. In contrast to post-contrast T1-WI black-
blood MRI used for identification of lipid core,
neovasculature and inflammation detection
requires fast and continuous acquisition of images
bright blood technique is preferred to measure the
arterial input function. Therefore, a dynamic
contrast-enhancement (DCE) MRI sequence is
started prior to gadolinium injection. In practice,
this is a T1-WI bright-blood gradient echo
sequence with inflow saturation of both arterial
and venous flow. After two dynamics are
acquired, half dose of gadolinium contrast bolus
is injected at a rate of 1 cc/s. Since the T1 relax-
ation time of blood is drastically reduced, a bright
blood effect is created post-contrast. About
20 dynamic scans with scan times of 15 s per
dynamic scan are collected. Kinetic modeling
using the Patlak model provides two quantitative
measurements: fractional plasma volume (vp) and
permeability rate constant (Ktrans). Both vp and
Ktrans have been found to be correlated with his-
tologically measured neovasculature and macro-
phages in a validation study using carotid
endarterectomy specimens [4]. Figure 4 shows a
pseudo-color image with vp in red and Ktrans in
green.

Fig. 4 Vasa vasorum

imaging for evaluating
plaque revascularization
and inflammation. Cross-
sectional images of right
carotid bifurcation on
pre-contrast T1-WI QIR (a),
post-contrast T1-WI QIR
(b), and corresponding vasa
vasorum image (c) obtained
from contrast kinetic
modeling of DCE-MRI.
Red areas correspond to
partial volume of blood (vp)
and green areas correspond
to transfer constant (Ktrans).
Asterisk indicates lumen;
circle, jugular vein; arrow
head, lipid-rich necrotic
core (Reprinted with
permission from Balu
et al. [51])
106
Ultrasmall super paramagnetic particles of iron
oxides (USPIO) are a class of iron oxide magnetic
particles that are also used for imaging plaque
inflammation [25]. After systemic injection,
USPIOs are concentrated in macrophages and
can be detected by their strong magnetic
susceptibility.
Long TE gradient echo sequences can be used
to detect the T2 or T2* effect of USPIOs. Similar
to gadolinium contrast injection, areas of plaque
with macrophages can be identified by comparing
pre and post USPIO injection. In contrast to gad-
olinium, areas of macrophage concentration are
identified by loss of signal, and the signal differ-
ence is maximized at 36 h as compared to
5–10 min for gadolinium. Use of positive suscep-
tibility contrast techniques may improve detection
of USPIOs [26].
Measurement of Plaque Morphology
and Composition
Several important measurements can be derived
from a multi-contrast plaque MRI protocol. Out-
lines of the lumen and outer wall on black-blood
MRI provide a measure of the plaque burden. The
wall area defined as the difference between the
two outlines provides a first measurement of
plaque in the vessel wall. Plaque area normalized
for vessel size by the outer wall area provides the
normalized wall index (NWI), which is similar to
percent atheroma volume used in intravascular
ultrasound imaging of coronary atherosclerosis.
While plaque burden can be measured on a single
contrast weighting such as pre-contrast black-
blood T1-WI sequence, outlining plaque bound-
aries on multi-contrast MRI is more robust for
advanced lesions in view of superficial calcifica-
tion and plaque-mimicking flow artifacts.
Measurement of plaque composition requires a
series of steps for multi-contrast image review.
These typically include multi-contrast image reg-
istration and segmentation of plaque component
boundaries. Due to the small components and
complex nature of plaque constituents, special
algorithms targeted at atherosclerotic plaque anal-
ysis are typically used. For example, the plaque
J. Sun et al.
analysis software, CASCADE, developed at the
University of Washington, has processing rou-
tines for multi-contrast image registration and
semiautomated lumen and wall detection [27]. In
order to assist readers in measurement of plaque
components, a morphology-enhanced probabilis-
tic plaque segmentation (MEPPS) algorithm can
be used to automatically segment plaque compo-
nents [28]. The component boundaries are reader
correctable if required. Following the analysis,
total volume of plaque components as well as
percent volumes relative to plaque volume can
be calculated. A 3D distribution of plaque com-
ponents and plaque burden can also be used for
assessment as shown in Fig. 5.
Current Applications
Imaging-Based Patient Stratification
Patients with carotid atherosclerosis are managed
with stringent risk factor control that covers life-
style changes and medications. Nonetheless,
revascularization could be highly beneficial to
those with a high risk of future stroke. However,
this potential benefit needs to be evaluated against
non-negligible surgical risk for complications
such as perioperative myocardial infarction and
distal embolism on an individual basis. In the
NASCET study which compared carotid endarter-
ectomy with medical treatment alone in patients
with transient ischemic attack or nondisabling ische-
mic stroke within the previous 6 months, there
was an absolute risk reduction of 17  3.5 % at
2 years favoring surgery in patients with ipsilateral
high-grade stenosis (70–99 %) [29]. In contrast,
patients with mild stenosis (<50 %) did not benefit
from surgery, although the remaining risk with med-
ical therapy was still considerable at 18.7 % over
5 years. Risk reduction with carotid endarterectomy
in asymptomatic patients was more moderate, even
in the presence of high-grade stenosis, as shown in
ACAS and the Asymptomatic Carotid Stenosis
Trial (ACST), which estimated that about 17 patients
need to be treated with carotid endarterectomy to
prevent one stroke in 5 years [30, 31]. Improvement
in cost-effectiveness of carotid endarterectomy
7 Imaging of Carotid Atherosclerosis
Fig. 5 3D volume rendering of vessel wall morphology
and plaque components. Segmentation results of vessel
wall morphology and plaque components can be displayed
using 3D volume rendering to illustrate distribution of
in asymptomatic patients is urgent in light of the
much larger patient population with asymptomatic
carotid atherosclerosis. In fact, 76.7 % of the
795,000 strokes that happen each year are first
attacks.
The unmet need to identify who is at risk of
future stroke and therefore would benefit from
surgical intervention has highlighted the impor-
tance of imaging-based patient stratification in
clinical decision making. The aforementioned
clinical trials were performed at a time when the
assessment of carotid atherosclerosis was largely
based on angiography for measuring luminal
narrowing as the sole criterion to inform surgeons.
Since then, the understanding of carotid athero-
sclerosis biology has advanced to appreciate the
fundamental role of plaque rupture and thrombo-
sis in the pathophysiology of neurological ische-
mic events. In the new paradigm built on plaque
instability, luminal stenosis is more appropriately
viewed as one of the many risk factors for plaque
rupture rather than the primary cause of stroke. A
number of morphological and pathological fea-
tures that differentiate ruptured from
non-ruptured plaques can now be characterized
by carotid MRI, which holds the promise to be
used in prospective clinical trials that use
107
plaque burden and components. Orange indicates
intraplaque hemorrhage (arrow); blue, calcification; yel-
low, lipid-rich necrotic core
MRI-detected plaque characteristics to increase
survival benefits from carotid endarterectomy.
In a study comparing symptomatic and asymp-
tomatic carotid plaques that cause at least 50 %
stenosis, multi-contrast MRI was used to compre-
hensively characterize plaque morphological and
compositional features [32]. Fibrous cap rupture
as well as large necrotic core and intraplaque
hemorrhage were more prevalent in symptomatic
plaques, whereas no difference was found in lumi-
nal stenosis or plaque volumes. As such, plaque
compositional features appear to further differen-
tiate symptomatic plaques from those with a
benign natural history among plaques causing
severe stenosis. The convenience of using a single
T1-WI imaging sequence to detect intraplaque
hemorrhage has led to studies that specifically
examined the association of intraplaque hemor-
rhage with cerebrovascular events in broad
populations. Yamada et al. studied 392 carotid
arteries from 222 patients with suspected or con-
firmed carotid atherosclerosis using the MPRAGE
sequence [33]. Carotid arteries were categorized
as mild (0–29 %), moderate (30–69 %), or severe
(70–99 %) stenosis measured on contrast-
enhanced MR angiography. They noted that
carotid plaques with high signal intensity,
108
indicative of intraplaque hemorrhage, were more
likely to have had ipsilateral ischemic events within
the previous 6 months. The association between
ischemic events and MRI-detected ipsilateral
intraplaque hemorrhage was strongest in the
group with moderate stenosis, implying that infor-
mation on plaque characteristics may be particu-
larly helpful in selecting high-risk patients where
assessment of luminal stenosis is not able to
do. Howarth et al. compared ten symptomatic
patients and ten asymptomatic patients with a
focus on plaque inflammation measured with
USPIO-enhanced MRI [25]. More regions with
signal drop, indicative of more extensive macro-
phage infiltration, were observed in symptomatic
patients. These retrospective imaging studies
have reaffirmed the role of plaque instability in
the pathophysiology of ischemic stroke as noted
in histopathological studies and have extended the
finding to carotid atherosclerosis without severe
stenosis.
The prognostic value of plaque characteristics
has been demonstrated in prospective studies of
both symptomatic and asymptomatic patient
cohorts. Takaya et al. followed 154 consecutive
patients with asymptomatic 50–79 % stenosis.
Baseline plaque characteristics, including thin/
ruptured fibrous cap, intraplaque hemorrhage,
larger hemorrhage area, large necrotic core, and
maximal wall thickness, were predictive of subse-
quent ischemic events [34]. Most of these features
appear to predict recurrent strokes as well, as
shown by Kwee et al. in a study involving
126 symptomatic patients with 30–69 % stenosis
[35]. In particular, MRI detection of intraplaque
hemorrhage has been shown in various clinical
populations to be associated with increased
atherothrombotic risk. Three independent meta-
analyses were recently published which estimated
the predictive value of MRI-detected intraplaque
hemorrhage for planning future observational
studies and clinical trials [36–38]. Intraplaque
hemorrhage alone appears to be associated with
a more than tenfold increased risk of neurological
ischemic events in symptomatic patients.
Leveraging such prognostic information assess-
able via carotid MRI could prove to be invaluable
in future clinical practice.
J. Sun et al.
Carotid Artery Stenting
Carotid artery stenting has become a reasonable
alternative to carotid endarterectomy when
patients are considered for revascularization due
to its less invasive nature. A few randomized trials
have compared the two approaches in terms of
clinical outcomes including perioperative compli-
cations and long-term event rates. Although long-
term event rates are comparable between the two,
it has been noted that carotid artery stenting has a
higher perioperative risk of embolic stroke com-
pared to carotid endarterectomy [39]. This has
raised concerns for carotid artery stenting, as
those perioperative microembolizations adversely
affect patients’ cognitive function and has been
shown to lower the quality of life after procedure.
Plaque composition and thrombogenicity could
be a major determinant of the risk of distal embo-
lization in the setting of carotid artery stenting.
Yoshimura et al. tested the hypothesis in a study
involving 112 patients undergoing carotid artery
stenting [40]. Diffusion-weighted imaging was
performed before and after stenting to detect
new ischemic lesions due to intervention.
Pre-operation time-of-flight MR angiography
images were used to identify ipsilateral plaques
with intraplaque hemorrhage. Both new ischemic
lesions and perioperative symptoms were found to
be more frequent in patients with intraplaque
hemorrhage. Carotid MRI performed prior to sur-
gical intervention therefore may contribute to an
accurate evaluation of the risk of perioperative
microembolization, which can be translated into
improved clinical outcomes of both carotid artery
stenting and carotid endarterectomy.
Cryptogenic Stroke
As many as a third of ischemic strokes are con-
sidered cryptogenic, which means that no definite
etiology can be determined after clinical workup.
Imaging carotid atherosclerosis with MRI may
enable visualization of telltale features of culprit
lesions, which will facilitate targeted secondary
prevention strategies for individual patients by
revealing or excluding large artery atherosclerosis
7 Imaging of Carotid Atherosclerosis
as the underlying etiology. Freilinger
et al. performed a preliminary study in a series of
32 consecutive patients with clinically diagnosed
cryptogenic stroke and nonstenosing (<50 %)
carotid atherosclerosis [41]. A multi-contrast
MRI protocol detected complicated lesions,
including fibrous cap rupture, intraplaque hemor-
rhage, and mural thrombus, in 12 (37.5 %)
plaques ipsilateral to the cryptogenic stroke,
whereas none was found on the contralateral side
which showed MR signals consistent with
fibroatheroma and pathological intimal thickening
as the predominant lesion type.
For the purpose of differential diagnosis in
patients presenting with cryptogenic stroke, it is
preferable to use a protocol that affords extended
coverage as patients can have multiple tandem
lesions and the location of culprit features, if
any, is not known a priori. The newly developed
3D gradient echo-based sequences are showing
capability to cover both extracranial and intracra-
nial carotid arteries, which may be a viable solu-
tion in the near future.
In Vivo Insights on Pathobiological
Mechanisms
As an increasing number of biological targets
implicated in the pathogenesis of atherosclerosis
can now be successfully imaged with advances in
MR sequence design and post-processing, carotid
MRI has become a powerful tool for studying
disease mechanisms. The advantages of carotid
MRI as a research tool for clinical studies include
its noninvasive nature, absence of ionizing radia-
tion, high reproducibility facilitating serial imag-
ing design, and the opportunity of combining
different contrast weightings for better plaque
characterization.
Epidemiological studies have traditionally
used carotid ultrasound or coronary calcium scan
to measure subclinical atherosclerosis. Recent
studies have attempted to introduce MR plaque
imaging to epidemiological studies. This is a
novel concept as MR plaque imaging not only
provides measurement of subclinical atheroscle-
rotic burden but also allows phenotypic
109
classification based on plaque morphology and
composition. Several studies have utilized MR
imaging to look into the association of traditional
and novel atherosclerotic risk factors with plaque
composition, such as calcification, necrotic core,
and intraplaque hemorrhage, in population-based
study samples. Wasserman et al. studied 214 sub-
jects with carotid atherosclerosis enrolled in the
Multi-Ethnic Study of Atherosclerosis (MESA)
study [42]. After adjustment for potential
confounding factors, subjects in the middle and
highest tertile of total plasma cholesterol had a
higher likelihood of showing a lipid core com-
pared to the lowest tertile (odds ratio: 2.76 and
4.63, respectively). The MRI study of the
Rotterdam study participants found that
intraplaque hemorrhage was associated with a
number of clinical risk factors including age,
sex, hypertension, and smoking, whereas necrotic
core was associated with sex and
hypercholesterolemia [43].
Lesions in animal models are usually in the
early stage of atherosclerosis and lack important
advanced features seen in human lesions. This has
led to a significant knowledge gap in understand-
ing how atherosclerosis progresses from subclin-
ical lesions to clinically overt, culprit lesions.
Asymptomatic carotid atherosclerosis coupled
with in vivo imaging offers a unique opportunity
for clinical researchers to advance understanding
of important pathological events leading to clini-
cal events. In a case-control study comparing
carotid plaques with and without intraplaque hem-
orrhage, an 18-month imaging follow-up demon-
strated distinctively different progression patterns
between the two groups [44]. While plaques with-
out intraplaque hemorrhage tended to be station-
ary over the study period, those with intraplaque
hemorrhage showed a larger increase in wall vol-
ume and necrotic core volume, which was trans-
lated mainly into luminal narrowing. The ability
of MRI to serially track changes of carotid athero-
sclerosis was shown in a recent study, where mul-
tiple MRI scans with a constant time interval were
retrospectively retrieved from subjects with
intraplaque hemorrhage to examine the immediate
and long-term effects of intraplaque hemorrhage
on plaque progression [45]. Images spanning 4–5
110
years successfully captured how plaques
progressed from the stage without intraplaque
hemorrhage to the stage with intraplaque hemor-
rhage. Rather than a transient mass effect on
plaque volume, the development of intraplaque
hemorrhage was found to have gradually altered
the trajectory of plaque progression by posing an
immediate and long-lasting promoting effect on
plaque progression. Given that animal models do
not mimic advanced, clinical relevant plaques,
knowledge so-obtained from imaging human
carotid atherosclerosis is invaluable to the under-
standing of the natural history of high-risk
plaques, which may point to promising directions
for reducing residual cardiovascular risk in con-
temporary medicine.
Fig. 6 Serial MR scans showing plaque stabilization
under pharmacotherapy. Top panel: three consecutive
cross sections from the baseline scan show a plaque with
a big necrotic core, which is delineated as non-enhanced
areas on contrast-enhanced T1-WI (CE-T1) images. Bot-
tom panel: Matched cross sections from the follow-up scan
show morphological and compositional changes of the
J. Sun et al.
Efficacy Markers for Testing
Therapeutic Strategies
The ability of MRI to track morphological and
compositional changes of atherosclerotic plaques
has also led to research efforts using quantitative
MRI measurements of carotid atherosclerosis
as efficacy markers for testing therapeutic
strategies that aim to stabilize atherosclerotic
plaques (Fig. 6). As a surrogate for systemic ath-
erosclerosis, information on carotid atherosclero-
sis should give clues to the underlying coronary
atherosclerosis, in which vessel wall information
currently obtainable is still limited due to
technical constraints. This may be particularly
important in the current era due to several reasons.
same plaque, consistent with plaque stabilization. The
distinct lumen (L ) and outer wall boundaries (long arrows)
suggest a reduction in plaque volume. However, the most
dramatic change is the decrease in non-enhanced areas,
indicative of smaller necrotic core size, at follow-up. A
region with thin fibrous cap at baseline had increased cap
thickness at follow-up (hollow arrows)
7 Imaging of Carotid Atherosclerosis
First, the prevalence of atherosclerosis is high,
but the incidence of atherothrombotic complica-
tions of any given population is low, especially in
the primary prevention setting. Consequently,
while a high volume of myocardial infarction
and ischemic strokes are seen clinically, the
large number of subjects needed to power a
clinical trial has been daunting. Thus, early sig-
nals from surrogate markers that can display
the efficacy of experimental agents are highly
desirable. Second, novel surrogate markers,
to certain extent, target biological activities or
pathways in atherogenesis. Characterizing
changes in those biomarkers will help clarify
the mechanisms of actions exerted by the
experimental agent, which often deviate
from what are known from preclinical studies.
Third, it is common for patients with different
demographics or clinical conditions to have het-
erogeneous responses to the same treatment,
which has led to the common practice of
predefined subgroup analysis in clinical
trials. Given the dramatic morphological and
pathological differences, plaques at different
stages or with different phenotypes are likely to
respond heterogeneously to the same medication,
which, however, is less understood. Intraplaque
hemorrhage appears to be one such factor
influencing therapeutic response [46]. In this
regard, plaque characteristics may not only
serve as efficacy markers but also enable a
deeper understanding of the efficacy of
antiatherogenic medications in terms of individual
responses.
A number of studies have been conducted
using MRI-based plaque measurements as end
points, predominantly using statins. Corti
et al. studied 32 carotid plaques under the
treatment of simvastatin [47]. Reduction in wall
area but not lumen area was seen at 1 year. When
follow-up was extended to 2 years, there was a
further decrease in wall area and a slight
increase in lumen area. These findings indicate
that carotid atherosclerosis is subject to rapid
progression or regression. Change in wall area
appeared to precede change in lumen area.
Another study by Underhill et al. used
multi-contrast MRI to examine the effect of
111
rosuvastatin on carotid plaque morphology and
composition in more advanced lesions causing
16–79 % stenosis [48]. A significant reduction in
necrotic core size was noted at 2 years, but not in
wall volume, which suggests plaque lipid deple-
tion, and thus stability, may be one of the under-
lying reasons for clinical benefits with statin
therapy. The time course of plaque lipid depletion
under lipid-lowering treatment was later
addressed by Zhao et al. [49]. In 33 subjects
with documented necrotic core at baseline,
necrotic core size gradually decreased over
3 years. While a significant reduction in necrotic
core could be seen in year 1, reduction in wall
volume was not apparent until year 2, lagging
behind the change in necrotic core. When plaque
inflammation is targeted, previous studies have
shown a reduction in imaging signals within a
few weeks [50]. Tang et al. compared two doses
of atorvastatin treatment in 47 patients with over
40 % carotid stenosis and USPIO uptake at base-
line. The high-dose group showed a significant
reduction in USPIO-defined inflammation at
6 weeks [50].
The aforementioned studies are all single cen-
ter studies. However, in cases of large-scale
recruitment, short recruitment period, or clinical
conditions with low prevalence, multicenter
design is usually the norm. In view of the substan-
tial variations in technique instrumentation
between imaging sites, the recent completion of
a few multicenter carotid MRI studies represents a
solid step toward the standardization of MRI
protocol and quantitative image analysis. A mul-
ticenter clinical trial involving 13 imaging
sites and two vendor platforms at 3-T was carried
out to examine change in necrotic core size of
carotid atherosclerosis over a short 6-month
period [46]. Under current standard-of-care to
medically control modifiable cardiovascular risk
factors, it was shown that plaques without
intraplaque hemorrhage had a small but signifi-
cant reduction in necrotic core size. In contrast,
plaques with intraplaque hemorrhage had an
increase in necrotic core size and wall volume
and a decrease in lumen volume – a progression
pattern consistent with previous natural history
studies.
112
Summary
Carotid atherosclerosis is a common clinical find-
ing, which almost always involves the bifurcation
segment where recirculating flow is most
severe. Vessel wall MRI, with high soft tissue
contrast and flexibility of sequence design, affords
inherent advantages for imaging carotid
atherosclerosis, yet several key technical points
must be addressed to achieve satisfactory visual-
ization of vessel wall pathology, including flow
and fat suppression, submillimeter spatial resolu-
tion, and sufficient SNR within clinically applica-
ble scan times. Traditional 2D imaging has gained
success in providing clinically relevant informa-
tion of atherosclerosis, demonstrated by a broad
spectrum of applications in previous literature.
But newly developed 3D sequences hold the
promise for future clinical practice due to short-
ened scan times and increased scan coverage.
Atherosclerotic plaques display considerable
morphological and pathological variations.
Appropriate interpretation of image findings may
depend on a sensible understanding of both
pathology and the capability and limitations
of MRI.
With the changing landscape in the field of
atherosclerosis management including pharmaco-
therapy and surgical intervention, clinicians are
facing either a shortage of data or data that are
not generalizable to current practice, when
selecting between medical treatment, carotid end-
arterectomy, and carotid artery stenting. Further-
more, the limitations of luminal stenosis in
predicting patients’ future risk of ischemic stroke
are increasingly recognized. New opportunities
have emerged that leverage technical develop-
ments in MRI of carotid atherosclerosis to assist
clinical decision making. As such, management of
patients with symptomatic or asymptomatic
carotid atherosclerosis is awaiting significant par-
adigm shift. With continued efforts in clinical
studies using direct MR imaging of carotid ath-
erosclerosis, the clinical outcomes of carotid ath-
erosclerosis may eventually be improved to a new
level.
J. Sun et al.
References
1. de Weerd M, Greving JP, de Jong AW, Buskens E, Bots
ML (2009) Prevalence of asymptomatic carotid artery
stenosis according to age and sex: systematic review
and metaregression analysis. Stroke 40:1105–1113
2. Cai JM, Hatsukami TS, Ferguson MS, Kerwin WS,
Saam T, Chu BC, Takaya N, Polissar NL, Yuan C
(2005) In vivo quantitative measurement of intact
fibrous cap and lipid-rich necrotic core size in athero-
sclerotic carotid plaque: comparison of high-
resolution, contrast-enhanced magnetic resonance
imaging and histology. Circulation 112:3437–3444
3. Moody AR, Murphy RE, Morgan PS, Martel AL,
Delay GS, Allder S, MacSweeney ST, Tennant WG,
Gladman J, Lowe J, Hunt BJ (2003) Characterization
of complicated carotid plaque with magnetic resonance
direct thrombus imaging in patients with cerebral
ischemia. Circulation 107:3047–3052
4. Kerwin WS, O’Brien KD, Ferguson MS, Polissar N,
Hatsukami TS, Yuan C (2006) Inflammation in carotid
atherosclerotic plaque: a dynamic contrast-enhanced
MR imaging study. Radiology 241:459–468
5. Trivedi RA, Mallawarachi C, U-King-Im JM, Graves
MJ, Horsley J, Goddard MJ, Brown A, Wang L,
Kirkpatrick PJ, Brown J, Gillard JH (2006) Identifying
inflamed carotid plaques using in vivo USPIO-
enhanced MR imaging to label plaque macrophages.
Arterioscler Thromb Vasc Biol 26:1601–1606
6. Virmani R, Kolodgie FD, Burke AP, Farb A, Schwartz
SM (2000) Lessons from sudden coronary death – a
comprehensive morphological classification scheme
for atherosclerotic lesions. Arterioscler Thromb Vasc
Biol 20:1262–1275
7. Leitinger N, Schulman IG (2013) Phenotypic polariza-
tion of macrophages in atherosclerosis. Arterioscler
Thromb Vasc Biol 33:1120–1126
8. Virmani R, Ladich ER, Burke AP, Kolodgie FD (2006)
Histopathology of carotid atherosclerotic disease. Neu-
rosurgery 59:S219–S227, S3-S13
9. Spagnoli LG, Mauriello A, Sangiorgi G, Fratoni S,
Bonanno E, Schwartz RS, Piepgras DG, Pistolese R,
Ippoliti A, Holmes DR (2004) Extracranial
thrombotically active carotid plaque as a risk factor
for ischemic stroke. JAMA 292:1845–1852
10. Fisher M, Paganini-Hill A, Martin A, Cosgrove M,
Toole JF, Barnett HJ, Norris J (2005) Carotid plaque
pathology: thrombosis, ulceration, and stroke patho-
genesis. Stroke 36:253–257
11. Redgrave J, Lovett JK, Gallagher PJ, Rothwell PM
(2006) Histological assessment of 526 symptomatic
carotid plaques in relation to the nature and timing of
ischemic symptoms – the Oxford plaque study. Circu-
lation 113:2320–2328
12. Narula J, Nakano M, Virmani R, Kolodgie FD,
Petersen R, Newcomb R, Malik S, Fuster V, Finn AV
(2013) Histopathologic characteristics of
7 Imaging of Carotid Atherosclerosis
atherosclerotic coronary disease and implications of
the findings for the invasive and noninvasive detection
of vulnerable plaques. J Am Coll Cardiol
61:1041–1051
13. Yarnykh VL, Terashima M, Hayes CE, Shimakawa A,
Takaya N, Nguyen PK, Brittain JH, McConnell MV,
Yuan C (2006) Multicontrast black-blood MRI of
carotid arteries: comparison between 1.5 and 3 Tesla
magnetic field strengths. J Magn Reson Imaging
23:691–698
14. Balu N, Yarnykh VL, Scholnick J, Chu B, Yuan C,
Hayes C (2009) Improvements in carotid plaque imag-
ing using a new eight-element phased array coil at 3 T.
J Magn Reson Imaging 30:1209–1214
15. Parker DL, Goodrich KC, Masiker M, Tsuruda JS,
Katzman GL (2002) Improved efficiency in double-
inversion fast spin-echo imaging. Magn Reson Med
47:1017–1021
16. Yarnykh VL, Yuan C (2003) Multislice double
inversion-recovery black-blood imaging with simulta-
neous slice reinversion. J Magn Reson Imaging
17:478–483
17. Wang J, Yarnykh VL, Hatsukami T, Chu B, Balu N,
Yuan C (2007) Improved suppression of plaque-
mimicking artifacts in black-blood carotid atheroscle-
rosis imaging using a multislice motion-sensitized
driven-equilibrium (MSDE) turbo spin-echo (TSE)
sequence. Magn Reson Med 58:973–981
18. Wang J, Yarnykh VL, Yuan C (2010) Enhanced image
quality in black-blood MRI using the improved
motion-sensitized driven-equilibrium (iMSDE)
sequence. J Magn Reson Imaging 31:1256–1263
19. Li L, Miller KL, Jezzard P (2012) DANTE-prepared
pulse trains: a novel approach to motion-sensitized and
motion-suppressed quantitative magnetic resonance
imaging. Magn Reson Med 68:1423–1438
20. Balu N, Yarnykh VL, Chu B, Wang J, Hatsukami T,
Yuan C (2011) Carotid plaque assessment using fast 3D
isotropic resolution black-blood MRI. Magn Reson
Med 65:627–637
21. Fan Z, Zhang Z, Chung YC, Weale P, Zuehlsdorff S,
Carr J, Li D (2010) Carotid arterial wall MRI at 3T
using 3D variable-flip-angle turbo spin-echo (TSE)
with flow-sensitive dephasing (FSD). J Magn Reson
Imaging 31:645–654
22. Yarnykh VL, Yuan C (2002) T-1-insensitive flow sup-
pression using quadruple inversion-recovery. Magn
Reson Med 48:899–905
23. Ota H, Yarnykh VL, Ferguson MS, Underhill HR,
DeMarco JK, Zhu DC, Oikawa M, Dong L, Zhao
XH, Collar A, Hatsukami TS, Yuan C (2010) Carotid
intraplaque hemorrhage imaging at 3.0-T MR imaging:
comparison of the diagnostic performance of three
T1-weighted sequences. Radiology 254:551–563
24. Wang J, Bornert P, Zhao H, Hippe DS, Zhao X, Balu N,
Ferguson MS, Hatsukami TS, Xu J, Yuan C, Kerwin
WS (2013) Simultaneous noncontrast angiography and
intraPlaque hemorrhage (SNAP) imaging for carotid
113
atherosclerotic disease evaluation. Magn Reson Med
69:337–345
25. Howarth SP, Tang TY, Trivedi R, Weerakkody R,
U-King-Im J, Gaunt ME, Boyle JR, Li ZY, Miller SR,
Graves MJ, Gillard JH (2009) Utility of USPIO-
enhanced MR imaging to identify inflammation and
the fibrous cap: a comparison of symptomatic and
asymptomatic individuals. Eur J Radiol 70:555–560
26. Stuber M, Gilson WD, Schar M, Kedziorek DA, Hof-
mann LV, Shah S, Vonken EJ, Bulte JW, Kraitchman
DL (2007) Positive contrast visualization of iron oxide-
labeled stem cells using inversion-recovery with
ON-resonant water suppression (IRON). Magn Reson
Med 58:1072–1077
27. Kerwin WS, Xu D, Liu F, Saam T, Underhill HR,
Takaya N, Chu BC, Hatsukami TS, Yuan C (2007)
Magnetic resonance imaging of carotid atherosclerosis:
plaque analysis. Top Magn Reson Imaging
18:371–378
28. Liu F, Xu DX, Ferguson MS, Chu BC, Saam T,
Takaya N, Hatsukami TS, Yuan C, Kerwin WS
(2006) Automated in vivo segmentation of carotid
plaque MRI with morphology-enhanced probability
maps. Magn Reson Med 55:659–668
29. Barnett HJ, Taylor DW, Eliasziw M, Fox AJ, Ferguson
GG, Haynes RB, Rankin RN, Clagett GP, Hachinski
VC, Sackett DL, Thorpe KE, Meldrum HE, Spence JD
(1998) Benefit of carotid endarterectomy in patients
with symptomatic moderate or severe stenosis. North
American Symptomatic Carotid Endarterectomy Trial
Collaborators. N Engl J Med 339:1415–1425
30. Walker MD, Marler JR, Goldstein M, Grady PA, Toole
JF, Baker WH, Castaldo JE, Chambless LE, Moore
WS, Robertson JT, Young B, Howard VJ, Purvis S,
Vernon DD, Needham K, Beck P, Celani VJ,
Sauerbeck L, von Rajcs JA, Atkins D (1995) Endarter-
ectomy for asymptomatic carotid artery stenosis.
JAMA 273:1421–1428
31. Halliday A, Mansfield A, Marro J, Peto C, Peto R,
Potter J, Thomas D (2004) Prevention of disabling
and fatal strokes by successful carotid endarterectomy
in patients without recent neurological symptoms:
randomised controlled trial. Lancet 363:1491–1502
32. U-King-Im JM, Tang TY, Patterson A, Graves MJ,
Howarth S, Li ZY, Trivedi R, Bowden D, Kirkpatrick
PJ, Gaunt ME, Warburton EA, Antoun NM, Gillard JH
(2008) Characterisation of carotid atheroma in symp-
tomatic and asymptomatic patients using high resolu-
tion MRI. J Neurol Neurosurg Psychiatry 79:905–912
33. Yamada N, Higashi M, Otsubo R, Sakuma T,
Oyama N, Tanaka R, Iihara K, Naritomi H,
Minematsu K, Naito H (2007) Association between
signal hyperintensity on T1-weighted MR imaging of
carotid plaques and ipsilateral ischemic events. AJNR
Am J Neuroradiol 28:287–292
34. Takaya N, Yuan C, Chu BC, Saam T, Underhill H, Cai
JM, Tran N, Polissar NL, Isaac C, Ferguson MS, Gar-
den GA, Cramer SC, Maravilla KR, Hashimoto B,
114
Hatsukami TS (2006) Association between carotid
plaque characteristics and subsequent ischemic cere-
brovascular events: a prospective assessment with MRI
– initial results. Stroke 37:818–823
35. Kwee RM, van Oostenbrugge RJ, Mess WH, Prins
MH, van der Geest RJ, Ter Berg JW, Franke CL,
Korten AG, Meems BJ, van Engelshoven JM,
Wildberger JE, Kooi ME (2013) MRI of carotid ath-
erosclerosis to identify TIA and stroke patients who are
at risk of a recurrence. J Magn Reson Imaging
37:1189–1194
36. Saam T, Hetterich H, Hoffmann V, Yuan C,
Dichgans M, Poppert H, Koeppel T, Hoffmann U,
Reiser MF, Bamberg F (2013) Meta-analysis and sys-
tematic review of the predictive value of carotid plaque
hemorrhage on cerebrovascular events by magnetic
resonance imaging. J Am Coll Cardiol 62:1081–1091
37. Hosseini AA, Kandiyil N, Macsweeney ST, Altaf N,
Auer DP (2013) Carotid plaque hemorrhage on mag-
netic resonance imaging strongly predicts recurrent
ischemia and stroke. Ann Neurol 73:774–784
38. Gupta A, Baradaran H, Schweitzer AD, Kamel H,
Pandya A, Delgado D, Dunning A, Mushlin AI, Sanelli
PC (2013) Carotid plaque MRI and stroke risk: a sys-
tematic review and meta-analysis. Stroke
44:3071–3077
39. Brott TG, Hobson RW, Howard G, Roubin GS, Clark
WM, Brooks W, Mackey A, Hill MD, Leimgruber PP,
Sheffet AJ, Howard VJ, Moore WS, Voeks JH, Hop-
kins LN, Cutlip DE, Cohen DJ, Popma JJ, Ferguson
RD, Cohen SN, Blackshear JL, Silver FL, Mohr JP, Lal
BK, Meschia JF (2010) Stenting versus endarterec-
tomy for treatment of carotid-artery stenosis. N Engl J
Med 363:11–23
40. Yoshimura S, Yamada K, Kawasaki M, Asano T,
Kanematsu M, Takamatsu M, Hara A, Iwama T
(2011) High-intensity signal on time-of-flight magnetic
resonance angiography indicates carotid plaques at
high risk for cerebral embolism during stenting. Stroke
42:3132–3137
41. Freilinger TM, Schindler A, Schmidt C, Grimm J,
Cyran C, Schwarz F, Bamberg F, Linn J, Reiser M,
Yuan C, Nikolaou K, Dichgans M, Saam T (2012)
Prevalence of nonstenosing, complicated atheroscle-
rotic plaques in cryptogenic stroke. J Am Coll Cardiol
Img 5:397–405
42. Wasserman BA, Sharrett AR, Lai S, Gomes AS,
Cushman M, Folsom AR, Bild DE, Kronmal RA,
Sinha S, Bluemke DA (2008) Risk factor associations
with the presence of a lipid core in carotid plaque of
asymptomatic individuals using high-resolution MRI:
the multi-ethnic study of atherosclerosis (MESA).
Stroke 39:329–335
43. van den Bouwhuijsen QJ, Vernooij MW, Hofman A,
Krestin GP, van der Lugt A, Witteman JC (2012)
J. Sun et al.
Determinants of magnetic resonance imaging detected
carotid plaque components: the Rotterdam study. Eur
Heart J 33:221–229
44. Takaya N, Yuan C, Chu BC, Saam T, Polissar NL,
Jarvik GP, Isaac C, McDonough J, Natiello C,
Small R, Ferguson MS, Hatsukami TS (2005) Presence
of intraplaque hemorrhage stimulates progression of
carotid atherosclerotic plaques: a high-resolution mag-
netic resonance imaging study. Circulation
111:2768–2775
45. Sun J, Underhill HR, Hippe DS, Xue Y, Yuan C,
Hatsukami TS (2012) Sustained acceleration in carotid
atherosclerotic plaque progression with intraplaque
hemorrhage: a long-term time course study. J Am
Coll Cardiol Img 5:798–804
46. Sun J, Balu N, Hippe DS, Xue Y, Dong L, Zhao X, Li F,
Xu D, Hatsukami TS, Yuan C (2013) Subclinical
carotid atherosclerosis: short-term natural history of
lipid-rich necrotic core–a multicenter study with MR
imaging. Radiology 268:61–68
47. Corti R, Fuster V, Fayad ZA, Worthley SG, Helft G,
Smith D, Weinberger J, Wentzel J, Mizsei G,
Mercuri M, Badimon JJ (2002) Lipid lowering by
simvastatin induces regression of human atheroscle-
rotic lesions: two years’ follow-up by high-resolution
noninvasive magnetic resonance imaging. Circulation
106:2884–2887
48. Underhill HR, Yuan C, Zhao XQ, Kraiss LW, Parker
DL, Saam T, Chu B, Takaya N, Liu F, Polissar NL,
Neradilek B, Raichlen JS, Cain VA, Waterton JC,
Hamar W, Hatsukami TS (2008) Effect of rosuvastatin
therapy on carotid plaque morphology and composi-
tion in moderately hypercholesterolemic patients: a
high-resolution magnetic resonance imaging trial. Am
Heart J 155:581–584
49. Zhao XQ, Dong L, Hatsukami T, Phan BA, Chu B,
Moore A, Lane T, Neradilek MB, Polissar N,
Monick D, Lee C, Underhill H, Yuan C (2011) MR
imaging of carotid plaque composition during
lipid-lowering therapy: a prospective assessment of
effect and time course. J Am Coll Cardiol Img
4:977–986
50. Tang TY, Howarth SP, Miller SR, Graves MJ, Patterson
AJ, U-King-Im JM, Li ZY, Walsh SR, Brown AP,
Kirkpatrick PJ, Warburton EA, Hayes PD, Varty K,
Boyle JR, Gaunt ME, Zalewski A, Gillard JH (2009)
The ATHEROMA (Atorvastatin Therapy: Effects on
Reduction of Macrophage Activity) study. Evaluation
using ultrasmall superparamagnetic iron oxide-
enhanced magnetic resonance imaging in carotid dis-
ease. J Am Coll Cardiol 53:2039–2050
51. Balu N, Wang J, Dong L, Baluyot F, Chen H, Yuan C
(2009) Current techniques for MR imaging of
atherosclerosis. Top Magn Reson Imaging
20(4):203–215
 Carotid Artery Dissection
8
Olivier Naggara, Myriam Edjlali-Goujon, Eric Bodiguel,
Marie Pierre Gobin-Metteil, Denis Trystram,
Christine Rodriguez-Regent, Jean-Louis Mas,
Jean Francois Meder, and Catherine Oppenheim

Contents Mural Hematoma Consequences . . . . . . . . . . . . . . . . . . . . 119

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Pathogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
Traumatic Carotid Artery Dissection . . . . . . . . . . . . . . . . 117
Spontaneous Carotid Artery Dissection . . . . . . . . . . . . . 117
Insights on the Mural Hematoma . . . . . . . . . . . . . . . . . 118
Mural Hematoma Constitution . . . . . . . . . . . . . . . . . . . . . . 118
Mural Hematoma Location . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Clinical Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Local Symptoms and Signs . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Ischemic Symptoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Outcome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
Digital Subtraction Angiography . . . . . . . . . . . . . . . . . . . . 122
Ultrasound . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
Cross-Sectional Imaging: CT or MRI? . . . . . . . . . . . . . . 123
Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Medical Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Mechanical Revascularization Strategies . . . . . . . . . . . . 135

O. Naggara (*) Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

Service de Neuroradiologie, Centre Hospitalier Sainte- References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Anne, INSERM UMR894, Université Paris Descartes,
Sorbonne Paris Cité, Paris, France
Service de Radiologie pédiatrique, Faculté de Médecine,
Hôpital Necker Enfants Malades, Université Paris
Descartes, Sorbonne Paris Cité, Paris, France
e-mail: o.naggara@ch-sainte-anne.fr
M. Edjlali-Goujon • M.P. Gobin-Metteil • D. Trystram •
C. Rodriguez-Regent • J.F. Meder • C. Oppenheim
Service de Neuroradiologie, Centre Hospitalier Sainte-
Anne, INSERM UMR894, Université Paris Descartes,
Sorbonne Paris Cité, Paris, France
e-mail: m.edjlali@ch-sainte-anne.fr;
gobin-metteil@ch-sainte-anne.fr;
d.trystram@ch-sainte-anne.fr;
c.rodriguez@ch-sainte-anne.fr;
jf.meder@ch-sainte-anne.fr;
c.oppenheim@ch-sainte-anne.fr
E. Bodiguel • J.-L. Mas
Service de Neurologie, Faculté de Médecine, Centre
Hospitalier Sainte-Anne, Université Paris Descartes,
Sorbonne Paris Cité, Paris, France
e-mail: e.bodiguel@ch-sainte-anne.fr;
jl.mas@ch-sainte-anne.fr

# Springer Science+Business Media New York 2016 115

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_16
116
Abstract
Carotid artery dissections (CAD) are a leading
cause of nonatherosclerotic stroke in young
adults, responsible for up to 25 % of ischemic
strokes. The cornerstone of CAD pathophysiol-
ogy and diagnosis is the presence of a mural
hematoma of unknown etiology, possibly
caused by an intimal tear, a primary rupture of
vasa vasorum, or an underlying arteriopathy
impairing vasomotion. This mural hematoma
occurs within the media layer extending distally
and circumferentially. The diagnosis of CAD is
generally based on a suggestive clinical presen-
tation, exclusion of atherosclerosis, and support-
ive radiologic evidence. An early and reliable
diagnosis is important as anticoagulation or
antithrombotic treatment is recommended in
order to reduce the risk of an early thromboem-
bolic event. The optimal imaging method for the
diagnosis of CAD is still debated. Doppler ultra-
sound examination (DUS), the first-line screen-
ing tool, may show normal findings whenever
the mural hematoma results in subtle lumen
alterations or is located in a segment that cannot
readily be displayed by DUS. Despite the ability
of CT angiography to reveal imaging findings of
CAD, brain CT is known to be poorly sensitive
for the detection of ischemic lesions. Thus, cer-
vical contrast-enhanced MR angiography
(CE-MRA) coupled with T1-WI fat-suppressed
axial sequences has gradually replaced conven-
tional angiography for the diagnosis of cervical
artery dissection. However, the diagnosis of
CAD often remains challenging, and it is not
rare for the diagnosis of CAD to remain pre-
sumptive despite an extensive imaging work-
up. The diagnosis can be retrospectively con-
firmed on imaging follow-up by monitoring the
lumen healing or progression in response to
treatment. High-resolution MRI (HR-MRI) of
the arterial wall is a noninvasive imaging tech-
nique that provides a delineation of the lumen
and arterial wall. This technique, available at 1.5
and 3-T, which requires standard or dedicated
surface coils, has been extensively used for
carotid atherosclerosis and, more recently, for
carotid dissection.
O. Naggara et al.
Keywords
Carotid artery dissection • Mural hematoma •
High resolution MRI • Stroke
Introduction
Carotid artery dissection (CAD) is caused by the
formation of a mural hematoma, of the subpetrous
segment of the internal carotid artery (ICA).
Although the clinical presentation of CAD can
be benign, cervical dissections are a major cause
of ischemic stroke in middle-aged adults [1]. The
increased diagnosis of this condition in the last
decade can be attributed to the availability of
diagnostic imaging techniques, such as color
Doppler ultrasound (US), computed tomography
(CT), magnetic resonance (MR) imaging, and,
more recently, high-resolution MR imaging of
the arterial wall (HR-MRI). Although several
comprehensive reviews have focused on different
aspects of CAD such as risk factors or treatment,
the present review proposes a global overview of
the epidemiology, pathophysiology, and
predisposing factors, clinical and imaging charac-
teristics, and therapeutic options. Additionally,
published data on the outcome of CAD, including
mortality rates and recurrences, were summa-
rized. This review deliberately focuses on sponta-
neous (nontraumatic) CADs in adults because
these are managed by stroke physicians, whereas
traumatic ICAD and cases in children are more
commonly handled by surgeons and pediatricians
and can require different management strategies.
Moreover, purely intracranial dissections or aortic
dissections extending to the cervical arteries are
outside the field of this review because their
mechanisms, symptoms, and management are
different.
Epidemiology
The incidence of CAD in the general population is
low, around 2,6 (95 % CI 1.9–3.3) per 100,000
inhabitants per year [2]. Vertebral artery dissec-
tions (VAD) are less common than CAD, although
8 Carotid Artery Dissection
this difference has been reduced by the increased
availability of MRI, which enables an accurate
diagnosis of vertebral artery dissection [3].
In the North American study, CAD occurred in
the community at a mean age of 45, 8 years, which
is similar to reports from two large published
multicenter series of ICAD in European
populations (44,0 years [459 patients] and 45,3
years [696 patients]) [2, 4].
A slight predominance in men was reported in
the European multicenter hospital-based series
(53–57 %), whereas 0–52 % of the patients were
women in the North American population-based
study [2, 4]. CADs seem to occur more frequently
in men and at an older age (47,0 vs. 43,4 years)
than do vertebral artery dissections [2, 4]. These
age and gender differences are difficult to explain.
However, it is possible that risk factors or trigger-
ing events, such as traumas in young patients, are
more involved in the pathogenesis of VAD
than ICAD.
Pathogenesis
The pathophysiology of CAD remains controver-
sial. Patients with ICAD may have a constitu-
tional, at least in part, genetically determined
weakness of the vessel wall. Environmental fac-
tors such as minor trauma or acute infection could
act as triggers [5]. ICADs are usually classified as
either traumatic in case of major trauma preceding
the onset of clinical symptoms or spontaneous,
otherwise.
Traumatic Carotid Artery Dissection
Major Traumatism
Traumatic ICADs can occur as a result of major
penetrating or nonpenetrating traumas. The most
common cause of a traumatic dissection is blunt
(severe hyperextension) or penetrating (stab
wound) injury to the neck. CAD has been shown
to occur in about 1–2 % of patients who have had
blunt trauma, and the risk is also increased with
trauma-associated injuries, such as facial or skull-
base fractures, and traumatic brain injury [6].
117
The risk of ICAD is increased after major thoracic
injuries, whereas the risk of vertebral artery dis-
sections, over the scope of this review, is
increased in case of cervical spine fractures [6].
Minor Traumatism or Mechanical Trigger
Event
A dissection is termed spontaneous whenever
there is no history of major trauma. Major pene-
trating or nonpenetrating trauma to the artery is
often lacking, but precipitating events involving
hyperextension, rotation, or lateroversion of the
neck are often reported by patients with ICAD
[7]. These include whiplash injury, chiropractic
manipulation, various sporting activities, neck
movements, and severe coughing. The exact role
of these precipitating events in the pathophysiol-
ogy of ICAD is unclear but suggests a mechanical
contribution. The styloid process and hyoid bone,
both in the vicinity to the ICA, have been impli-
cated in the pathogenesis of CAD with shorter
distances between the styloid process and hyoid
bone and dissected ICA than in nondissection
carotid pathology [8, 9].
However, such traumas are recurrent and result
in ICAD in only a few individuals. If anatomic
factors contribute so strongly to CAD risk, one
could wonder why ICAD recurrence is rare. If
prior cervical trauma seems to be an important
environmental determinant of CAD, it is not an
independent outcome predictor [7]. Because of
the characteristics of most prior cervical trauma,
the term mechanical trigger event rather than
trauma may be more appropriate [7].
Spontaneous Carotid Artery Dissection
In the absence of a traumatic trigger, the etiology
often remains undefined, likely to be multifactorial,
including infections, environmental exposures, and
collagen abnormalities of the arterial wall.
Infection
An association of ICAD with history of recent
infection has been reported [10, 11], potentially
predisposing individuals to CAD via endothelial
damage or prothrombotic mechanisms [10–14].
118
A possible causal role of recent infection is
supported by a seasonal peak of CAD in autumn.
This pattern has been attributed to an increased
occurrence of infection and weather-related
changes in blood pressure, coagulation parame-
ters, and physical activity [13]. The failure to
prove a specific infective agent led to the hypoth-
esis that the inflammatory response rather than a
specific infective agent might be responsible for a
transient arteriopathy nonspecific inflammation-
mediated injury of the arterial wall linked to
CAD [15]. Reports of elevated C-reactive protein
(CRP) and peripheral leukocyte counts in CAD
support this hypothesis [16].
Genetic Risk Factors
Genetic factors might also have a role in the
pathophysiology of CAD, mainly as part of a
multifactorial predisposition. The prevailing the-
ory is that CAD is a multifactorial disease caused
by several genetic variants and environmental
factors, each probably having a modest and
potentially synergistic effect. The high preva-
lence of multiple dissection events and of long-
term (>1 year) recurrent dissections in patients
with a familial history of CAD indicates that a
specific predisposition for familial cervical dis-
sections may exist [15]. Less often (<2 % of
cases), ICAD has been associated with a known
monogenic connective tissue disease, such as
vascular Ehlers-Danlos syndrome type IV [15],
or, more rarely, in patients with Marfan’s
syndrome or other known monogenic connective
tissue disorders, such as osteogenesis imperfect
[15]. About half of patients with cervical dissec-
tions have abnormalities in skin connective tissue
that follow an autosomal-dominant inheritance
pattern [15].
Although these findings suggest that genetic
factors play a role in CAD [15], all published
genetic association studies on ICAD have been
underpowered. More accurate data on genetic
risk factors for CAD can only be obtained from
much larger multicenter genetic association stud-
ies, such as the ongoing genome-wide associa-
tion study within the Cervical Artery Dissections
and Ischemic Stroke Patients (CADISP)
consortium [17].
O. Naggara et al.
Transient Arteriopathy
The epidemiology of CAD with an incidence peak
in autumn, the clinical observation that some
patients with CAD develop multiple consecutive
dissections within weeks [18, 19], and the associ-
ation with hereditary connective tissue diseases,
as well as ultrastructural abnormalities of connec-
tive tissue in skin biopsy specimens, support the
hypothesis that a general susceptibility state is
involved in the pathogenesis of CAD.
The presence of an underlying vasculopathy is
also suggested by the fact that CAD patients com-
monly present with concomitant arterial abnor-
malities, such as fibromuscular dysplasia, aortic-
root dilation, hyper-distensibility [20], and
increased stiffness of the arterial wall, as well as
impaired spontaneous and endothelial-dependent
vasodilation [21]. About 50 % of patients with
CAD have connective tissue aberrations in their
skin, including composite collagen fibrils and
fragmentation of elastic fibers [22]. In addition,
signs of tissue weakening along the tunica media
and the tunica adventitia junction in superficial
temporal artery biopsy specimens of patients
with CAD but not of controls suggest the presence
of a generalized arteriopathy leading to impair-
ment of the stability of the arterial wall in patients
with CAD [23]. Association of CAD with visceral
artery dissection, or reversible cerebral vasocon-
striction syndrome [24], strengthens this inflamma-
tory hypothesis [5, 15]. This latter hypothesis is
further reinforced by high-resolution MRI findings
(see “High-Resolution MRI” section) [25, 26].
Insights on the Mural Hematoma
Mural Hematoma Constitution
As in aortic or coronary dissections, CAD is the
mere consequence of a wall hematoma, either
secondary to an intimal tear or to direct bleeding
within the arterial wall caused by ruptured vasa
vasorum. Current insight into the pathogenesis of
coronary arteries and thoracic aorta dissection
suggests that dissecting aneurysms are disorders
of the outer arterial layers, involving the vasa
vasorum. Histological evidences of pathological
8 Carotid Artery Dissection
alterations of the tunica adventitia and the tunica
media suggest that CAD is due to an “outside-in”
process starting in the outer arterial layers rather
than an “inside-out” injury starting from the endo-
thelium leading to endothelial rupture as the pri-
mary event [27].
Regardless of its mechanism, the mural hema-
toma can expand toward the intima or the adven-
titia, resulting in a stenosis or pseudoaneurysm
formation. Mural blood then dissects longitudi-
nally. Dissections can tear through the intima,
with the consequence for mural blood to enter
the lumen of the artery. Increased thickness of
the arterial wall by mural blood accumulation
causes compression of the lumen, causes pertur-
bation of the blood flow stream and of the vascular
endothelium, and, finally, causes activation of
platelets and the coagulation cascade. These
changes contribute to the formation of a luminal
thrombus. CAD may start from the luminal side at
the intimal surface and, through an intimal flap,
dissect into the media. The mural hematoma can
create a false lumen that might reconnect with the
true lumen, creating parallel circulatory channels,
separated by the intimal flap [28]. This false
lumen may either remain patent, resolve
completely, or thrombose and cause narrowing
of the true lumen. The mural hematoma some-
times extends between the media and the adven-
titia, leading to the formation of an aneurysmal
lesion.
Mural Hematoma Location
The commonest location for a spontaneous CAD
is the cervical segment 2–3 cm distal to the carotid
bulb, an area that is probably subject to most of the
stretch during extension or rotation of the neck.
Although several studies have described the prox-
imal anatomical location of spontaneous CAD,
data on the cranial extension of the mural hema-
toma are scarce. One study reported that the mural
hematoma extended in the cranial direction,
involving the petrous segment of the ICA in
nearly 75 % of CAD [29]. If CADs extend for a
variable length cranially, they do not usually past
the point of entry of the ICA into the petrous
119
temporal bone [29]. Since intracranial arteries
have no external elastic limitans and have a thin-
ner media and adventitia compared with cervical
arteries, intracranial extension of a CAD has a
high risk of intracranial rupture.
Mural Hematoma Consequences
Dissections can be divided into two groups. The
first group is mainly subintimal, and the dissec-
tions compromise the arterial lumen or cause
thrombus to enter the lumen. The second group
is subadventitial, and these dissections either per-
turb structures adjacent to the outer arterial wall.
Extracranial subintimal CAD causes symptoms
primarily through luminal compromise and pres-
ence of luminal clot. Ischemic symptoms and
infarction are caused by reduced perfusion in the
brain territory supplied by the artery or embolism.
Subadventitial cervical CAD can perturb the sym-
pathetic fibers located along the dilated artery and
might compress or cause ischemia to the lower
cranial nerves (IX–XII) that exit near the
skull base.
Clinical Features
The consequences of the mural hematoma are
local symptoms and signs including headache
and neck pain (supposedly attributable to a dis-
tension of the artery by the mural hematoma stim-
ulating pain-sensitive receptors), Horner’s
syndrome, or cranial-nerve palsies resulting from
the stretching of sympathetic-nerve and cranial-
nerve fibers by an enlarged carotid artery. The
mechanism by which CAD leads to cerebral or
retinal ischemia is thought to be embolic, from
intraluminal thrombi forming at the site of the
intimal tear. Hemodynamic infarcts are also pos-
sible. Occasionally, CAD can lead to subarach-
noid hemorrhage when the dissection extends
intracranially.
The classical triad of presentation of a CAD is
ipsilateral headache, facial or neck pain, and
Horner’s syndrome, followed hours or days later
by cerebral or retinal ischemia. This triad of
120
symptoms is found in less than one third of
patients, but the presence of two elements of this
triad strongly suggests the diagnosis [1, 5].
Local Symptoms and Signs
The following symptoms and signs can occur
separately or in combination: [2, 19, 30] Horner’s
syndrome, unusual neck pain or headache,
cranial-nerve palsy, tinnitus, and, rarely,
cervical-root injury. Horner’s syndrome and
cranial-nerve palsy occur in carotid artery dissec-
tions. Sudden-onset Horner’s syndrome, particu-
larly if associated with headache or neck pain or
with an ipsilateral ischemic stroke in the carotid
territory, can be considered to be specific to CAD
and should lead to urgent imaging of the cervical
arteries.
The characteristics of pain associated with
CAD are not specific and can sometimes mimic
migraine or even cluster headache. CAD with
isolated pain might be more common than
expected [31] and is more often caused by extra-
cranial vertebral artery dissection but can also be
caused by carotid artery dissections. The onset
type ranged from thunderclap headache to
progressive pain.
Cranial-nerve palsies are rare, representing less
than 7 % of CAD cases in large hospital-based
series. The hypoglossal nerve is the most com-
monly affected, followed by the CN IX and CN X
cranial nerves, which are topographically close to
the carotid artery in its cervical trajectory. The
most likely mechanism is compression of the
nerves by an enlarged carotid artery. Cranial-
nerve ischemia is another potential mechanism,
particularly in very rare cases of upper cranial-
nerve palsy.
Ischemic Symptoms
Stroke (transient ischemic attack [TIA] or cerebral
infarction) is the most frequent and severe mani-
festation of extracranial CAD. The clinical pre-
sentation of cerebral ischemia caused by CAD
does not differ from that of cerebral ischemia
O. Naggara et al.
attributable to other factors. Although sometimes
inaugural, it is most often preceded by local symp-
toms that started shortly before the ischemic
event, a finding that is highly suggestive of CAD.
Cerebral infarction and TIA are inaugural in
respectively 10–15 % and 10–20 % of cases
[30]. Middle cerebral artery territory is involved
in more than 80 % of cases [30]. In rare instances,
carotid artery dissection can (MCA) lead to retinal
ischemia [19, 32].
The literature provides support for several con-
cepts, regarding the primary cervical lesion and
the secondary intracranial lesion in CAD. An
autopsy case has provided the only direct demon-
stration of secondary intracranial emboli
[33]. Three studies described indirect signs of
emboli on transcranial Doppler monitoring stud-
ies [34–36], but the sample size of these studies
limits the significance of these results. According
to current concepts relating to mechanism and
stroke pattern, strokes in pial or perforating artery
territory are more likely to be embolic, whereas
border-zone infarcts are more likely to be hemo-
dynamic. Using this simplification scheme, sev-
eral authors discussed the mechanism of
infarction in CAD patients, with conflicting
results [37, 38]. Based on the stroke pattern on
CT, one group demonstrated an embolic mecha-
nism in 92.2 % of 65 patients [39], whereas others
reported a similar frequency of embolic and
hemodynamic infarcts in 11 patients
[38]. Recently, the occlusive intracranial throm-
bus in vivo was directly demonstrated using T2*
sequence. Other findings that substantiate the con-
cept of artery-to-artery embolism in the pathogen-
esis of stroke in CAD were the multiple acute
DWI lesions in the majority of patients and the
occurrence of pial artery or perforating artery ter-
ritory stroke in 96 % of patients [40–43].
If the simplification of the relationship between
the mechanism and stroke pattern provides an
easy approach to the presumed mechanisms of
cerebral infarction in most patients, one should
bear in mind that, even at the individual level,
both mechanisms may be at play and may rein-
force each other in causing the ischemic lesion.
Border-zone infarcts might result from mixed
mechanisms by impaired clearance of emboli in
8 Carotid Artery Dissection
hypoperfused regions. Indeed, during the consti-
tution of the mural hematoma, only the embolic
mechanism may explain the occurrence of an
infarction, whereas the hemodynamic mechanism
may appear progressively with the narrowing of
the lumen due to the growth of the mural hema-
toma. Once the lumen is severely stenosed, hemo-
dynamic impairment may encourage the
formation of secondary clots in border-zone
regions, impede the clearance of distal clots, and
increase the impact on cerebral tissues. The fact
that border-zone infarction rarely occurred in iso-
lation but mainly occurred in association with pial
or perforating artery territory infarction supports
the hypothesis of an impaired clearance of small
emboli [40].
Outcome
Clinical Outcome
Mortality rates in the acute phase of CAD are
generally low (<5 %) [2, 19], although higher
rates of up to 23 % have been reported in previous
series of subsets of patients with severe CAD. In
general, mortality after CAD might be
underestimated, as patients with severe forms of
CAD sometimes die before arterial imaging and
etiological diagnosis can be done. In support of
this possibility, data from studies on patients with
acute complete MCA infarction have suggested
that CAD could be a major cause of “malignant”
cerebral infarction. Functional outcome is good in
about three-quarters of patients with CAD, but
impact on quality of life and socioprofessional
integration can be important [1, 32]. The func-
tional outcome after CAD does not seem to be
better than in other types of ischemic strokes in
young individuals. Factors associated with a poor
functional outcome are presence of cerebral ische-
mia, arterial occlusion, carotid location, older age,
and a severe National Institutes of Health stroke
scale score at onset.
Recurrent Ischemic Events
Recurrent ischemic events seem to be rare,
although early recurrences (before the patient is
discharged from the stroke unit) are not always
121
accounted for in published series. The rate of
ischemic recurrences ranges from 0 % to 13,3 %
at 1 year, with the largest recurrence rates
observed when recurrent events that occurred
before the diagnosis of CAD are taken into
account [19, 44]. Recurrent ischemic events usu-
ally occur during the first weeks after the dissec-
tion. Factors associated with an increased risk of
recurrent ischemic events are multiple dissections
and a history of hypertension [19]. Although iso-
lated cases of ischemic strokes caused by chronic
dissecting aneurysms of the carotid artery have
been reported, two prospective series of aneurys-
mal CAD found no ischemic events after about
3 years of follow-up [44]. Similar rates of ische-
mic recurrences have been reported in 46 patients
with CAD with a transient stenosis or occlusion
and in 46 patients with CAD with a permanent
stenosis or occlusion [45]. However, in another
series of 130 consecutive patients with CAD,
ischemic recurrences were attributed to a worsen-
ing carotid stenosis in five of six cases [46].
Arterial Lesion Evolution
The proportion of patients with complete resolu-
tion of arterial abnormalities varies between stud-
ies: 46 % for stenoses, 33 % for occlusions, and
12 % for dissecting aneurysms in the general
population [46–48]. Complete resolution or stable
residual luminal irregularity was documented
after a median duration of 0,29 years and in
82 % of cases within the first year [49]. The like-
lihood of complete recanalization seems higher in
patients with CAD who present with only local
symptoms and signs [50]. Occlusions can lead to
residual stenoses, and residual aneurysms can
appear after the acute phase in initially stenotic
or occluded arteries [44].
Recurrences of Dissections
Recurrences of dissections are defined by a new
mural hematoma, remote the initial lesion. Early
recurrences of dissection in the days or weeks
after the initial event could be a manifestation of
a unique transient disorder, whereas late recur-
rences occurring several months or years later
could indicate an underlying connective tissue
weakness. Both early and late recurrences are
122
rare; they seem to be most frequent within the first
2 months after the initial event [19], but some
dissections that were diagnosed as multiple
might have occurred sequentially within a short
time frame. Late recurrences are possibly
underestimated as studies with long-term follow-
up are scarce [51]. Recurrent dissections were not
reported in the only population-based series with
follow-up data of 48 patients with CAD (mean
duration of follow-up: 7–8 years). In hospital-
based series, the rate of recurrent dissections
ranged between 0,6 % and 25,0 %. The recurrence
rate was probably overestimated in tertiary refer-
ral series. Risk factors for recurrent CAD are
younger age [18], a familial history of CAD,
vascular Ehlers-Danlos syndrome, and
fibromuscular dysplasia [32]. The prognosis of
recurrent CAD has been described as benign [51].
Imaging
Dissection of the craniocervical arteries, which
was once considered as an uncommon diagnosis,
has become increasingly recognized as a cause of
stroke in young and middle-aged patients. In part,
this increased recognition is due to increased use
of noninvasive imaging studies. The latter has
allowed the screening of larger numbers of
patients. In some populations, the rate of dissec-
tion of the craniocervical arteries has been diag-
nosed at rates 3–10 times greater than that
determined before the use of MRI, MR angiogra-
phy (MRA), and, more recently, CT angiography
(CTA) studies.
Digital Subtraction Angiography
In the past, the diagnosis of arterial dissection
relied on standard catheter angiography (DSA).
As some authors have pointed out, DSA is an
imperfect reference standard because of question-
able interobserver variability. Indeed, a number of
entities can mimic arterial dissection and DSA is
an invasive procedure. In addition, if DSA may
detect vessel wall irregularities, such as
fibromuscular dysplasia, it may miss dissections,
O. Naggara et al.
since it does not allow direct visualization of the
vessel wall, whenever the false lumen does not
opacify with contrast medium or when the lumen
contains thrombus. It has been replaced by non-
invasive imaging modalities, both for the primary
diagnosis and the follow-up of dissections.
Pathognomonic signs are double lumen and inti-
mal flap. These are detected in less than 10 % of
cases [52], while the following patterns are found
more commonly. The string sign, the angiographic
hallmark of CAD, is a long, tapered, usually eccen-
tric, and irregular stenosis that begins beyond the
ICA bulb and might extend to the skull base
[52, 53]. The “string and pearl” sign corresponds
to a focal narrowing with a distal dilatation. The
flame-shaped tapering of the lumen, a tapered
occlusion that spares the carotid bulb, is suggestive
of dissection. Although these patterns are indicative
of dissection, they are not specific. DSA findings of
spontaneous CAD may also include a dissecting
aneurysm, alone or associated with a tapered steno-
sis [52, 53], with changing lumen patterns over
time, often returning to a normal caliber.
Ultrasound
Noninvasive imaging is required for the detection
of possible CAD. Ultrasound imaging provides
direct visualization of pathological findings that
may be related to CAD and demonstrates
impaired lumen patency [54]. Pathognomonic
findings on gray-scale US include mural hema-
toma, an intimal flap, or coexistence of a true and
a false lumen (Fig. 1) [55]. Unlike DSA, US is
able to demonstrate a false lumen even if this
lumen is thrombosed.
Spectral Doppler US waveforms in CAD can
be normal or may demonstrate nonspecific fea-
tures also occurring in high-grade stenosis and
occlusions (damped spectral waveform with
lower amplitude and biphasic pattern, high-
resistance spectral waveform, or absence of flow
with no spectral Doppler waveform) [54].
The diagnostic sensitivity of US decreased if
CAD results in a low-grade stenosis [54].
Sturzenegger et al. showed high accuracy in
detecting a dissection with US in case of ICA
8 Carotid Artery Dissection 123

Fig. 1 Longitudinal color

(a, b) and power (c)
Doppler ultrasound image
demonstrating a
hypoechoic thickened wall,
a finding consistent with a
mural hematoma (a)
narrowing and displacing
the true lumen (b) of the
ICA with enlargement of its
external diameter to 7 mm
(normal 5 mm). An
abnormal high-resistance
spectral Doppler US
waveform is demonstrated
in the dissected lumen (d)

occlusion or high-grade stenosis but a 20 % sen- Magnetic Resonance Imaging

sitivity in low-grade stenosis [49]. Another limi-
tation is that US is less reliable for CAD outside
the “window” of the US examination such as
dissections located in the subpetrous segment
and the carotid canal or in case of small mural
hematoma [54, 55].
Cross-Sectional Imaging: CT or MRI?
MRI, including MR angiography (MRA) and high-
resolution MRI (HR-MRI), and CT angiography
(CTA) are the main noninvasive cross-sectional
imaging tools used for assessing CAD. The use of
MRI and CTA is twofold: (1) demonstration of a
mural hematoma and (2) evaluation of the conse-
quence of the mural hematoma on the lumen
patency. MRI is currently the recommended imag-
ing technique [5], thanks to its higher ability to
demonstrate arterial wall abnormalities and its bet-
ter specificity as compared with CTA. CTA can be
performed as a first-line screening tool, particularly
in case of trauma.
Dissections of the craniocervical arteries are best
assessed with cross-sectional fat-suppressed T1
sequence with cranial and caudal saturation pulses
and with MRA techniques. The former can pro-
vide direct visualization of the mural hematoma,
which is the pathological hallmark of a dissection,
and the latter allows noninvasive visualization of
arteries.
Lumen Imaging: MR Angiography
Contrast-enhanced MRA (CE-MRA) is the tech-
nique of choice, thanks to recent improvements in
MR gradients and software, which allow rapid
acquisitions with centric k-space filling and auto-
matic bolus triggering. Like DSA, CE-MRA can
demonstrate luminal irregularity; long narrowing
of the vessel, the so-called string sign; or occlu-
sion and pseudoaneurysm formation (Fig. 2).
CE-MRA also has the advantage of being easily
preceded with fat-suppressed T1-WI axial images
through narrowed segments to confirm or refute a
suspected dissection. The combination of cross-
sectional MRI and MRA is not only a powerful
124
Fig. 2 Luminal patterns of carotid artery dissection on
contrast-enhanced MR angiography (CE-MRA). (a, b)
Stenotic type of CAD. In (a) CE-MRA demonstrates a
long, tapered, eccentric, and irregular stenosis (arrows)
that begins beyond the ICA bulb and that extends to the
skull base, the “string” sign. In (b) focal eccentric stenosis
diagnostic modality but can also be used effec-
tively for noninvasive follow-up of CAD to mon-
itor lumen patency evolution, resolution of a
mural hematoma, or development of complica-
tions (Figs. 3, 4, and 5).
Mural Hematoma Imaging
Carotid artery dissection is characterized, on
fat-suppressed T1 sequence, by a narrowed eccen-
tric flow void, which is surrounded by a crescent-
shaped, hyperintense area expanding the vessel
diameter (Fig. 6) [53]. MR signal of the mural
hematoma has a similar temporal evolution than
intracerebral counterpart. [56] An acute mural
hematoma can be hypointense on T2- and T1-WI
images and therefore be difficult to distinguish
from an area of flow void. The mural hematoma
may therefore be missed on MRI within the first
48 h after CAD [53]. Thereafter, the hematoma
becomes of intermediate signal intensity on
T1-WI images and hyperintense on proton density
and T2-WI images (Fig. 7). After a few days, the
hematoma remains hyperintense on T1-WI images
for approximately 2 months (Figs. 7 and 8). Sub-
sequent loss of signal intensity can mask the
O. Naggara et al.
(arrow) that begins beyond the ICA bulb, with a distal
dilatation, the “string and pearl” sign. (c) Occlusive pattern
of CAD. Flame-shaped tapering followed by an occlusion
of the lumen that spares the carotid bulb. (d)
Pseudoaneurysmal pattern of CAD (arrow) associated
with a distal long-tapered stenosis (double arrows)
hematoma after 6 months, as it becomes isointense
with the surrounding soft tissue. These patterns are,
however, not specific for a dissection as fresh
thrombus due to an atheromatous or embolic occlu-
sion that may exhibit a similar evolution of MR
signal intensity. It may, indeed, be impossible to
distinguish an intraluminal clot from an occlusive
mural hematoma. In such cases, a crescentic thick-
ening with eccentric signal void proximal to com-
plete occlusion should suggest dissection rather
than atheromatous occlusion (Fig. 9).
At least four causes of false-positive MRI or
MRA findings have been reported [57].
Atheromatous plaques causing subtotal occlu-
sion can be hyperintense on T2-WI images and
mimic a CAD [57]. This rim is, however, typically
isointense on T1-WI images, whereas the
periarterial rim in subacute CAD is usually
hyperintense on T1-WI images [56].
Structures adjacent to arteries can simulate a
periarterial rim of hyperintense signal which can
mimic a mural hematoma or a pseudoaneurysm.
For instance, carotid arteries are surrounded
by fat. Systematic fat-saturation imaging is the
best solution.
8 Carotid Artery Dissection
Fig. 3 Evolution of a pseudoaneurysm type of carotid
artery dissection. The pseudoaneurysm (arrow) was initially
associated with a focal-tapered stenosis (a). Resolution of
Fig. 4 Evolution of a stenotic type of carotid artery dissection. The long, irregular stenosis (arrow) improves with
progressive normalization seen at 12 months
Turbulent or slow flow on MRI and MRA can
be another confounding factor for CAD by pro-
ducing hyperintense signal in the periphery of the
arterial flow void, along the surface of the arterial
lumen. For example, marked narrowing of the
extracranial segment of the internal carotid artery
125
the stenosis associated with a progression of the
pseudoaneurysm 2 months after onset of the CAD (b).
Progressive normalization seen at 6 and 12 months (c, d)
can produce hyperintense slow flow distally,
which can be manifested by loss of part or all of
the intracranial flow void. This signal can be mis-
taken for mural hematoma. Inflow phenomena
generally produce signal abnormalities that are
located centrally within the flow void rather than
126 O. Naggara et al.

Fig. 5 Twelve-month evolution of a pseudoaneurysm type of carotid artery dissection with true and false lumen
remaining patent

Fig. 6 Right carotid artery

dissection on contrast-
enhanced MR angiography
(CE-MRA) (a) and axial
fat-suppressed T1-WI
image (b). Carotid artery
dissection is characterized,
on fat-suppressed T1-WI by
a narrowed eccentric flow
void, which is surrounded
by a crescent-shaped,
hyperintense mural
hematoma expanding the
outer vessel diameter

peripherally. Homogeneity of the hyperintense
signal on all slices associated with vessel expan-
sion supports dissection rather than slow flow.
Flow-related enhancement in arteries and veins
can simulate mural hematoma. Bright signal in
arteries at the extreme of an imaging volume due
to entry of unsaturated spins (i.e., the so-called
entry slice phenomenon) has long been recog-
nized as a cause of abnormal arterial signal on
selected images. This phenomenon can simulate
mural abnormal signal or complete absence of the
expected flow void within an artery and be
mistaken for dissection. The recognition that the
image is obtained at the extreme of the imaging
volume (i.e., first or second image in a stack of
images) and correlation with the signal of the
arteries on other imaging sequences resolve the
issue.
A false-negative diagnosis of dissection on
MR is most likely at the very early stage, when
the mural hematoma is not yet hyperintense.
Indeed, at this stage, the blood products within
the hematoma have not yet progressed to the stage
of methemoglobin. In this case, the signal
8 Carotid Artery Dissection 127

Fig. 7 Temporal evolution

of MR signal of the mural
hematoma on T1 and T2
sequences. An acute mural
hematoma, e.g., within
2 days after onset, is
isointense on T2- and
T1-WI images. The
hematoma becomes of
intermediate signal
intensity on T1-WI images
and hyperintense on T2-WI
images after day 2. The
hematoma remains
hyperintense on both T2-
and T1-WI images for
approximately 2 months.
Subsequent loss of signal
intensity can mask the
hematoma after 6 months,
as it becomes isointense
with the surrounding soft
tissue (not shown)

Fig. 8 Temporal evolution of MR signal of the mural isointense to the surrounding tissue with bright heteroge-
hematoma on T1 sequence. Contrast-enhanced MR angi- neous signal of its peripheral part (b). It becomes homo-
ography (a) demonstrated a long irregular stenosis sparing geneously hyperintense at day 3 (c) and was no longer
the right ICA bulb. The mural hematoma was initially visible on fat-suppressed T1-WI image at day 60 (d)
128 O. Naggara et al.

Fig. 9 Occlusive cases of

carotid artery dissection. In
the first patient (a, b),
CE-MRA demonstrated
right ICA occlusion and
long stenosis of the left
ICA. On fat-suppressed
T1-WI image (b), it was
difficult in the right ICA
lumen to distinguish an
intraluminal clot from an
occlusive mural hematoma
(arrow). CAD diagnosis
was done, thanks to left ICA
mural hematoma (double
arrows). In the second
patient (c, d), CE-MRA
demonstrated left ICA
occlusion (arrow). On
fat-suppressed T1-WI
image (d), it was also
difficult distinguishing an
intraluminal clot from an
occlusive mural hematoma.
In such a case, the
crescentic hyperintense
thickening associated with
increase outer external
diameter of the occluded
ICA suggested dissection
rather than atheromatous
occlusion

intensity of the hematoma does not differ from
that of background fat-suppressed tissue. Other
signs, such as the presence of a pseudoaneurysm,
or an increase in external diameter may help in
these difficult cases.
In addition, whenever a CAD produces arterial
occlusion (as opposed to stenosis), the mural
hematoma may not be conspicuous against the
background of the hyperintense signal abnormal-
ity of an occluded arterial lumen on
fat-suppressed T1-WI images. In such cases,
although the arterial occlusion is recognized, the
cause may not be discernible.
The mural hematoma frequently extends cra-
nially to the petrous carotid segment in the case of
carotid artery dissection. This location is within
the limits of the field of view of standard brain
MRI. Consequently, apart from dedicated cervical
fat-suppressed T1 sequences, brain MRI may be
useful to demonstrate CAD. More than three-
quarters of such acute CAD can be diagnosed
using stroke brain MRI protocol only, including
DWI, FLAIR, T2* sequences, and time-of-flight
MRA (Fig. 10) [29]. This result might have clin-
ical implications for patients who are not initially
suspected of having CAD based on available
8 Carotid Artery Dissection
Fig. 10 Illustration of carotid artery dissection on brain
MRI sequences in five patients. Axial gradient-recalled
echo T2 (a), diffusion-weighted imaging (b), FLAIR (c),
and native slices of 3D time-of-flight angiography of the
history at the time of protocol and at institutions
where cervical CE-MRA is not coupled with MR
brain imaging. The interpretation of direct signs of
CAD on brain MRI is a learnable skill and
requires little specialized training and can also be
achieved in regular clinical practice.
High-Resolution MRI
High-resolution magnetic resonance imaging
(HR-MRI) is an attractive tool for the assessment
of carotid artery diseases. Using dedicated surface
radiofrequency coils at 1.5-T or standard coils at
3-T, it provides high-resolution images of the
arterial wall in vivo. Optimal image quality relies
on correct patient’s preparation, on adequate coil
positioning, and on the use of pulse sequences
with signal suppression of the flowing blood.
HR-MRI has initially been applied for the
in vivo analysis of carotid atherosclerosis with
now solid evidence that HR-MRI can identify
the major components of atherosclerotic plaque,
i.e., lipid core, hemorrhage, calcifications, as well
as the fibrous cap [58]. Beyond atherosclerosis,
HR-MRI can be used in case of suspicion of
cervical artery dissection, providing an excellent
129
circle of Willis (d). Sagittal T1-WI imaging (e). Increased
external diameter, crescentic mural thickening (arrows),
and eccentric lumen (arrows) were clearly seen in all
patients
visualization of the cervical arterial mural hema-
toma on both carotid and vertebral arteries [59].
High-Resolution MRI Using Surface Coils
at 1.5-T
The aim of MRI of the arterial wall of cervical
arteries is to obtain in plane, submillimetric
voxels. The loss of signal to noise induced by
reducing the voxel size (500  500 μm  2–3
mm) is compensated for by the use of phased-
array surface coils, which increases the signal-to-
noise ratio by 40 %. The so-called HR-MRI cor-
responds to images with submillimetric voxels
that imply the use of dedicated surface coils.
These coils are commercially available and com-
patible with most MR manufacturers. They con-
sist of a pair of multichannel coils. The coils,
made of flexible material, are placed bilaterally
around the neck. The carotids, being superficial
structures, are well suited for surface coil imaging.
To explore the patient with suspected CAD,
the radiologist should ensure that the coil is at the
level of the carotid-presumed lesion. A multicentric
study reports a rate of 90 % of interpretable MRI
exams performed for the purpose of carotid
130
bifurcation analysis. For a given magnetic field
when using the same surface coil, image quality
seems stable across different platforms [60].
The MR protocol comprises multiple
sequences in the axial plane, covering about
3 cm in height. Consequently, the slices must be
focused on the index lesion or artery.
As in atherosclerotic plaque exploration, it is
currently recommended to combine multiple con-
trasts to fully characterize potential mural hema-
toma. Pulse sequences designed for vascular
imaging are classed as “bright blood” or “black
blood,” depending upon the flowing blood signal.
Both of these sequences provide a good contrast
between the lumen and the arterial wall. Bright
blood sequence corresponds to 3D TOF gradient
echo sequences used for MR angiography. On this
sequence, the flowing blood signal is enhanced,
and the lumen is brighter than the vessel’s wall.
The lack of 180 refocusing pulse creates T2*-
sensitive signal. The aim of black blood
(BB) sequences is to cancel the signal of the
flowing blood so as to increase the contrast
between the lumen (in black) and the wall. For
that reason, the acquisition rate is synchronized to
the heartbeat and is preceded by a double-
inversion-recovery (IR) magnetization preparation
pulse. This process maximizes flow suppression
due to outflow and minimizes artifacts due to vessel
motion. The time of inversion (TI) for the double-
IR preparatory pulse is fixed as close to the null
point of the blood signal. The double-IR prepara-
tion and rapid SE acquisition are typically repeated
using 2RR cardiac gating until all the rapid SE
shots are acquired. Black blood sequence can be T2
or proton density (PD) weighted (2RR) or T1-WI
(1RR). By applying the double-IR preparation
just after the cardiac trigger, the slice-selective
re-inversion pulse is placed in the “end-diastolic”
part of the cardiac cycle. Because the TI is usually
close to 650 ms, the rapid SE acquisition is also
played out during diastole. A third inversion pulse
can be added for fat suppression.
Several principles should be followed when
designing an MR protocol for imaging cervical
arterial wall in CAD: (1) the coil positioning is of
importance and should be guided by lumen abnor-
malities seen on CE-MRA or ultrasonographic
O. Naggara et al.
examination. The surface coils should be posi-
tioned anteriorly and laterally, facing the mandib-
ular bone for the petrous portion of internal
carotid artery, whereas it should be placed poste-
rior and under the mastoid for the V3 segment;
(2) acquire an oblique scout view (3D phase con-
trast, 2D time of flight, 1 min) of the entire index
artery so as to visualize the plaque distribution and
the level of the carotid lesion; (3) acquire HR
images (voxel <500  500 μm in plane); (4) use
multicontrasts, by combining black and bright
blood sequences, to identify the mural hematoma
presence and age; (5) keep the total acquisition
time to less than 30 min. Indicative parameters at
1.5-T are detailed in Table 1 [58].
High-Resolution MRI at 3-T
Although most of the HR-MRI studies have been
performed on 1.5-T MR units, 3-T units are now
more widely available. Surface coils for carotid
imaging are commercially available for 3-T MR
units. These are slightly different from those used
on 1.5-T units (tuned to a different frequency and
adapted to the 3-T connectors and/or interfacing).
They provide significant improvement in signal-
to-noise and contrast-to-noise ratios [61],
allowing reduction in both the acquisition time
and the in-plane resolution. For a given acquisi-
tion time, thinner slices or larger anatomic cover-
age in the z-axis can be obtained. At 3-T, longer
repetition times are advised for PD- and T2-WI
images. Double-IR blood suppression techniques
are available at 3-T, but they require an adjustment
of the TI for the blood T1 at 3-T, i.e., slightly
longer TI for 3-T compared to that used for
1.5-T [61]. Demonstration of strong agreement
between 3 and 1.5-T HR-MRI of the carotid
plaques, particularly in the identification of ath-
erosclerotic plaque composition, including hem-
orrhage, supports the translation of histologically
validated 1.5-T criteria to 3-T carotid images
[62]. However, there are differences in the mea-
sured size of arterial wall hemorrhage measures,
larger at 1.5-T than at 3-T, suggesting that patients
enrolled in serial prospective trials of hematoma
measurement in CAD should be imaged at the
same field strength in order to minimize interstudy
errors during analysis [62].
8 Carotid Artery Dissection 131

Table 1 Indicative imaging parameters for 1.5-T high-resolution plaque MR imaging using dedicated surface coils [60]
Parameters 3D TOF T1 BB DP BB T2 BB
TR (ms) 30 1 RR 2 RR 2 RR
TE (ms) 6.9 9 16–20 50
FOV (mm) 130 130 130 130
Slice thickness (mm) 2.2 3 3 3
Matrix 288  224 352  256 256  512 256  512
NEX 2 3 2 2
Synchronization – Cardiac Cardiac Cardiac
Number of slices 20 8 8 8
Resolution (μm)a 254  254 254  254 254  254 254  254
Scan time (min)b 4 4–5b 4–6b 4–6b
PD indicates proton density, TR repetition time, TE echo time, TI inversion time, NEX no. of excitations, BB black blood
sequence with double-inversion-recovery preparation pulse
a
After zero filling interpolation
b
Reported scan times are based on a heart rate of 70 beats/min. Times shorten with faster rates

Fig. 11 1.5-T high-

resolution MRI using
surface coils. Axial
fat-suppressed T1-WI
images showing the mural
hematoma (a) and an
intimal flap (b) creating two
parallel circulatory
channels. Although most
often seen under CTA, the
intimal flap with the double
lumen sign can be seen on
HR-MR images

High-Resolution MRI Findings in CAD excellent depiction of the structural characteristics

Using standard MRI, an early and reliable identi-
fication of acute CAD might be impaired by the
limited spatial resolution, the tortuous course of
the arteries, the great variability in normal vessel
caliber, and the presence of the thick bony cover-
ing and adjacent veins. This is the case for dissec-
tion of the upper portions of the vertebral arteries
(V2 and V3 segments), outside the field of this
review, and for dissection of petrous ICA. In some
cases, the diagnosis of CAD remains presumptive,
further confirmed by sequential follow-up MR
showing signs of healing or progression in
response to treatment.
1.5-T HR-MRI has demonstrated its usefulness
for the assessment of CAD [59] and permits an
of the vessel wall and the vessel lumen in acute
CAD. The eccentric vessel lumen (occlusion, ste-
nosis, luminal thrombus) and the vessel wall
(crescent mural hematoma, pseudoaneurysm,
double lumen, intimal tear) of the dissected artery
can then be exquisitely analyzed using HR-MRI
(Fig. 11) [63–65]. The mural hematoma is well
defined between intimal and adventitial wall
boundary. The signal intensity of mural hematoma
varies with the time elapsed since the onset of
symptoms [56]. More specifically, the age of the
mural hematoma may be classified into acute,
early subacute, late subacute, and chronic hema-
tomas according to its signal intensities on the T1-
and T2-WI sequences. One study suggested that
132
the MR signal changes of mural hematoma had
only moderate agreement with the dynamics seen
in cerebral hematomas, even if it can be difficult to
determine the exact delay between the occurrence
of CAD and the HR-MRI scan. Initially, the mural
hematoma is isointense to the sternocleidomastoid
muscle on T1-WI and hypointense on T2 and TOF
imaging. At a mean delay of 5,8 days, mural
hematoma may appear hyperintense on all
sequences (Figs. 7 and 8). Progressively, the sig-
nal drops on all pulse sequences [56].
HR-MRI can help in distinguishing between
dissection and atherosclerosis in case of cervical
artery occlusion [66]. The signal of the mural
hematoma is usually homogeneous in CAD
while being heterogeneous in hemorrhagic ath-
erosclerotic plaques. In addition, unequivocal dis-
tinction between mural hematoma and thrombus
is allowed by HR-MRI.
Recent studies have shown the presence of
perivascular soft tissue signal changes, related to
periarterial edema (PAE) on 3-T HR-MRI in
patients with recent CAD. PAE can be detected
on a 1.5-T MR unit (Fig. 12), with a high preva-
lence (83.3 %) of PAE in the spontaneous CAD
group including 18 patients. These changes may
Fig. 12 1.5-T high-resolution MRI using surface coils.
Periarterial edema in spontaneous carotid artery dissection.
High-resolution fat-suppressed T2-WI MR image. Note
the signal changes in the soft tissues (arrows) surrounding
the right dissected ICA
O. Naggara et al.
correspond to a localized inflammation of the
arterial wall. Spontaneous CAD is more fre-
quently associated with the presence of PAE com-
pared to traumatic CAD [25]. Another study
demonstrated that a subset of patients with spon-
taneous CAD showed signs of a generalized tran-
sient inflammatory arteriopathy in contrast-
enhanced HR-MRI and [18F]-
fluorodeoxyglucose positron emission tomogra-
phy CT (PET-CT) [26]. This subset of patients
may be more prone to multiple dissections. Inter-
estingly, such HR-MRI pattern is similar to those
observed in inflammatory diseases, such as
Takayasu’s disease and giant cell arteritis. PAE
and inflammation could simply result from com-
pression by the mural hematoma. However, such
HR-MRI pattern was more frequently found in
spontaneous CAD and rarely observed in trau-
matic CAD despite a larger hematoma in this latter
group in one study [25]. This minimizes the like-
lihood of periarterial abnormalities simply
resulting from a mechanical compression. How-
ever, we cannot exclude the possibility that
inflammatory HR-MRI pattern could be associ-
ated with thrombogenicity after CAD rather than
with the CAD mechanism itself.
Because of the limited accessibility of HR-MRI
using dedicated surface coils and the long scanning
time inherent to the technique, the idea is not to
perform HR-MRI in all patients suspected of CAD
but rather to improve our imaging protocols to get
supportive imaging evidence in most patients. This
is already feasible using 3-T MR magnets with
standard multichannel head and neck coils. The
3-T’s increase signal-to-noise ratio can be targeted
at improving spatial resolution without the need of
dedicated surface coil. In daily practice, 3-T
fat-suppressed T1- and T2-WI sequences with
3 mm slice thickness challenge the high resolution
obtained using dedicated surface coil with 1.5-T
MR unit to image the cervical artery wall.
Recent technological innovations, including
parallel imaging, optimized k-space trajectories,
and variable flip angles, allow the acquisition, at
3-T, of fat-suppressed tridimensional T1-WI
sequences (3D T1), with neck and head coverage,
submillimetric voxels (spatial resolution of
8 Carotid Artery Dissection
Fig. 13 3-T fat-suppressed 3D FSE T1 sequence, with
luminal black blood effect. Axial (a), coronal (b), and
sagittal views (c) showing a bright T1 hematoma and its
extension. Fusion images combining a volume rendering
0.8  0.8  1 mm [3] before and 0.4  0.4 
0.5 mm [3] after interpolation), and acceptable scan
duration for clinical practice (total scan time in
between 5 and 7 min). In a preliminary study, it
was demonstrated that fat-suppressed 3D T1
sequence can demonstrate the mural hematoma
[64]. The 3D T1 sequence may offer several advan-
tages over standard axial T1 sequence: (1) absence
of inflow artifacts, thanks to a 3D acquisition and to
an outer volume suppression pulse; (2) a
noncardiac gated black blood effect due to the
presence of different intravoxel velocities in the
blood vessels, causing a phase dispersion, which
results in signal loss. This phenomenon is ampli-
fied with the use of long echo trains and low
refocusing flip angles; (3) homogeneous fat sup-
pression; (4) a tridimensional acquisition with
133
of contrast-enhanced MR angiography and 3D FSE T1
images on coronal (d, f) and axial (e) views: the
hyperintense mural hematoma is facing luminal
irregularities
submillimetric quasi isotropic voxels, allowing
for multiplanar reformations; (5) a large field of
view in the z-axis, allowing whole neck coverage.
This potentially allows 3D T1 to be used as a first-
line diagnostic tool, when the CAD location is
unknown, as opposed to HR-MRI using dedicated
surface coil sequences that are centered on the
abnormal segment of the dissected artery because
of a limited number of axial sections in clinically
compatible acquisition time (Fig. 13). Another
group confirmed the added value of the 3D T1
sequence, compared to the fat-suppressed 2D T1
sequence, in patients with acute or subacute carotid
or vertebral artery dissection using a standard
multi-array surface coil at 1.5-T, with the limitation
of incomplete fat saturation in the lower cervical
region [67].
134
CT Angiography
Multisection CT angiography (CTA) provides
high-resolution and high-contrast images of the
arterial lumen and vessel wall. Axial source
images provide excellent visualization of the
craniocervical arteries as a consequence of the
contrast difference between the enhanced arterial
lumen and the surrounding structures.
Postprocessing of the source data as MIPs or
curved multiplanar reformats MPR produces
images that closely resemble those obtained with
DSA or MRA, and voxel-rendering techniques
allow a 3D display of the vessels and surrounding
soft tissues. CTA of CAD has shown excellent
agreement with DSA [68]. A 100 % sensitivity
and specificity was achieved using the presence of
a narrowed eccentric lumen in association with
enlargement of the overall vessel diameter as a
criterion for acute carotid dissection. Other signs
of extracranial carotid artery dissection on CT
include stenosis, mural thickening, occlusion,
aneurysm formation, and thin annular contrast
enhancement [69].
Another advantage of CTA examinations over
MRI resides in identification of the rare but patho-
gnomonic intimal flap. The higher sensitivity of
CTA is likely related to both the faster acquisition
time and higher spatial resolution.
Treatment
Medical Treatment
Anticoagulants or antiplatelets are usually
recommended in the acute phase of CAD to pre-
vent primary or recurrent stroke. However, no
randomized trial has successfully compared the
efficacy of these treatments. The Cervical Artery
Dissection in Stroke Study (CADISS) is an ongo-
ing randomized multicentric prospective study
comparing antiplatelet therapy with
anticoagulation for patients with carotid and ver-
tebral artery dissection. Randomization must be
within 7 days of onset of symptoms. In the
absence of randomized data, the best evidence at
O. Naggara et al.
present is from a systematic meta-analysis of non-
randomized studies that reported on outcomes
with antiplatelets versus anticoagulants [70].
There was no significant difference in the rates
of death or disability between both treatment
groups. Thus, at present, empirical results are
used to choose between anticoagulants and
antiplatelet drugs, in a case-by-case-based deci-
sion. Anticoagulation is sometimes preferred over
antiplatelet drugs when there is a severe stenosis,
an occlusion (with a risk of embolization before
recanalization), or a pseudoaneurysm, provided
that the infarct is not too large. Anticoagulants
might also be preferred whenever a thrombus is
seen in the arterial lumen or when the presence of
multiple ischemic lesions in the same arterial terri-
tory suggests an embolic mechanism. Conversely,
antiplatelet drugs are commonly favored when
there is a general contraindication to anticoagulants
or when CAD is associated with a large infarct,
exposing the individual to a high risk of hemor-
rhagic transformation. Whether antiplatelets
should be preferred in cases of local symptoms
only is still a matter of debate. Antiplatelet drugs
might also be preferred to anticoagulants in the rare
case of intracranial extension of CAD that could
theorically lead to a subarachnoid hemorrhage.
The information of evolution of the mural
hematoma during the first week after initiation of
the treatment may be important. Since CAD is
characterized by a mural accumulation of blood,
the dissecting hematoma may enlarge under
anticoagulation, with subsequent lumen
narrowing. It was recently demonstrated, using
HR-MRI, that heparin did not promote enlarge-
ment of the mural hematoma, progression of lumi-
nal narrowing, or dissection recurrence much
more than aspirin [71].
The currently available observational studies
suggest that thrombolytic therapy should not be
withheld in patients in whom CAD is associated
with an acute ischemic stroke. However, the lack
of any trend toward a benefit of thrombolysis in
such patients explains the research for more effi-
cient treatment options including mechanical
revascularization strategies.
8 Carotid Artery Dissection
Mechanical Revascularization
Strategies
Recently, stent-supported angioplasty has been
used to treat CAD in case of neurological symp-
toms while on anticoagulation or as an alternative
to the traditionally accepted use of anticoagulation
[72]. Stenting permitted resolution of the stenosis
with immediate recanalization of the artery. In a
recent systematic review of 140 patients
(153 CAD), including traumatic (48 %, n = 64),
spontaneous (37 %, n = 49), and iatrogenic (16 %,
n = 21) CAD, the technical success rate of
stenting was 99 %, and the procedural complica-
tion rate was 1.3 %. At a mean angiographic
follow-up of 12.8 months, in-stent stenosis or
occlusion was encountered in 2 % of patients,
and at a mean clinical follow-up of 17.7 months,
neurological events were seen in 1.4 % of patients
[73]. Further trials comparing stenting with med-
ical management are needed to elucidate the role
of stenting in selected situations.
Summary
An early and reliable identification of CAD is of
importance in order to prevent ischemic brain
lesion. The diagnosis of carotid artery dissection
is currently based on (1) a suggestive clinical
presentation, (2) exclusion of atherosclerosis,
and (3) supportive radiologic evidence. Obtaining
radiological evidence of CAD may be a real chal-
lenge. The emergence of high-resolution, rapid
imaging methods has enabled MRI to noninva-
sively image the fine internal structure of the
cervical arterial walls. HR-MRI can be used in
case of suspicion of cervical artery dissection,
providing an excellent visualization of the cervi-
cal arterial mural hematoma. Because of the lim-
ited accessibility of 1.5-T HR-MRI and the long
scanning time inherent to the technique, the idea is
not to perform HR-MRI in all patients with CAD
but rather to improve our imaging protocols to get
supportive imaging evidence in most patients.
This is already feasible using 3-T MR magnets
135
with standard multichannel head and neck coils
and volumic fat-suppressed T1-WI sequence.
References
1. Schievink WI (2001) Spontaneous dissection of the
carotid and vertebral arteries. N Engl J Med
344:898–906
2. Lee VH, Brown RD Jr, Mandrekar JN, Mokri B (2006)
Incidence and outcome of cervical artery dissection: a
population-based study. Neurology 67:1809–1812
3. von Babo M, De Marchis GM, Sarikaya H, Stapf C,
Buffon F, Fischer U, Heldner MR, Gralla J, Jung S,
Simonetti BG, Mattle HP, Baumgartner RW, Bousser
MG, Arnold M (2013) Differences and similarities
between spontaneous dissections of the internal carotid
artery and the vertebral artery. Stroke 44:1537–1542
4. Touze E, Oppenheim C, Zuber M, Meary E, Meder JF,
Mas JL (2003) Early asymptomatic recurrence of cer-
vical artery dissection: three cases. Neurology
61:572–574
5. Debette S, Leys D (2009) Cervical-artery dissections:
predisposing factors, diagnosis, and outcome. Lancet
Neurol 8:668–678
6. Majidi S, Hassan AE, Adil MM, Jadhav V, Qureshi AI
(2014) Incidence and outcome of vertebral artery dis-
section in trauma setting: analysis of national trauma
data base. Neurocrit Care 21(2):253–258
7. Engelter ST, Grond-Ginsbach C, Metso TM, Metso AJ,
Kloss M, Debette S, Leys D, Grau A, Dallongeville J,
Bodenant M, Samson Y, Caso V, Pezzini A, Bonati LH,
Thijs V, Gensicke H, Martin JJ, Bersano A, Touze E,
Tatlisumak T, Lyrer PA, Brandt T, Cervical Artery D,
Ischemic Stroke Patients Study G (2013) Cervical
artery dissection: trauma and other potential mechani-
cal trigger events. Neurology 80:1950–1957
8. Renard D, Azakri S, Arquizan C, Swinnen B,
Labauge P, Thijs V (2013) Styloid and hyoid bone
proximity is a risk factor for cervical carotid artery
dissection. Stroke 44:2475–2479
9. Raser JM, Mullen MT, Kasner SE, Cucchiara BL,
Messe SR (2011) Cervical carotid artery dissection is
associated with styloid process length. Neurology
77:2061–2066
10. Grau AJ, Brandt T, Buggle F, Orberk E, Mytilineos J,
Werle E, Conradt, Krause M, Winter R, Hacke W
(1999) Association of cervical artery dissection with
recent infection. Arch Neurol 56:851–856
11. Guillon B, Berthet K, Benslamia L, Bertrand M,
Bousser MG, Tzourio C (2003) Infection and the risk
of spontaneous cervical artery dissection: a case-
control study. Stroke 34:e79–e81
12. Grau AJ, Brandt T, Forsting M, Winter R, Hacke W
(1997) Infection-associated cervical artery dissection.
Three cases. Stroke 28:453–455
136
13. Lindsberg PJ, Grau AJ (2003) Inflammation and infec-
tions as risk factors for ischemic stroke. Stroke
34:2518–2532
14. Rubinstein SM, Peerdeman SM, van Tulder MW,
Riphagen I, Haldeman S (2005) A systematic review
of the risk factors for cervical artery dissection. Stroke
36:1575–1580
15. Debette S, Markus HS (2009) The genetics of cervical
artery dissection: a systematic review. Stroke 40:
e459–e466
16. Genius J, Dong-Si T, Grau AP, Lichy C (2005)
Postacute c-reactive protein levels are elevated in cer-
vical artery dissection. Stroke 36:e42–e44
17. Debette S, Metso TM, Pezzini A, Engelter ST, Leys D,
Lyrer P, Metso AJ, Brandt T, Kloss M, Lichy C,
Hausser I, Touze E, Markus HS, Abboud S, Caso V,
Bersano A, Grau A, Altintas A, Amouyel P,
Tatlisumak T, Dallongeville J, Grond-Ginsbach C,
Group C (2009) Cadisp-genetics: an international pro-
ject searching for genetic risk factors of cervical artery
dissections. Int J Stroke 4:224–230
18. Schievink WI, Mokri B, O’Fallon WM (1994) Recur-
rent spontaneous cervical-artery dissection. N Engl J
Med 330:393–397
19. Touze E, Gauvrit JY, Moulin T, Meder JF, Bracard S,
Mas JL (2003) Risk of stroke and recurrent dissection
after a cervical artery dissection: a multicenter study.
Neurology 61:1347–1351
20. Calvet D, Boutouyrie P, Touze E, Laloux B, Mas JL,
Laurent S (2004) Increased stiffness of the carotid wall
material in patients with spontaneous cervical artery
dissection. Stroke 35:2078–2082
21. Baumgartner RW, Lienhardt B, Mosso M, Gandjour J,
Michael N, Georgiadis D (2007) Spontaneous and
endothelial-independent vasodilation are impaired in
patients with spontaneous carotid dissection: a case-
control study. Stroke 38:405–406
22. Brandt T, Orberk E, Weber R, Werner I, Busse O,
Muller BT, Wigger F, Grau A, Grond-Ginsbach C,
Hausser I (2001) Pathogenesis of cervical artery dis-
sections: association with connective tissue abnormal-
ities. Neurology 57:24–30
23. Volker W, Besselmann M, Dittrich R, Nabavi D,
Konrad C, Dziewas R, Evers S, Grewe S, Kramer SC,
Bachmann R, Stogbauer F, Ringelstein EB,
Kuhlenbaumer G (2005) Generalized arteriopathy in
patients with cervical artery dissection. Neurology
64:1508–1513
24. Mawet J, Boukobza M, Franc J, Sarov M, Arnold M,
Bousser MG, Ducros A (2013) Reversible cerebral
vasoconstriction syndrome and cervical artery dissec-
tion in 20 patients. Neurology 81:821–824
25. Naggara O, Touze E, Marsico R, Leclerc X, Nguyen T,
Mas JL, Pruvo JP, Meder JF, Oppenheim C (2009)
High-resolution mr imaging of periarterial edema asso-
ciated with biological inflammation in spontaneous
carotid dissection. Eur Radiol 19(9):2255–2260
26. Pfefferkorn T, Saam T, Rominger A, Habs M, Gerdes
LA, Schmidt C, Cyran C, Straube A, Linn J,
O. Naggara et al.
Nikolaou K, Bartenstein P, Reiser M, Hacker M,
Dichgans M (2011) Vessel wall inflammation in spon-
taneous cervical artery dissection: a prospective, obser-
vational positron emission tomography, computed
tomography, and magnetic resonance imaging study.
Stroke 42:1563–1568
27. Volker W, Dittrich R, Grewe S, Nassenstein I, Csiba L,
Herczeg L, Borsay BA, Robenek H, Kuhlenbaumer G,
Ringelstein EB (2011) The outer arterial wall layers are
primarily affected in spontaneous cervical artery dis-
section. Neurology 76:1463–1471
28. Caplan LR (2008) Dissections of brain-supplying
arteries. Nat Clin Pract Neurol 4:34–42
29. Naggara O, Soares F, Touze E, Roy D, Leclerc X,
Pruvo JP, Mas JL, Meder JF, Oppenheim C (2011) Is
it possible to recognize cervical artery dissection on
stroke brain mr imaging? A matched case-control
study. AJNR Am J Neuroradiol 32:869–873
30. Biousse V, D’Anglejan-Chatillon J, Touboul PJ,
Amarenco P, Bousser MG (1995) Time course of
symptoms in extracranial carotid artery dissections. A
series of 80 patients. Stroke 26:235–239
31. Biousse V, Woimant F, Amarenco P, Touboul PJ,
Bousser MG (1992) Pain as the only manifestation of
internal carotid artery dissection. Cephalalgia
12:314–317
32. Leys D, Bandu L, Henon H, Lucas C, Mounier-Vehier-
F, Rondepierre P, Godefroy O (2002) Clinical outcome
in 287 consecutive young adults (15 to 45 years) with
ischemic stroke. Neurology 59:26–33
33. Steinke W, Schwartz A, Hennerici M (1996) Topogra-
phy of cerebral infarction associated with carotid artery
dissection. J Neurol 243:323–328
34. Molina CA, Alvarez-Sabin J, Schonewille W,
Montaner J, Rovira A, Abilleira S, Codina A (2000)
Cerebral microembolism in acute spontaneous internal
carotid artery dissection. Neurology 55:1738–1740
35. Srinivasan J, Newell DW, Sturzenegger M, Mayberg
MR, Winn HR (1996) Transcranial doppler in the eval-
uation of internal carotid artery dissection. Stroke
27:1226–1230
36. Droste DW, Junker K, Stogbauer F, Lowens S,
Besselmann M, Braun B, Ringelstein EB (2001) Clin-
ically silent circulating microemboli in 20 patients with
carotid or vertebral artery dissection. Cerebrovasc Dis
12:181–185
37. Bonati LH, Wetzel SG, Gandjour J, Baumgartner RW,
Lyrer PA, Engelter ST (2008) Diffusion-weighted
imaging in stroke attributable to internal carotid artery
dissection: the significance of vessel patency. Stroke
39:483–485
38. Weiller C, Mullges W, Ringelstein EB, Buell U, Reiche
W (1991) Patterns of brain infarctions in internal
carotid artery dissections. Neurosurg Rev 14:111–113
39. Lucas C, Moulin T, Deplanque D, Tatu L, Chavot D
(1998) Stroke patterns of internal carotid artery dissec-
tion in 40 patients. Stroke 29:2646–2648
40. Morel A, Naggara O, Touze E, Raymond J, Mas JL,
Meder JF, Oppenheim C (2012) Mechanism of
8 Carotid Artery Dissection
ischemic infarct in spontaneous cervical artery dissec-
tion. Stroke 43:1354–1361
41. Naggara O, Morel A, Touze E, Raymond J, Mas JL,
Meder JF, Oppenheim C (2012) Stroke occurrence and
patterns are not influenced by the degree of stenosis in
cervical artery dissection. Stroke 43:1150–1152
42. Naggara O, Touze E, Seiller N, Gobin-Metteil MP, Mas
JL, Meder JF, Oppenheim C (2008) Asymmetry of
intracranial internal carotid artery on 3d tof mr angiog-
raphy: a sign of unilateral extracranial stenosis. Eur
Radiol 18:1038–1042
43. Iancu-Gontard D, Oppenheim C, Touze E, Meary E,
Zuber M, Mas JL, Fredy D, Meder JF (2003) Evalua-
tion of hyperintense vessels on flair mri for the diagno-
sis of multiple intracerebral arterial stenoses. Stroke
34:1886–1891
44. Touze E, Randoux B, Meary E, Arquizan C, Meder JF,
Mas JL (2001) Aneurysmal forms of cervical artery
dissection: associated factors and outcome. Stroke
32:418–423
45. Kremer C, Mosso M, Georgiadis D, Stockli E,
Benninger D, Arnold M, Baumgartner RW (2003)
Carotid dissection with permanent and transient occlu-
sion or severe stenosis: long-term outcome. Neurology
60:271–275
46. Arauz A, Hoyos L, Espinoza C, Cantu C,
Barinagarrementeria F, Roman G (2006) Dissection
of cervical arteries: long-term follow-up study of
130 consecutive cases. Cerebrovasc Dis 22:150–154
47. Charbonneau F, Gauvrit JY, Touze E, Moulin T,
Bracard S, Leclerc X, Mas JL, Meder JF (2005) Diag-
nosis and follow-up of cervical arterial
dissections–results of the sfnv-sfnr study. J
Neuroradiol 32:255–257
48. Leclerc X, Lucas C, Godefroy O, Nicol L, Moretti A,
Leys D, Pruvo JP (1999) Preliminary experience using
contrast-enhanced mr angiography to assess vertebral
artery structure for the follow-up of suspected dissec-
tion. AJNR Am J Neuroradiol 20:1482–1490
49. Sturzenegger M, Mattle HP, Rivoir A, Baumgartner
RW (1995) Ultrasound findings in carotid artery dis-
section: analysis of 43 patients. Neurology 45:691–698
50. Nedeltchev K, Bickel S, Arnold M, Sarikaya H,
Georgiadis D, Sturzenegger M, Mattle HP,
Baumgartner RW (2009) R2-recanalization of sponta-
neous carotid artery dissection. Stroke 40:499–504
51. Dittrich R, Nassenstein I, Bachmann R, Maintz D,
Nabavi DG, Heindel W, Kuhlenbaumer G, Ringelstein
EB (2007) Polyarterial clustered recurrence of cervical
artery dissection seems to be the rule. Neurology
69:180–186
52. Houser OW, Mokri B, Sundt TM Jr, Baker HL Jr,
Reese DF (1984) Spontaneous cervical cephalic arte-
rial dissection and its residuum: angiographic spec-
trum. AJNR Am J Neuroradiol 5:27–34
53. Provenzale JM, Morgenlander JC, Gress D (1996)
Spontaneous vertebral dissection: clinical, conven-
tional angiographic, ct, and mr findings. J Comput
Assist Tomogr 20:185–193
137
54. Benninger DH, Georgiadis D, Gandjour J,
Baumgartner RW (2006) Accuracy of color duplex
ultrasound diagnosis of spontaneous carotid dissection
causing ischemia. Stroke 37:377–381
55. Dittrich R, Dziewas R, Ritter MA, Kloska SP,
Bachmann R, Nassenstein I, Kuhlenbaumer G,
Heindel W, Ringelstein EB, Nabavi DG (2006) Nega-
tive ultrasound findings in patients with cervical artery
dissection: negative ultrasound in cad. J Neurol
253:424–433
56. Habs M, Pfefferkorn T, Cyran CC, Grimm J,
Rominger A, Hacker M, Opherk C, Reiser MF,
Nikolaou K, Saam T (2011) Age determination of
vessel wall hematoma in spontaneous cervical artery
dissection: a multi-sequence 3t cardiovascular mag-
netic resonance study. J Cardiovasc Magn Reson 13:76
57. Provenzale JM, Sarikaya B, Hacein-Bey L, Wintermark
M (2011) Causes of misinterpretation of cross-sectional
imaging studies for dissection of the craniocervical arter-
ies. AJR Am J Roentgenol 196:45–52
58. Oppenheim C, Naggara O, Touze E, Lacour JC,
Schmitt E, Bonneville F, Crozier S, Guegan-
Massardier E, Gerardin E, Leclerc X, Neau JP,
Sirol M, Toussaint JF, Mas JL, Meder JF (2009)
High-resolution mr imaging of the cervical arterial
wall: what the radiologist needs to know. Radio-
graphics 29:1413–1431
59. Bachmann R, Nassenstein I, Kooijman H, Dittrich R,
Kugel H, Niederstadt T, Kuhlenbaumer G, Ringelstein
EB, Kramer S, Heindel W (2006) Spontaneous acute
dissection of the internal carotid artery: high-resolution
magnetic resonance imaging at 3.0 Tesla with a dedi-
cated surface coil. Invest Radiol 41:105–111
60. Touze E, Toussaint JF, Coste J, Schmitt E,
Bonneville F, Vandermarcq P, Gauvrit JY, Douvrin F,
Meder JF, Mas JL, Oppenheim C (2007) Reproducibil-
ity of high-resolution mri for the identification and the
quantification of carotid atherosclerotic plaque compo-
nents: consequences for prognosis studies and thera-
peutic trials. Stroke 38:1812–1819
61. Yarnykh VL, Terashima M, Hayes CE, Shimakawa A,
Takaya N, Nguyen PK, Brittain JH, McConnell MV,
Yuan C (2006) Multicontrast black-blood mri of
carotid arteries: comparison between 1.5 and 3 Tesla
magnetic field strengths. J Magn Reson Imaging
23:691–698
62. Underhill HR, Yarnykh VL, Hatsukami TS, Wang J,
Balu N, Hayes CE, Oikawa M, Yu W, Xu D, Chu B,
Wyman BT, Polissar NL, Yuan C (2008) Carotid
plaque morphology and composition: initial compari-
son between 1.5- and 3.0-t magnetic field strengths.
Radiology 248:550–560
63. Naggara O, Louillet F, Touze E, Roy D, Leclerc X, Mas
JL, Pruvo JP, Meder JF, Oppenheim C (2010) Added
value of high-resolution mr imaging in the diagnosis of
vertebral artery dissection. AJNR Am J Neuroradiol
31:1707–1712
64. Edjlali M, Roca P, Rabrait C, Naggara O, Oppenheim C
(2012) 3d fast spin-echo t1 black-blood imaging for the
138
diagnosis of cervical artery dissection. AJNR Am J
Neuroradiol 34(9):E103–E106
65. Naggara O, Oppenheim C, Toussaint JF, Calvet D,
Touze E, Mas JL, Meder JF (2007) Asymptomatic
spontaneous acute vertebral artery dissection: diagno-
sis by high-resolution magnetic resonance images with
a dedicated surface coil. Eur Radiol 17:2434–2435
66. Bachmann R, Nassenstein I, Kooijman H, Dittrich R,
Stehling C, Kugel H, Niederstadt T, Kuhlenbaumer G,
Ringelstein EB, Kramer S, Heindel W (2007) High-
resolution magnetic resonance imaging (mri) at 3.0
Tesla in the short-term follow-up of patients with proven
cervical artery dissection. Invest Radiol 42:460–466
67. Cuvinciuc V, Viallon M, Momjian-Mayor I, Sztajzel R,
Pereira VM, Lovblad KO, Vargas MI (2013) 3d
fat-saturated t1 space sequence for the diagnosis of
cervical artery dissection. Neuroradiology 55:595–602
68. Leclerc X, Godefroy O, Salhi A, Lucas C, Leys D,
Pruvo JP (1996) Helical ct for the diagnosis of extra-
cranial internal carotid artery dissection. Stroke
27:461–466
69. Vertinsky AT, Schwartz NE, Fischbein NJ,
Rosenberg J, Albers GW, Zaharchuk G (2008) Com-
parison of multidetector ct angiography and mr
O. Naggara et al.
imaging of cervical artery dissection. AJNR Am J
Neuroradiol 29:1753–1760
70. Kennedy F, Lanfranconi S, Hicks C, Reid J,
Gompertz P, Price C, Kerry S, Norris J, Markus HS,
Investigators C (2012) Antiplatelets vs anticoagulation
for dissection: cadiss nonrandomized arm and meta-
analysis. Neurology 79:686–689
71. Machet A, Fonseca AC, Oppenheim C, Touze E,
Meder JF, Mas JL, Naggara O (2013) Does
anticoagulation promote mural hematoma growth or
delayed occlusion in spontaneous cervical artery dis-
sections? Cerebrovasc Dis 35:175–181
72. Lavallee PC, Mazighi M, Saint-Maurice JP,
Meseguer E, Abboud H, Klein IF, Houdart E,
Amarenco P (2007) Stent-assisted endovascular throm-
bolysis versus intravenous thrombolysis in internal
carotid artery dissection with tandem internal carotid
and middle cerebral artery occlusion. Stroke
38:2270–2274
73. Pham MH, Rahme RJ, Arnaout O, Hurley MC, Bern-
stein RA, Batjer HH, Bendok BR (2011) Endovascular
stenting of extracranial carotid and vertebral artery
dissections: a systematic review of the literature. Neu-
rosurgery 68:856–866, discussion 866
 Nonatherosclerotic Nondissection
Diseases of Carotid Artery 9
Santosh Kumar Kannath and T. R. Kapilamoorthy

Contents Carotid Aneurysms and Carotid Blowout . . . . . . . . 154

Carotid Aneurysm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140 Carotid Blowout Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Fibromuscular Dysplasia (FMD) . . . . . . . . . . . . . . . . . . 140 Miscellaneous . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
Natural History and Clinical Features . . . . . . . . . . . . . . . 140 Carotidynia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
Pathology and Imaging-Pathologic Correlation . . . . . 141 Eagle Syndrome or Stylohyoid Syndrome . . . . . . . . . . 157
Vasculitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
Takayasu Arteritis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
Giant Cell Arteritis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
Other Uncommon Vasculitis . . . . . . . . . . . . . . . . . . . . . . . 145
HIV Vasculopathy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Connective Tissue Diseases . . . . . . . . . . . . . . . . . . . . . . . . 147
Arterial Tortuosity Syndrome . . . . . . . . . . . . . . . . . . . . . . . . 147
Ehler-Danlos Syndrome Type 4 (EDS4) . . . . . . . . . . . . 147
Loeys Dietz Syndrome (LDS) . . . . . . . . . . . . . . . . . . . . . . . 148
Pseudoxanthoma Elasticum (PXE) . . . . . . . . . . . . . . . . . . 150
Marfan Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
Other Systemic Disease with Carotid
Involvement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
Neurofibromatosis Type 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
PHACE(S) Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Tuberous Sclerosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Behcet Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154

S.K. Kannath (*)

Neurointervention Center, Department of Imaging
Sciences and Interventional Radiology, Neurointervention
Center, Sree Chitra Institute of Medical Sciences and
Technology, Trivandrum, Kerala, India
e-mail: santhoshkannath@sctimst.ac.in;
santhoshkannath@rediffmail.com
T.R. Kapilamoorthy
Department of Imaging Sciences and Interventional
Radiology, Sree Chitra Institute of Medical Sciences and
Technology, Trivandrum, Kerala, India
e-mail: kapil@sctimst.ac.in

# Springer Science+Business Media New York 2016 139

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_38
140
Abstract
Nonatherosclerotic and nondissection diseases
of carotid are often unrecognized clinical
entity, primarily due to nonspecific imaging
findings and unavailability of diagnostic tests.
This entity comprises of wide spectrum of clin-
ical conditions such as nonspecific arteritis,
vasculitis, connective tissue diseases, infective
processes, and some acquired conditions.
Oftentimes, carotid involvement may be the
first and only manifestation of these diseases.
Ancillary clinical and examination findings
lend clue to the underlying disease process
and help the clinician to further investigate
into the specific pathology. The list is exhaus-
tive; however, some of the important disease
entities and their clinical implications are
reviewed in this chapter.
Keywords
Carotid aneurysms • Carotid stenosis •
Takayasu arteritis • Fibromuscular dysplasia •
Giant cell arteritis • Vasculitis • Connective
tissue disorders • Marfan syndrome • Ehler-
Danlos syndrome • Neurofibromatosis •
Arterial tortuosity syndrome • Carotidynia •
Eagle syndrome
Introduction
Though atherosclerotic manifestation of the carotid
artery is a more common and important vascular
pathology in middle and elderly patients, nonather-
osclerotic vascular diseases assume prominence in
younger population and remain an important cause
for stroke in this age group. Nonatherosclerotic
diseases may remain asymptomatic and are inci-
dentally detected or may present with catastrophic
events. Due to its variable age of presentation, it
may be confused with atherosclerotic pathologies
if the latter presents prematurely. Thus understand-
ing of this group of pathologies is vital to design
appropriate investigations for optimal therapeutic
management. In this chapter, some of the important
nonatherosclerotic and nondissection diseases of
carotid artery are reviewed.
S.K. Kannath and T.R. Kapilamoorthy
Fibromuscular Dysplasia (FMD)
Fibromuscular dysplasia is an idiopathic, segmen-
tal, nonatherosclerotic, and noninflammatory dis-
ease involving the small- and medium-sized
arteries, commonly affecting renal and carotid
arteries. The exact prevalence of the disease is
unknown. The prevalence of the symptomatic
disease based on earlier angiographic studies is
estimated to be between 0.3 and 3 %, though
obviously this could not be extrapolated to general
population due to the evident bias and its actual
incidence may even be much lower [1–3].
Natural History and Clinical Features
The etiology of FMD is not known, and various
factors such as familial predisposition, hormonal
influence, smoking, and mechanical forces are
attributed to the causation of FMD. FMD com-
monly affects the truncal portion of the internal
carotid artery which is prone for mechanical
stretching and shearing over the cervical verte-
brae. As this segment relatively lacks vasa
vasorum, an ischemic cause has also been attrib-
uted. A congenital theory presumes that the arte-
rial lesions are persistent embryonic intimal
cushions that normally appear in the main trunk
during fetal life. The fact that proximal cervical
ICA is formed from dorsal aortic arch lends cre-
dence to this theory [1, 4].
Cervicocranial FMD presents at an older age,
though classical FMD is commonly discovered in
younger and female population [4]. Carotid and
vertebral arteries are nearly as frequently affected
as renal arteries. FMD may be detected inciden-
tally or may have ischemic or hemorrhagic pre-
sentation. Stroke or transient ischemic attacks
(TIA) can occur due to thromboembolic phenom-
ena, dissection, or hemodynamic compromise.
Spontaneous carotid artery dissections are a
known complication, and FMD changes may be
identified in up to 15 % of spontaneous dissec-
tions [5]. Recent registry data showed incidence
of carotid dissections and aneurysms to be 75 %
and 21 %, respectively. Besides, vertebral
artery dissections are also not uncommon [6].
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery
The incidence of cerebral aneurysms is reported to
be 20–30 % in FMD, higher than the general
population. Subarachnoid hemorrhage is a
dreaded complication of aneurismal rupture.
Besides, rarer manifestations such as intracerebral
hemorrhage, carotidocavernous fistula or
vertebrovenous fistulas are also described. Rarely,
FMD can involve intracranial arteries such as
middle cerebral artery, basilar artery, and distal
carotid artery in isolation or as extension of cervi-
cal FMD. Intracranial FMDs are observed in pedi-
atric population and usually have fatal outcome
[1]. In some follow-up studies, the risk of recur-
rent TIAs or stroke is reported up to 5 % per year,
and hence it is prudent to keep such patients on
long strict clinical surveillance. Natural history of
incidental intracranial aneurysms in FMD is
unknown [1, 7].
Pathology and Imaging-Pathologic
Correlation
Three types of FMDs are identified histologically
depending on the arterial layer involvement. The
angiographic appearance can be correlated with
histological features; however, categorization
into different subtypes is successful only in few
entities. The various pathological subtypes
and its angiographic appearances are described
below.
1. Intimal fibroplasias: Occurs in less than 10 %.
There is circumferential thickening of intima
by collagenous tissue. It is usually seen in
children and young adults with no gender pre-
dilection. The lesions are localized or a long
segment; however, this feature is not specific to
this entity.
2. Medial hyperplasia: Occurs in 60–70 %. There
are areas of thickened fibrotic ridges alternate
with thinned media producing a beaded
appearance due to alternate hyperplasic change
and local aneurysm formation. The beads are
aneurismal outpouchings that exceed the nor-
mal proximal caliber of the artery, and constric-
tions represent fibrotic ridges. The appearance
is a diagnostic of FMD and usually occurs in
141
the distal carotid artery sparing the carotid
bulb. These lesions are stable and do not pro-
gress beyond 40 years.
3. Perimedial fibroplasias: Occurs in 10 %.
Lesions have a collagen collar of variable
thickness in the outer half or the media or at
media-adventitia junction. It can be localized
or diffused. Producing a corrugated appear-
ance, the caliber of outpouching does not
exceed the diameter of the normal proximal
artery.
4. Medial hyperplasia : Occurs in 1–2 % . There is
smooth muscle hyperplasia without fibrosis.
Beads are less numerous than other subtypes.
5. Periarterial fibroplasia: Fibrous thickening of
adventitia with cicatrix formation causes local/
diffuse area of narrowing in the artery . This is a
very rare type occurring in <1 % of cases.
Traditionally, angiography has been used to
diagnose FMD. The frequently affected segments
include mid and distal parts of ICA and vertebral
artery at C1 and C2 levels. Bilateral involvement
is common, and carotid and vertebral artery
pathologies may coexist. The most common
angiographic abnormality is the appearance of a
string of beads which is a characteristic of FMD
and usually suggests subtype medial FMD
(Fig. 1). Fine bristling of lumen may be noted
indicating the presence of microaneurysms.
Standing columns of spastic contractions are the
differential diagnosis. Tubular stenosis accounts
for 20 % of angiographic observations which may
be focal or diffuse. It is not related to specific
subtypes and can be confused with other vascular
diseases such as atherosclerosis or vasculitis.
Rarely, a weblike defect at the origin of internal
carotid artery is also described. Other rare atypical
manifestations include beading of external carotid
branches, diverticular outpouching of carotid
artery [1, 2, 4, 8].
Intracranial FMD is described in conjunction
with carotid involvement or as isolated findings.
Angiography may show a typical string of bead
appearance or multiple aneurysms and arterial
tortuosities. Intima is as commonly affected as
media, and most cases are reported in pediatric
population [2].
142
Fig. 1 Diagnostic angiogram of bilateral common carotid
artery in a 3-year-old boy reveals tapering and occlusion of
the right internal carotid artery and typical arterial beading
and moderate stenosis of proximal left internal carotid
Though typical angiographic appearance is
confirmatory of FMD, the test is invasive, and
similar findings can be observed by other nonin-
vasive studies such as US, CTA, or MRA. Distal
carotid, vertebral artery, and intracranial circula-
tion are better evaluated with CT or MRI, and in
the current scenario, angiography is only
performed if revascularization is planned. Besides
diagnosis, imaging also helps in the surveillance
of intracranial aneurysms.
Vasculitis
Takayasu Arteritis
Takayasu arteritis (TA) is a chronic inflammatory
disease of large arteries affecting predominantly
aorta, subclavian, carotid, renal, and pulmonary
arteries. Though reported from all over the world,
the disease is more prevalent in Asian population.
The affected individuals are usually young and
female, and since the initial symptoms are
nonspecific, the time to diagnosis is often delayed
by months to years. Initial inflammatory symp-
toms include anorexia, weight loss, malaise, fever,
myalgia, arthralgia, or carotidynia. The disease
shows progressive or relapsing remitting course
and finally burns out. The clinical presentation is
often related to the extent of compensatory collat-
eral development in the affected vascular territory,
S.K. Kannath and T.R. Kapilamoorthy
artery (a, b). MRI revealed bilateral parieto-occipital
region acute infarcts (not shown). Volumetric reconstruc-
tion shows moderate focal stenosis of midsegment of the
main renal artery (c)
and symptoms can significantly hamper the daily
activities in more than two thirds of TA
patients [9].
The cerebral manifestations of TA include
headache, dizziness, visual disturbances, stroke,
TIA, or rarely subarachnoid hemorrhage and
intracerebral bleeds. The reported incidence of
these symptoms is quite variable with minor
symptoms noted in more than 50 % of patients
and dreaded events such as TIA or stroke
observed in 10–35 % of patients [10, 11]. Ischemic
manifestations may be due to hemodynamic
insufficiency or thromboembolic phenomenon.
Intracranial stenosis is a rare important cause for
stroke in TA [12]. Subarachnoid hemorrhage sec-
ondary to aneurismal rupture is also rarely
reported in TA. The aneurysms are common in
the posterior circulation, and persistent hyperten-
sion and altered hemodynamic factors are thought
to play a role in its genesis. Necrosis and destruc-
tion of wall and elastic lamina are also reported;
however, primary arteritis seems unlikely [13].
Vasa vasorum (VV) is assumed to play an
important role in the pathogenesis of
TA. Inflammatory cells infiltrate the VV and incite
the chronic and granulomatous inflammatory
reaction that spreads across to involve the media
and adventitia layer of the artery. The adventitia
and media are thickened, and intimal layer shows
hyperplasia. The smooth muscles and elastin of
media are destroyed, and the panarterial layer is
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery
gradually replaced by diffuse or nodular fibrotic
lesions resulting in the development of stenosis or
eventual occlusion of the artery [9, 14, 15]. Exten-
sive destruction followed by inadequate fibrosis
response leads to the formation of arterial aneu-
rysms. The prognosis of TA has improved over the
last decade primarily due to earlier diagnosis, uti-
lization of imaging to assess disease progression,
and advances in medical therapy [16]. The major
morbidity or mortality is related to the develop-
ment of cardiac failure or stroke. Premature
atherosclerosis further adds to the increased cere-
bral risks. Diagnostic criteria introduced by
Ishikawa and the American College of Rheuma-
tology are purely clinical and demonstrated more
than 90 % sensitivity and specificity for TA.
European consortium for diagnosis of childhood
TA includes imaging cognates as well and had
reported very high sensitivity and specificity
nearing 100 % [9].
The gold standard imaging study for the eval-
uation of TA has been a conventional angiogram,
though recently, CT and MRI have been increas-
ingly employed to study the luminal and
extraluminal features with high degree of accu-
racy. Typical changes described in cerebral angi-
ography include:
1. Irregularity of contour of artery: This is the
earliest change seen and is mainly due to inti-
mal fibrous thickening.
2. Stenosis caused by circumferential thickening
indicating advanced disease.
3. Complete occlusion due to thrombosis in a
stenosed segment or due to the disease process
itself. Stenosis or occlusion often occurs at or
close to the origin of the branches of aorta.
4. Diffuse dilatation or ectasias due to destruction
of media leading to irregular uniform widening
of carotid.
5. Aneurysm saccular or fusiform : As there is a
thick adventitial covering rupture of blood ves-
sel which is uncommon.
6. Combination of lesions: Stenosis and
aneurysms.
The angiograms are infrequently performed
and are increasingly replaced by noninvasive
143
imaging modalities such as CT/MRI and
US. The angiography is performed as a part of
interventional procedure or to assess the hemody-
namic status of cerebral circulation in patients
presenting with TIA/stroke or hemorrhage.
Since carotid artery is superficial, US can
detect many of these changes involving common
carotid artery and internal carotid artery such as
stenosis, occlusion, or aneurysm formation.
Besides, it can demonstrate the extraluminal arte-
rial wall thickening and helps in the diagnosis as
well as monitoring of further progression. It also
provides information regarding the direction of
flow in carotid and vertebral artery, which is use-
ful while planning therapy. Limitations include
inability to visualize the arteries in entirety.
Cross-sectional imaging such as MRI and CT
can demonstrate mural thickening of the arterial
wall, extent of thickening, intraluminal thrombo-
sis, steno-occlusive lesions, collateral pathways,
and aneurysm formation. MRI has the advantage
of superior tissue characterization and helps to
delineate inflamed/fibrosed media and adventitia-
periadventitial regions and edematous intima [16].
Disease activity monitoring: Activity is best
followed up by hematological parameters includ-
ing erythrocyte sedimentation rate and C-reactive
protein assessment. Recent reports suggest good
correlation between the disease activity and the
delayed gadolinium enhancement of the wall of
the artery especially aorta. Thus MRI could
become an important modality in assessing the
disease progression of the patient. Caution needs
to be exercised; however, while interpretation as
MRI lacks specificity, wall edema and enhance-
ment may be observed in clinically dormant phase
as well [17, 18].
The imaging studies show stenosis and/or
occlusion of common carotid arteries, aneurysms
of unilateral or bilateral carotid arteries,
intraluminal thrombosis, and rarely dissection of
the carotid artery. Extensive collateralization of
external carotid artery through ascending cervical
and thyrocervical arteries may be visualized
reforming the internal carotid artery. Tortuous
vascular channels may be visualized within the
arterial wall which is in fact the vasa vasorum that
develops secondary to ischemia of the arterial wall
144
Fig. 2 This young lady presented with multiple events of
acute cerebral ischemia. Aortogram demonstrates tapering
of brachiocephalic artery, occlusion of bilateral subclavian
artery, and moderate intermediate length smooth stenosis
of proximal left common carotid artery (white arrows in a,
or cerebral parenchyma. Intracranial arteries may
show occlusion or arterial wall thickening
(Fig. 2). Rarely nonaneurysmal SAH has been
reported [19–21].
Giant Cell Arteritis
Giant cell arteritis (GCA), also known as temporal
arteritis or Horton disease, is the most common
primary systemic vasculitis affecting people older
than 50 years. The disease incidence increases
with age and shows mild female predominance.
Vasculitic involvement involves large- and
medium-sized arteries and frequently affects
cervicocranial arteries such as superficial tempo-
ral artery, posterior ciliary artery, and vertebral
artery. Aorta, proximal large aortic arteries, and
upper limb arteries are also involved in disease
process; however, intracranial arteries are not typ-
ically affected by this disease [22].
Clinical symptoms are variable and are related
to developing ischemia or nonspecific inflamma-
tory processes. Polymyalgic symptoms may be
seen in 30–50 % of GCA at the time of diagnosis
S.K. Kannath and T.R. Kapilamoorthy
b). Thin tortuous channels are seen along the course of the
right common carotid artery that reconstitutes antegrade
flow in the right internal carotid artery (c, d). MRI shows
old infarct in the right corona radiate (e)
or may develop later [22, 23]. Ischemic cranial
symptoms include headache, scalp tenderness, jaw
claudication, amaurosis fugax (AF), and partial or
complete visual loss. AF is an important premoni-
tory symptom portending worse visual outcomes if
not recognized and intervened at earliest. The
visual loss occurs due to narrowing or occlusion
of posterior ciliary arteries and rarely due to central
retinal artery occlusion, posterior ischemic optic
neuritis, or retrobulbar neuritis [24]. Cerebral
ischemic events are rare and are reported to occur
in 2–4 % of GCA patients [22, 25]. Ischemic events
are common in posterior circulation and may pre-
sent as TIAs or stroke. Atypical presentation
includes the presence of nonspecific constitutional
symptoms, unexplained fever, claudication pain of
extremities due to large vessel GCA involvement,
carotid tenderness, or development of aortic aneu-
rysm, which in the absence of systemic biomarkers
can delay timely diagnosis and institution of appro-
priate management.
Pathological studies demonstrated infiltration
of all the three layers of arterial wall by chronic
inflammatory cells such as lymphocytes, macro-
phages, and giant cells. The internal elastic lamina
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery
shows fragmentation, and smooth muscles are
destroyed. Typical histological findings such as
the presence of giant cells at the transition of
intima and media are found only in 50 % of the
cases. The hyperplasic changes of intima narrow
or obliterate the vascular lumen and contribute to
the ischemic pathogenesis. However, these
changes are neither specific nor sensitive, and
often a false negative result may occur due to
segmental nature of the disease [22].
Diagnosis of GCA is made as per the guide-
lines of the American College of Rheumatology
1990 classification criteria. Ultrasonography and
Doppler studies are valuable in the diagnosis of
GCA, though they are not commonly employed.
Ultrasonography may show a dark hypoechoic
circumferential arterial wall thickening termed as
“halo sign” that disappears 2–3 weeks after the
initiation of steroid therapy. A recent meta-
analysis revealed a sensitivity of 69 % and spec-
ificity of 91 % for unilateral halo sign. For bilat-
eral halo sign, the specificity approaches 100 %,
and some investigators initiate therapy without
confirmatory evidence from temporal artery
biopsy. Besides halo sign, stenosis or occlusion
of temporal artery could be identified; however,
these findings are neither specific nor sensitive.
Ultrasonography is operator dependent, and exper-
tise, knowledge of temporal artery anatomy, and
availability of high-resolution probes can affect the
diagnostic accuracy of GCA [22, 26, 27]. A recent
small study reported improved visualization of
carotid wall, and its vascularization and luminal
border by contrast enhanced ultrasonography
[28]. High-resolution MRI and MRA with 1.5T
and 3T may show inflammation of temporal artery
visualized as rim of thickening and enhancement
of the arterial wall with central dark lumen due to
spin flow (Fig. 3). The performance of MRI/MRA
was found to be comparable to USG and
Doppler [29].
Angiography is rarely performed these days
and is largely superseded by noninvasive modal-
ities. MRA and CTA are useful in detecting extra-
cranial vascular abnormalities of aorta or other
large arteries (Fig. 4). FDG-PET study has the
potential to detect early features of inflammation
145
of aortic or large arterial wall and has been found
to be useful in the diagnosis of GCA in its early
phase or when biopsy is negative. Also this
increased uptake in acute phase of GCA has
been correlated with the development of aortic
dilatation at late follow-up [30].
Other Uncommon Vasculitis
HIV Vasculopathy
Owing to improved survival rates among HIV
patients, the primary vascular pathology due to
HIV infection is increasingly being documented
in clinical practice. Young males are predomi-
nantly affected, and often the patients who are
immunocompetent may manifest the disease. Vas-
cular pathology may be aneurismal or occlusive in
nature and involve extracranial carotid as well as
intracranial arteries. Carotid involvement is noted
in 25 % of all HIV-related aneurysms, and majority
of these aneurysms present with clinical features
such as painful enlarging mass lesions, hoarseness
of voice, dysphagia, or stridor [31, 32]. Intracranial
aneurysms may be multiple large or giant fusiform
aneurysms, reported usually in children or isolated
saccular aneurysms in adults [33]. Among the
patients presenting with stroke, HIV vasculopathy
has been found to be the cause in 20 % of the
patients [34].
Exact pathogenesis in unclear and probably
involves multiple factors. Pathological studies
had demonstrated infiltration of vasa vasorum by
neutrophils, macrophages, and plasma cells. The
resultant vasa vasoritis leads to ischemia and
destruction of arterial wall and incites the devel-
opment of arterial aneurysms. There is fragmen-
tation and loss of internal elastic lamina, thinning
of outer arterial wall, and intimal fibrosis. Bland
thrombus occluding the lumen may also be dem-
onstrated. Opportunistic bacterial infection of
arterial wall is a distant possibility and should be
evaluated for if tissue specimen is available for
analysis [32, 34–36].
Vascular abnormalities are detected during rou-
tine surveillance or while evaluating specific
146
Fig. 3 Biopsy has proven giant cell arteritis. Extensive
aortic and arch vessel wall thickening and severe luminal
narrowing and irregularity are noted (a, b, d). Intraluminal
clinical manifestations. Ultrasound is widely
available and can easily demonstrate pseudo-
aneurysms, stenosis, or occlusion of extracranial
carotid artery. Pseudoaneurysms are commonly
localized to distal common carotid artery or
carotid bifurcation, and it is visualized as blow-
out defect in the arterial wall with turbulent color
flow. The adjacent arterial wall shows thickening
and hyperechoic spotting [37]. CT and MRI can
S.K. Kannath and T.R. Kapilamoorthy
thrombus is visualized in the proximal left common carotid
artery (arrow in panel c)
better delineate these pathologies and are useful
in detecting intracranial vascular abnormalities
as well. Intracranial aneurysms are common in
the circle of Willis, first- or second-order
branches of MCA, ACA, or PCA. Besides diag-
nosis, imaging is also useful in monitoring the
growth of the aneurysm and in assessing the
response to retroviral therapy as some of the
lesions resolve with treatment [31, 34, 38].
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery
Fig. 4 Giant cell arteritis. Fat saturated T1W contrast
enhanced MRI (a, b) shows soft tissue enhancement in
the right temporal fossa (arrow in a) which is readily
apparent when compared to opposite side. Thickening of
Connective Tissue Diseases
Arterial Tortuosity Syndrome
Arterial tortuosity syndrome is a rare autosomal
recessive connective tissue disorder of large- and
medium-sized arteries due to mutations of
SL2A10 gene encoding the glucose transporter
receptor GLUT10. Impaired GLUT10 activity
leads to upregulation of TGF beta activity in the
arterial wall and results in poor elastogenesis and
extracellular matrix formation. The hallmark
findings include arterial tortuosity and elonga-
tion, aneurysms, and dissections of large- and
middle-sized arteries, focal coarctation of aorta,
stenosis of great vessels of the arch, and pulmo-
nary artery (Fig. 5). Associated extravascular
147
arterial wall is subtly appreciable in source images of TOF
MRA (c, d), highlighted in panel e (arrow). The thickened
arterial wall shows contrast enhancement (f)
features include dysmorphic facies, skin laxity,
hyperextensible joints, arachnodactyly, high
arched palate, and bilateral inguinal hernia
which may mimic other CTDs such as Ehler-
Danlos or Marfan syndrome [39–41].
Ehler-Danlos Syndrome Type 4 (EDS4)
Also known as vascular EDS, EDS4 is an autoso-
mal dominant CTD caused by heterogeneous
mutation of COL3A1 gene located in the chromo-
some 2 and encodes type 3 collagen that predom-
inates the walls of arteries and abdominal hollow
visceral organs. This entity is rare and constitutes
<10 % of all EDS patients. The arterials walls are
fragile and thin and are easily susceptible to shear
148
Fig. 5 Arterial tortuosity syndrome in a 3-month-old child. Arch, brachiocephalic artery, left common carotid artery, and
descending aorta are tortuous and elongated (a–c). The left brachial artery is small in caliber and appears quite tortuous (d)
stress injuries especially in large- and medium-
sized arteries [42–45]. The four cardinal features
that help in the diagnosis of EDS4 include dys-
morphic facial features, thin skin with visible
subcutaneous veins, bruises, echymosis or hema-
tomas, and features of fragility or rupture of artery,
digestive tract, or uterus. The typical features of
EDS such as skin and joint laxity are however
uncommon in this entity. The diagnosis is con-
firmed by demonstration of deficient synthesis of
type 3 procollagen by cultured skin fibroblasts or
detection of mutation of COL3A1 gene.
Major catastrophic events such as visceral or
arterial rupture are uncommon in childhood; how-
ever, nearly 80 % suffer from at least one such
episode by 40 years. The average life expectancy
is around 50 years, and cause of death is due to
vascular complications in more than 90 % of the
patients [45, 46]. The carotid and intracranial
events are reported to occur in about 15 % of
patients. The typical complications include arte-
rial dissection, carotidocavernous fistula, carotid
or intracranial aneurysms, and arterial rupture
(Figs. 6 and 7). The EDS4 remains an important
differential diagnosis for ischemic stroke in youn-
ger patients [44, 47]. Due to inherent weakness of
S.K. Kannath and T.R. Kapilamoorthy
vessel wall, the arteries are easily damaged while
obtaining arterial access or during the cerebral
angiography. Catastrophic and life-threatening
complications are reported during the procedure,
and hence invasive diagnostic tests are best
avoided unless performed as a part of interven-
tional procedure [48–50]. Imaging surveillance by
CT/MRI or US is advocated to detect or monitor
the progress of vascular disease.
Loeys Dietz Syndrome (LDS)
LDS is an autosomal dominant connective tissue
disease caused by the heterozygous mutation of
genes encoding the receptors such as TGFBR1,
TGFBR2, SMAD3, and TGFB2 that are involved
in the transformation growth factor signaling
pathway. Phenotypic features include aortic aneu-
rysms, arterial tortuosity, hypertelorism, bifid
uvula, or cleft palate. The aortic aneurysms are
frequently observed in LDS patients, and it
remains the major cause of morbidity and mortal-
ity [51, 52]. Besides, generalized arterial tortuos-
ity, tortuosity of head and neck arteries,
dissections, and intracranial aneurysms are also
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery 149

Fig. 6 A middle-aged lady presenting with left carotico- proximal cavernous internal carotid artery (arrow in a).
cavernous fistula with clinical features of EDS. Dilatation Panel b shows direct carotico-cavernous fistula with sig-
of distal cervical internal carotid artery and old dissection nificant cortical venous reflux. Also note fusiform dilata-
sequelae are noted (Arrow in a, c). Note small aneurysm of tion of V4 segment of the left vertebral artery (arrow in d)

Fig. 7 Another patient

with EDS. Extensive
tortuosity and mild ectasias
of bilateral cervical
vertebral arteries are evident
150
reported. The other neurological features include
Chiari 1 malformation, craniosynostosis, and
hydrocephalus. The aneurysms are large and
symptomatic, reported to occur in 10 % of the
patients causing a mortality rate due to cerebral
hemorrhage of 7 % [51, 53]. Among the several
factors that are taken into consideration in the
management of these aneurysms, the presence of
arterial tortuosity cannot be overlooked as it sig-
nificantly hampers the distal access of interven-
tional devices. Both neurosurgical clipping and
endovascular coiling have been reported for the
management of cerebral aneurysms [51, 53, 54].
Pseudoxanthoma Elasticum (PXE)
PXE is an autosomal recessive disease due to the
mutation of ABCC6 gene on chromosome 16, the
defect of which results in the accumulation of
mineralized and fragmented elastic fibers in the
skin, eyes, and arterial wall. The intima of the
artery is thickened, calcified, and fragmented.
The typical manifestations of the disease include
skin laxity, angioid streaks in retina, and cardio-
vascular changes. The disease is rare and cerebral
complications are reported in 10 % of the PXE
patients. The neurological manifestations include
cerebral ischemia due to carotid or vertebral artery
occlusion, carotid artery calcifications, and cervi-
cal or intracranial aneurysms. Dissections and
association of rete mirabile of carotid artery are
also rarely reported [43, 55–57].
Marfan Syndrome
Marfan syndrome (MS) is a relatively common
connective tissue disease characterized by typical
cardiovascular, musculoskeletal, spinal menin-
geal, and ocular features. The disease is autosomal
dominant inherited; however, in one third of
patients, new mutations are observed. The defect
lies in the synthesis of fibrillin 1, an extracellular
matrix protein that forms one of the important
components of elastic microfibril. The microfi-
brils support elastic membrane and provide
S.K. Kannath and T.R. Kapilamoorthy
long-term structural stability, transfer and reorient
the stress load, stabilize the TGF beta binding
proteins, and regulate its signaling. Fibrillin 1
mutation is not identified in one third of the
patients with marfanoid features, and instead,
mutations of TGFBR1 and TGFBR2 may be
detected. These patients are grouped as MS type
2, and ocular features in this subset are uncommon
[58–60]. Revised Ghent criterion enables the
establishment of MS based on cardinal features,
family history, genetic screening tests, and
weighted systemic scores of ancillary clinical
findings [61].
Typical pathological features of large caliber
arteries include cystic medial necrosis, elastin
fragmentation, smooth muscle necrosis, and apo-
ptosis. Cystic medial necrosis is uncommon in
intracranial arteries. A few pathological studies
reported intimal hyperplasia, medial degenera-
tion, and internal elastic lamina fragmentation of
intracranial arteries – findings that may predispose
to the formation of aneurysms [62, 63]. The vas-
cular findings of carotid arteries include arterial
tortuosity (similar to LDS), intracranial and extra-
cranial aneurysms, and dissection. The isolated
involvement of the neck and intracranial arteries
is extremely uncommon, limited to small series or
case reports. Dissecting aneurysms secondary to
extension of aortic dissection may be more fre-
quent observation and may present with neurolog-
ical deficits. The saccular aneurysms are observed
in intracranial circulation, while fusiform type is
common in extracranial carotid arteries. The
patients may be asymptomatic or may present
with pulsatile swelling in the neck, TIA or strokes,
or rarely subarachnoid hemorrhage [62, 64–66].
Other Systemic Disease with Carotid
Involvement
Neurofibromatosis Type 1
Neurofibromatosis type 1 is a rare autosomal
dominant disease caused by the mutation of
tumor suppressor gene located in the chromosome
17 that encodes the protein neurofibromin.
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery
Neurofibromin regulates cellular mitogenesis by
inhibiting the reticular activating system (ras)
pathways, and hence a loss function leads to
overexpression of mitogenic signals within the
cell leading to cellular proliferation or differenti-
ation [67]. The clinical hallmarks of the disease
are café-au-lait spots, cutaneous neurofibromas,
bone changes, and iris hamartomas. The vascular
involvement is uncommon and is reported to
occur in 0.4–6.4 % of patients with NF1
[68]. NF1 vasculopathy affects large-, medium-,
and small-sized arteries. The arteries may be sec-
ondary involved when its wall is weakened by
extrinsic compression of extravascular neurofi-
bromas or due to the proliferation of Schwann
cell within the arterial layers. Recent studies dem-
onstrated a marked intimal and smooth muscle
cell proliferation, and this observation has been
related to decreased expression of neurofibromin
in the vascular wall. It is unsure whether the
vascular changes are congenital or acquired as a
pathological response to injury [68–71]. The path-
ophysiology of vascular abnormality is related to
the amount of intimal proliferation, spindle cell
proliferation in the media, and the arterial diame-
ter. Hence, small-sized arteries are often occluded,
whereas larger arteries show aneurismal changes
due to weakening of arterial wall [69–72].
The NF1 vasculopathy is the second leading
cause of death among the NF1 patients younger
than 40 years. The carotid, vertebral, or intracra-
nial arteries are affected in one fifth of the patients,
and vascular abnormality may be in the form of
arterial stenosis, aneurismal changes, or arteriove-
nous malformations. The aneurismal changes are
the commonest pathology and show female pre-
ponderance and predilection to rupture spontane-
ously during pregnancy. Extracranial aneurysms
are large and fusosaccular and may present as pul-
satile or enlarging neck mass or acute rupture. Intra-
cranial aneurysms are relatively uncommon and
may affect intradural ICA, middle cerebral, anterior
cerebral, or basilar arteries [68, 69, 71, 73, 74].
The aneurysms are incidentally detected or present
with subarachnoid hemorrhage. Rarely coexistent
intracranial and extracranial aneurysms also may be
found. The incidence of cerebral vasculopathy in
151
pediatric population was found to be 2.5 % in a large
institutional-based study. Besides moyamoya pat-
tern, isolated stenosis, occlusions, or ectasias of
intracranial arteries are observed (Figs. 8 and
9) [75].
Vertebral arteriovenous fistula is a well recog-
nized complication of neurofibromatosis with
fewer than 50 cases reported in the literature
(Fig. 10). The etiology of arteriovenous fistula is
not clearly understood. Spontaneous rupture
predisposed by arterial wall thinning and fragility
to adjacent venous plexus is thought to be one of
the mechanisms. A congenital theory assumes
preexisting abnormal congenital vascular commu-
nications within the arterial wall which develops
into the fistula. Clinical features of this condition
include pulsatile mass lesion, radiculo-
myelopathy, cranial neuropathy, or tinnitus. The
fistula may be found involving the upper or lower
cervical vertebral arteries or may involve one of
the branches of the vertebral artery. The epidural
venous sacs are often large and demonstrate rapid
shunting. Flow reversal and arterial steal may
often be identified, and sometimes, small commu-
nication channels open up when the main fistula is
occluded, and hence long-term surveillance is
advocated. Recent reports suggest good clinical
outcome with various endovascular strategies
[76–78].
PHACE(S) Syndrome
PHACE(S) is an acronym for a neurocutaneous
syndrome with salient features of Posterior fossa
malformation, Facial hemangiomas, Arterial
anomalies, Coarctation of aorta and Cardiac
defects, Eye anomalies, and ventral developmen-
tal anomalies such as Sternal clefts or
Supraumbilical raphe. The vascular lesions are
common in PHACES, seen in up to 78 % of
patients. The abnormalities are either congenital
or progressive and are related to the development
or phase of growth of cutaneous hemangioma.
The arterial anomalies are detected in the location
of hemangioma, and this association is hypothe-
sized due to vascular expression factors released
152
Fig. 8 A young patient with neurofibromatosis. MRI (a)
shows gliosis and atrophy of the right frontal operculum.
MR angiogram (b, c) shows occlusion of the right internal
carotid artery and small caliber of the right middle cerebral
artery. Diagnostic angiogram reveals occlusion of the right
by the hemangioma either during embryogenesis
or proliferative growth. The hemangiomas of face,
orbit, or scalp are associated with vascular
malformations of intracranial or carotid arteries,
whereas mandibular, neck, or chest lesions are
related to cardiac, aortic, or supraaortic abnormal-
ities [79, 80].
The congenital vascular anomalies observed
include arterial dysplasias, aberrant course or ori-
gin, hypoplasia, or agenesis of different intracra-
nial arteries, dolichoectasia, saccular intracranial
aneurysms, and presence of persistent embryonic
S.K. Kannath and T.R. Kapilamoorthy
supraclinoid internal carotid artery (d) and excessive tor-
tuosity of the left superior and inferior divisions of middle
cerebral artery (e). Cervical and intracranial vertebral arter-
ies are also tortuous (f)
carotid-vertebral communications [79–81]. Pro-
gressive vasculopathy such as stenosis,
moyamoya pattern, or occlusions is observed in
40 % of the patients. These vascular findings are
observed in more than 90 % of the patients
presenting with clinical features of acute stroke
[82]. Internal carotid arteries are commonly
affected followed by middle cerebral, anterior
cerebral, and vertebrobasilar arteries. Associated
brain structural anomalies are evident in the brain
in 41 % of the patients, more commonly in poste-
rior fossa [81].
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery 153

Fig. 9 Young girl with neurofibromatosis 1, presenting middle cerebral artery (a, c) with striate collaterals (b)
with acute stroke. Diagnostic angiogram of left internal suggestive of Moyamoya disease
carotid artery reveals occlusion of proximal segment of

Fig. 10 A middle aged

female with pulsatile mass
lesion in the neck. MRI (a)
show tortuous and dilated
left cervical vertebral artery
emptying into a large
pouch. This is confirmed in
the diagnostic angiogram
(b) She was diagnosed to
have neurofibromatosis 1

Tuberous Sclerosis mapped on chromosome 9 and chromosome

16, respectively. Besides typical hamartomatous
Tuberous sclerosis (TS) is an autosomal dominant lesions discovered in different organs, vascular
multisystem disorder caused by the mutation of involvement is extremely rare and is more com-
tumor suppressor genes, TSC1 and TSC2, monly observed in the aorta. To date, only three
154
extracranial carotid aneurysms and 19 intracranial
aneurysms are reported in the literature. Intracra-
nial aneurysms are more often reported with the
most commonly affected being the internal carotid
artery. The aneurysms are giant and fusiform and
may affect multiple arteries. Vascular wall disin-
tegration due to accumulation of mucopolysac-
charides is posited in its genesis [83–85].
Behcet Disease
Behcet disease (BD) is a multisystem inflamma-
tory disease of unknown etiology that presents
with typical clinical triad of uveitis and oral and
genital ulcers. Vascular involvement is observed
in 7–29 % of patients, and venous systems are
predominantly affected. Arterial involvement is
noted in 3–5 % of the patients, and aneurismal
rupture remains to be the most common cause of
mortality in patients with BD. Pathomechanism of
aneurysm formation is thought to be due to vasa
vasoritis and vasculitis of the arterial wall. Carotid
artery involvement is very rare and may affect the
common carotid or internal carotid artery. The
patient may present with enlarging pulsatile
lump in the neck or subarachnoid hemorrhage
[86–88]. Intradural aneurysms are rarer and
show no segmental predilection. Multiple aneu-
rysms may be found, and these patients show
tendency to develop aneurysms at different loca-
tions on follow-up [89, 90].
Carotid Aneurysms and Carotid
Blowout
Carotid Aneurysm
Aneurysms of extracranial carotid artery are
uncommonly observed and constitute about 4 %
of all peripheral aneurysms and less than 5 % of all
carotid surgeries [91]. Carotid aneurysms are cat-
egorized into true aneurysms or pseudoaneurysms
depending on whether the layers of arterial wall
are intact or not. Atherosclerosis remains the most
common cause of true aneurysm, seen in elderly
hypertensives, and presents with pulsatile neck
S.K. Kannath and T.R. Kapilamoorthy
mass or neurological symptoms such as TIA or
stroke. Indeed, luminal thrombus is noted in one
fifth of such lesions. Detailed discussion of this
entity is dealt with elsewhere. Other causes of true
aneurysms are rare and are described in the pre-
ceding sections. Pseudoaneurysms are relatively
uncommon and accounts for 14 % of all extracra-
nial carotid aneurysms. The most common cause
of such aneurysm includes blunt or penetrating
trauma and previous carotid surgeries. Direct pen-
etrating injury by sharp objects, shrapnels, or iat-
rogenic carotid puncture either intentionally or
while central venous access can injure the vessel
wall and produce localized or expanding hema-
toma and pseudoaneurysm. Blunt injuries damage
the arterial wall, when the shear or rotary forces
compress and stretch the arterial wall against the
cervical vertebra or skull base and promote aneu-
rysm formation. Also the dissecting process of
carotid artery splits the internal elastic lamina
and muscular media and weakens the wall which
may eventually turn aneurismal in one third of
dissection [92, 93]. The isolated aneurysms of
external carotid artery are very rare [94].
Untreated extracranial aneurysms carry poor
prognosis with high mortality [93, 95]. Besides
central neurological features, aneurysms may
bulge into throat or compress adjacent structures
such as the esophagus and lower cranial nerves.
The most feared complication is aneurysm expan-
sion and rupture resulting in a massive exsangui-
nation and fatal outcome. Initial diagnostic
evaluation is done with ultrasonography and
Doppler which confirm the clinical diagnosis,
delineate the aneurysm boundary and extent, and
demonstrate intraluminal thrombosis. Larger
aneurysms, aneurysms close to skull base and
presence of calcifications hinder detailed analysis.
However, it is a useful tool in following the
growth of the aneurysm in susceptible individuals
or in patients on conservative therapy. CT angiog-
raphy is currently the investigation of choice for
the evaluation of cervical aneurysms. CTA delin-
eates the extent and boundary of the aneurysm, its
relation to other vascular and airway structures,
presence of thrombosis or soft tissue hematoma,
and status of intracranial vasculature. Besides
diagnosis, it provides wealth of information
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery
Fig. 11 Carotid aneurysms in two patients (a, b, and c, respectively) with no identifiable etiology. Luminal irregularity
and thrombosis with calcifications are present (not shown)
influencing the choice of treatment as well as over
anticipated difficulties. Access vessel tortuosities,
presence of kinks, thrombosis, and disrupted cir-
cle of Willis are some of the factors considered
while planning the therapeutic approach [96, 97].
Rare involvement of external carotid artery or
concomitant intra-extracranial disease can also
be detected by CT angiography. Moreover,
craniocervical CTA as part of whole-body CT
protocol in trauma patients enables early recogni-
tion of vascular abnormalities and allows institu-
tion of appropriate therapy at the earliest [98, 99].
Disadvantages such as contrast overload and radi-
ation concerns are largely superseded by the over-
all benefits. Potential pitfalls include overlooking
subtle imaging features such as vessel wall irreg-
ularity, spasm, or small peripheral arterial aneu-
rysms especially when images are degraded by
swallowing artifacts.
MRI and MRA are complementary to CT in the
diagnosis of these aneurysms. MRI is useful in the
diagnosis and characterization of luminal thrombus
and delineation of extent of the aneurysm and
status of adjacent soft tissue structures. Besides
distinct advantage of absent ionizing radiation,
CT scores over MRI in the imaging workup of
carotid aneurysms due to its easy availability and
reduced scanning time except in situations of clin-
ical dilemma where CT is inconclusive and strong
suspicion of pseudoaneurysm exists. Identification
155
of characteristic signal abnormalities of hematoma
helps to clinch the diagnosis. Due to rarity of the
disease, the role of CT and MRI as a comprehen-
sive imaging study is not explored, and the current
practice is dependent on local institutional protocol
and availability of resources. Conventional angiog-
raphy is an invasive modality, usually performed as
a part of interventional procedure (Fig. 11) [100].
Carotid Blowout Syndrome
The carotid blowout syndrome (CBS) is a life-
threatening emergency caused by the rupture of
the carotid artery or its branches due to the weak-
ening of adventitial layer. The important factors
predisposing CBS include trauma, prior neck irra-
diation for head and neck malignancies, or radical
neck surgeries. Radiation increases the risk by
7.6-fold and can occur anytime between 2 and
20 years after the therapy. Radiation induces
inflammatory changes, fibrosis, and ischemia of
adventitia and results in the weakening of arterial
wall. The patient may be clinically silent or may
present with acute massive exsanguination. The
latter often carries high morbidity and mortality.
Despite successful treatment by endovascular par-
ent artery occlusion or stent graft placement,
recurrent CBS and procedural or postprocedural
complications are reported to occur in more than
156
one fourth of the patients, and overall prognosis is
poor due to natural progression of the disease or
treatment-related issues [101, 102].
Clinically, CBS is categorized into three
groups – acute, impending, and threatened.
Acute CBS presents with massive hemorrhage
that is not self-limiting and cannot be controlled
by packing or pressure. The artery ruptures freely
into the communicating wound or into the airway.
Without immediate resuscitation and treatment,
the patient succumbs to rapid exsanguination.
The impending CBS bleeds intermittently through
the oral cavity or surgical wound indicating a
weakened or ruptured artery with pseudo-
aneurysm formation. Without further treatment,
full-blown arterial rupture may ensue anytime.
Exposed artery as a result of wound infection,
flap necrosis, or tissue breakdown suggests inev-
itable rupture without any treatment, and it is
termed as threatened CBS [103].
Recent study has shown that CTA is very use-
ful in identifying the patients presenting with the
clinical features of impending CBS [104] The
CTA findings include extravasation of contrast,
pseudoaneurysms, recurrence of tumor, necrosis,
and exposed main or branches of carotid arteries.
The pseudoaneurysm and extravasation may be
observed in the common carotid artery, internal
carotid artery, or external carotid artery branches.
Petrous ICA is a common location of aneurysm in
patients who have received radiotherapy for naso-
pharyngeal carcinoma. The artery is defined as
exposed if necrosis contacts more than half of its
circumference. These findings showed good cor-
relation with angiographic features and have
value in prognosticating the outcome following
intervention. The patients with normal CTA find-
ings and impending CBS had better clinical out-
come as compared to the group of patients who
demonstrated these findings. Disadvantages of
CTA include its inapplicability in acute CBS,
decreased sensitivity in detecting small
pseudoaneurysms, or contrast extravasations
[101, 104, 105]. Imaging appearance of CBS in
MRI is also reported. Pseudoaneurysm was visu-
alized as enhancing lesion close to the carotid
artery that appeared isointense on T1-WI
sequences and hyperintense on T2-WI sequences
S.K. Kannath and T.R. Kapilamoorthy
[106]. In the appropriate clinical setting, this
imaging appearance should raise the suspicion
of CBS.
Angiogram is performed as a diagnostic pro-
cedure prior to intervention or when patients pre-
sent with acute CBS. Angiography may
demonstrate acute contrast extravasation,
pseudoaneurysm formation, focal luminal
narrowing, or luminal irregularity. Findings may
involve common carotid, internal carotid, or
external carotid artery branches and may involve
more than one site too. In threatened CBS, the
angiographic findings are often normal [101, 103,
105]. Diffuse blush in the soft tissue suggests
inflammatory or telangiectatic changes secondary
to radiation and merits treatment if patient is
symptomatic.
Miscellaneous
Carotidynia
The idiopathic carotidynia (CD) is characterized by
unilateral neck pain and tenderness over the carotid
artery bifurcation. Due to the diagnostic confusion
with other diseases that have similar clinical symp-
toms, carotidynia was removed by the International
Headache Society from the 2004 classification of
headache disorders [107, 108]. However, there are
many isolated reports that describe radiological and
structural changes of the carotid wall, thus empha-
sizing its existence as a unique entity possibly
induced by inflammation. The symptoms of
carotidynia include unilateral neck pain, radiating
to face and ear, aggravated by lateral head move-
ment, swallowing, or chewing. The disease is self-
limited and may be preceded by flu-like symptoms.
CD commonly affects young and middle-aged
individuals and women. Imaging shows the
absence of structural vascular and nonvascular
causes, though many reports suggest specific imag-
ing appearances in carotidynia [109–115]. Acute
phase reactants such as C-reactive protein (CRP)
and serum amyloid antigen (SAA) are reported to
correlate with disease activity [115, 116]. A gamut
of differential diagnosis need to be ruled out before
confirming the diagnosis of CD.
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery
Pathological study was reported only in one
report where a patient with classical carotidynia
was found to have a thickened carotid adventitia
intraoperatively and histological analysis
revealed adventitial infiltration of lymphocytes,
vascular and fibroblasts proliferation, and early
fibrosis – picture suggesting chronic inflamma-
tory process [112]. Noninvasive imaging modal-
ities detect structural abnormalities in the
periarterial carotid space, though these changes
are not correlated with histological analysis.
Ultrasonography and Doppler studies showed
eccentric hypoechoic wall thickening of carotid
bulb and distal common carotid artery which may
or may not be associated with mild non-flow-
limiting luminal narrowing. The hypoechoic
soft tissue may extend outwardly and corroborate
with the location of tenderness or nodule
[109–111]. CT is not sensitive to detect subtle
soft tissue changes due to lower resolution. PET
CT has shown increased [18 F] fluorodeox-
yglucose (18 F-FDG) uptake in the region of
enhancing the soft tissue around the carotid
[117]. MRI study may reveal intermediate signal
intensity tissue around the distal common carotid
and carotid bulb limited to carotid sheath on
T1-WI sequence. These abnormal tissues show
a marked homogenous enhancement, more con-
spicuous on fat-saturated T1-WI sequence. The
luminal caliber of the artery is almost always
normal. Clearly, MRI delineates the abnormal
tissue as well as the extent of involvement and
also helps to rule out other causes that may mimic
CD [109, 111, 113–115].
Eagle Syndrome or Stylohyoid
Syndrome
The long styloid process defined as more than
3 cm in length is seen in 4 % of population,
though the typical symptoms attributed to it are
seen in just about 4 % of these patients. The
styloid process is a long slender bone attached
to the base of the temporal bone and is related
closely to many important neurovascular struc-
tures. The styloid process, together with
stylohyoid ligament and lesser cornua of hyoid
157
bone, forms stylohyoid complex (SHC), and a
reduced distensibility of this complex with sub-
sequent irritation of neurovascular structures is
implicated in the pathogenesis of Eagle syn-
drome (ES) [118]. The embryological develop-
ment of SHC is controversial and is thought to
differentiate from Reichert cartilage which is the
cartilage of the second branchial arch. One the-
ory focuses on the disintegration of this cartilage
into five components such as tympanohyal,
stylohyal, ceratohyal, hypohyal, and basohyal.
Tympanohyal and stylohyal contribute to the
base and styloid process, respectively; hypohyal
and basohyal form the hyoid bone; and
ceratohyal in humans is thought to degenerate
and forms the stylohyoid ligament. The second
theory assumes the division of the cartilage: the
proximal and larger part contributes to the for-
mation of styloid process, distal segment into
hyoid bone, and the intervening mesenchyme
develops into stylohyoid ligament [119–122].
The styloid process is related to important
neuro and vascular structures, and thus a deviant
and angled styloid process may impinge upon
these structures and cause clinical symptoms.
The styloid process is related to the 9th to 12th
cranial nerves, cervical sympathetic chain,
branches of external carotid artery, internal carotid
artery, and internal jugular vein. The syndrome is
commonly observed in women in the fourth or
fifth decades. Typical symptoms include unilat-
eral sharp or dull aching pain in the lateral neck,
angle of mandible, oral cavity, or base of the
tongue which are precipitated or aggravated by
neck movement, swallowing, or yawning. Com-
pression of carotid artery may cause dull aching
pain in the parietal region, TIA, stroke, and
Horner syndrome exacerbated by rotary move-
ments of head to contralateral side or it may
even cause sudden death [119, 120, 123]. Direct
mechanical impingement on the cervical carotid
artery also has been implicated as an important
risk factor for carotid artery dissection or aneu-
rysm [124, 125].
A gamut of possibilities needs to be excluded
before one considers the diagnosis of ES. Oral
examination of tonsillar fossa may reveal unduly
elongated styloid process. Palpation of tonsillar
158
fossa may provoke the pain which is abolished by
instillation of local anesthetic along the styloid
process. Radiographic studies such as submento-
vertical view or orthopantomogram are useful to
reveal enlarged styloid process. The styloid pro-
cess is considered enlarged if its length exceeds
one third of the length of the ramus of the mandi-
ble or if the length is >3 cm. CT and CTA studies
can accurately depict the length, orientation, and
relation of styloid process to major vascular struc-
tures [126]. Though invasive, 3D rotational angi-
ography can directly visualize the mechanical
impression on carotid or other vessels by elon-
gated styloid process, it is a useful tool to confirm
and localize disease in doubtful and confusing
situations [127].
Summary
The clinical disease entities encompassing
nonatherosclerotic and nondissection diseases of
carotid artery are exhaustive and have often
nonspecific clinical and imaging findings. The
specific clinical inputs and typical imaging fea-
tures are important clues to suspect these vascular
diseases and guide in further evaluation and man-
agement. Genetic and molecular analysis is
confirmative in hereditary and connective tissue
disorders.
References
1. Touze E, Oppenheim C, Tyrstram D et al (2010)
Fibromuscular dysplasia of cervical and intracranial
arteries. Int J Stroke 5:296–305
2. Plouin PF, Perdu J, Batide-Alanore AL et al (2007)
Fibromuscular dysplasia. Orphanet J Rare Dis 2:28
3. Schievink WI, Bjornsson J (1996) Fibromuscular
dysplasia of the internal carotid artery: a clinicopath-
ological study. Clin Neuropathol 15:2–6
4. Mettinger KL (1982) Fibromuscular dysplasia and the
brain. II. Current concept of the disease. Stroke
13:53–58
5. Schievink WI (2001) Spontaneous dissection of the
carotid and vertebral arteries. N Engl J Med
344:898–906
6. Olin JW, Froehlich J, Gu X, Bacharach JM, Eagle K,
Gray BH, Jaff MR, Kim ESH, Mace P, Matsumoto
AH, McBane RD, Kline-Rogers E, White CJ,
S.K. Kannath and T.R. Kapilamoorthy
Gornik L (2012) The United States registry
for fibromuscular dysplasia: results in the first
447 patients. Circulation 125:3182–3190
7. Wells RP, Smith RR (1982) Fibromuscular dysplasia
of the internal carotid artery: a long term follow-up.
Neurosurgery 10:39–43
8. Mettinger KL, Ericson K (1982) Fibromuscular dys-
plasia and the brain observations on angiographic,
clinical and genetic characteristics. Stroke 13:46–52
9. De Souza AWS, de Carvalho JF (2014) Diagnostic
and classification criteria of Takayasu arteritis.
J Autoimmun 48–49:79–83
10. Ishikawa K (1978) Natural history and classification
of occlusive thromboaortopathy. (Takayasu’s dis-
ease). Circulation 57:27–35
11. Kim JH, Suh DC, Kim JK, Kim SJ et al (2005) Cor-
relation of neurological manifestations of Takayasu’s
arteritis with cerebral angiographic findings. J Clin
Imaging 29:79–85
12. Ringleb PA, Strittmatter EI, Loewer M, Hartmann M
et al (2005) Cerebrovascular manifestations of
Takayasu arteritis in Europe. Rheumatology
44:1012–1015
13. Kim DS, Kim JK, Yoo DS, Huh PW, Cho KS, Kang
JK (2002) Takayasu arteritis presenting with sub-
arachnoid hemorrhage. Report of two cases. J Korean
Med Sci 17:695–698
14. Terao C, Yoshifuji H, Mimori T (2014) Recent
advances in Takayasu arteritis. Int J Rheum Dis
17:238–247
15. Hotchi M (1992) Pathological studies on Takayasu’s
arteritis. Heart Vessels Suppl 7:11–17
16. Hartlage GR, Palios J, Barron BJ, Stillman AE,
Bossone E et al (2014) Multimodality imaging of
aortitis. JACC Cardiovasc Imaging 7:605–619
17. Ohigashi H, Haraguchi G, Konishi M, Tezuka D,
Kamiishi T, Ishihara T et al (2012) Improved progno-
sis of Takayasu arteritis over the past decade – com-
prehensive analysis of 106 patients. Circ J
76:1004e11
18. Alibaz-Oner F, Aydin SZ, Direskeneli H (2013)
Advances in the diagnosis, assessment and outcome
of Takayasu’s arteritis. Clin Rheumatol 32:541–546
19. Caballero PE (2011) Common carotid artery aneu-
rysm revealing Takayasu’s arteritis. J Stroke
Cerebrovasc Dis 20(6):556–558
20. Santhosh K, Jaydadevan ER, Kapilamoorthy TR,
Sylaja PN (2013) Temporal evolution of
hypertrophied vasa vasorum of common carotid
artery triggered by cerebral ischemia: a serial angio-
graphic investigation. Neurol India 61(3):312–314
21. Shuaib UA, Kate M, Homik J, Jerrakathil T (2013)
Recurrent non-aneurysmal subarachnoid
haemorrhage in Takayasu arteritis: is the cause immu-
nological or mechanical? BMJ Case Rep
22. Borchers AT, Gershwin MG (2012) Giant cell arteri-
tis: a review of classification, pathophysiology,
geoepidemiology and treatment. Autoimmun Rev
11:A544–A554
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery
23. Smeeth L, Cook C, Hall AJ (2006) Incidence of
diagnosed polymyalgia rheumatica and temporal
arteritis in the United Kingdom, 1990–2001. Ann
Rheum Dis 65:1093–1098
24. Hayreh SS, Podhajsky PA, Zimmerman B (1998)
Ocular manifestations of giant cell arteritis. Am J
Ophthalmol 125:509–520
25. Gonzalez-Gay MA, Vazquez-Rodriguez TR, Gomez-
Acebo I, Pego-Reigosa R, Lopez-Diaz MJ, Vazquez-
Triñanes MC et al (2009) Strokes at time of disease
diagnosis in a series of 287 patients with biopsy-
proven giant cell arteritis. Medicine (Baltimore)
88:227–235
26. Nesher G (2014) The diagnosis and classification of
giant cell arteritis. J Autoimmun 48–49:73–75
27. Arida A, Kryprianou M, Kanakis M, Sfikakis PP
(2010) The diagnostic value of ultrasonography-
derived edema of the temporal artery wall in giant
cell arteritis: a second meta-analysis. BMC
Musculoskelet Disord 11:44
28. Schinkel AF, van den Oord SC, van der Steen AF, van
Laar JA, Sijbrands EJ (2014) Utility of contrast-
enhanced ultrasound for the assessment of the carotid
artery wall in patients with Takayasu or giant cell
arteritis. Eur Heart J Cardiovasc Imaging 15:541–546
29. Bley TA, Reinhard M, Hauenstein C, Markl M,
Warnatz K, Hetzel A, Uhl M, Vaith P, Langer M
(2008) Comparison of duplex sonography and high-
resolution magnetic resonance imaging in the diagno-
sis of giant cell (temporal) arteritis. Arthritis Rheum
52:2470–2477
30. Blockmans D, Coudyzer W, Vanderschueren S,
Stroobants S, Loeckx D, Heye S, De Ceuninck L,
Marchal G, Bobbaers H (2008) Relationship between
fluorodeoxyglucose uptake in the large vessels and
late aortic diameter in giant cell arteritis. Rheumatol-
ogy 47:1179–1184
31. Robbs JV, Paruk N (2010) Management of HIV
vasculopathy – a South African experience. Eur J
Vasc Endovasc Surg 39:S25–S31
32. Padayachy V, Robbs JV (2012) Carotid artery aneu-
rysms in patients with human immunodeficiency
virus. J Vasc Surg 55:331–337
33. Hamilton DK, Kassel NF, Jensen ME, Dumont AS
(2007) Subarachnoid hemorrhage and diffuse
vasculopathy in an adult infected with HIV.
J Neurosurg 106:478–480
34. Tipping B, de Villiers L, Wainwright H, Candy S,
Bryer A (2007) Stroke in patients with human immu-
nodeficiency virus infection. J Neurol Neurosurg Psy-
chiatry 78:1320–1324
35. Maniker AH, Hunt D (1996) Cerebral aneurysm in the
HIV patient: a report of six cases. Surg Neurol
46:49–54
36. Chetty R, Batitang S, Nair R (2000) Large artery
vasculopathy in HIV-positive patients: another
vasculitic enigma. Hum Pathol 31:374–379
37. Woolgar JD, Ray R, Maharaj K, Robbs JV (2002)
Colour Doppler and grey scale ultrasound features of
159
HIV-related vascular aneurysms. Br J Radiol
75:884–888
38. Mazzoni P, Chiriboga CA, Millar WS, Rogers A
(2000) Immunodeficiency virus infection: case report
and literature review. Pediatr Neurol 23:252–255
39. Ritelli M, Drera B, Vicchio M, Puppini G, Biban P,
Pilati M, Prioli MA, Barlat S, Colombi M (2009)
Arterial tortuosity syndrome in two Italian paediatric
patients. Orphanet J Rare Dis 4:20
40. Callewaert BL, Willaert A, Kerstjens-Frederikse WS,
De Backer J, Devriendt K, Albrecht B et al (2008)
Arterial tortuosity syndrome: clinical and molecular
findings in 12 newly identified families. Hum Mutat
29:150–158
41. Nauheim MR, Walcott BP, Nahed BV, Macrae CA,
Levinson JR, Oglivy CS (2011) Arterial tortuosity
syndrome with multiple intracranial aneurysms.
Arch Neurol 68(3):369–371
42. Germain DP (2007) Ehlers- Danlos syndrome type
4. Orphanet J Rare Dis 2:32
43. Gdynia HJ, Kuhnlein P, Ludolph AC, Huber R (2008)
Connective tissue disorders in dissections of the
carotid or vertebral arteries. J Clin Neurosci
15:489–494
44. Boutouyrie P, Germanin DP, Fiessinger JN, Laloux P,
Perdu J, Laurent S (2004) Increased carotid wall stress
in vascular Ehlers-Danlos syndrome. Circulation
109:1530–1535
45. Pepin M, Schwarze U, Superti-Furga A, Byers PH
(2000) Clinical and genetic features of Ehlers–Danlos
syndrome type IV, the vascular type. N Engl J Med
342:673–680
46. Oderich GS, Panneton JM, Bower TC, Lindor NM,
Cherry KJ, Noel AA, Kalra M, Sullivan T, Gloviczi P
(2005) The spectrum, management and clinical out-
come of Ehlers-Danlos syndrome type IV: a 30-year
experience. J Vasc Surg 42:98–106
47. Schievink WI (2004) Cerebrovascular involvement in
Ehlers-Danlos Syndrome. Curr Treat Options
Cardiovasc Med 6:231–236
48. Dohle C, Baehring JM (2012) Multiple strokes and
bilateral carotid dissections: a fulminant case of newly
diagnosed Ehlers–Danlos syndrome type IV. J Neurol
Sci 318:168–170
49. Horowitz MB, Purdy PD, Valentine RJ et al (2000)
Remote vascular catastrophes after neurovascular
interventional therapy for type 4 Ehlers-Danlos syn-
drome. AJNR Am J Neuroradiol 21:974–976
50. Usinskiene J, Mazighi M, Bisdorff A et al (2006)
Fatal peritoneal bleeding following embolization of
a carotid-cavernous fistula in Ehlers-Danlos syn-
drome type IV. Cardiovasc Intervent Radiol
29:1104–1106
51. Loeys BL, Schwarze U, Holm T et al (2006) Aneu-
rysm syndromes caused by mutations in the TGF-beta
receptor. N Engl J Med 355(8):788–798
52. Maccarrick G, Black JH III, Bowdin S, Hamamsy IE,
Frischmeier-Guerrerio PA, Guerrerio AL, Sponseller
PD, Loeys B, Dietz HC (2014) Loeys–Dietz
160
syndrome: a primer for diagnosis and management.
Genet Med 16:576–587
53. Rahme RJ, Adel JG, Bendok BR, Bebawy JF, Gupta
DK, Batjer HH (2011) Association of intracranial
aneurysm and Loeys-Dietz Syndrome: case illustra-
tion, management, and literature review. Neurosur-
gery 69:E488–E493
54. Levitt MR, Morton RP, Mai JC, Ghodke B, Hallam
DK (2012) Endovascular treatment of intracranial
aneurysms in Loeys-Dietz syndrome. J Neurointerv
Surg 4:e37
55. van den Berg JSP, Hennekam RCM, Cruysberg JRM,
Steijlen PM, Swart J, Tijmes N et al (2000) Prevalence
of symptomatic intracranial aneurysm and ischaemic
stroke in pseudoxanthoma elasticum. Cerebrovasc
Dis 10:315–319
56. Schievink WI, Michels VV, Piepgras DG (1994)
Neurovascular manifestations of heritable connective
disorders: a review. Stroke 25:889–903
57. Vasseur M, Carsin-Nichol B, Ebran JM, Willoteaux S,
Martin L, Leftheriotis G (2011) Carotid rete mirabile
and pseudoxanthoma elasticum: an accidental associ-
ation? Eur J Vasc Endovasc Surg 42:292–294
58. Dean JCS (2007) Marfan syndrome: clinical diagno-
sis and management. Eur J Hum Genet 15:724–733
59. Eberth JF, Taucer AI, Wilson E, Humphrey JD (2009)
Mechanics of carotid arteries in a mouse model of
Marfan syndrome. Ann Biomed Eng 37:1093–1104
60. Ferruzzi J, Collins MJ, Yeh AT, Humphrey JD (2011)
Mechanical assessment of elastin integrity in fibrillin-
1-deficient carotid arteries: implications for Marfan
syndrome. Cardiovasc Res 92:287–295
61. Loeys BL, Dietz HC, Braverman AC et al (2010)
Towards a revised Ghent nosology for the Marfan
syndrome. J Med Genet 47:476–485
62. Schievink WI, Parisi JE, Piepgras DG (1997) Intra-
cranial aneurysms in Marfan’s syndrome: an autopsy
study. Neurosurgery 41:866–871
63. Yuan SM, Jing H (2011) Cystic medial necrosis:
pathological findings and clinical implications. Rev
Bras Cir Cardiovasc 26:107–115
64. Bouziane Z, Bensaid B, Idrissi RE, Mesnaoui AE,
Bensaid Y, Ratbi I (2013) Marfan syndrome and cer-
vical internal carotid artery aneurysm. Vasa
42:457–460
65. Sztajzel R, Hefft S, Girardet C (2001) Marfan’s syn-
drome and multiple extracranial aneurysms.
Cerebrovasc Dis 11:346–349
66. Matsuda M, Matsuda I, Handa H, Okamoto K (1979)
Intracavernous giant aneurysm associated with
Marfan’s syndrome. Surg Neurol 12:119–121
67. Riccardi VM (2000) The vasculopathy of NF1 and
histogenesis control genes. Clin Genet 58:345–347
68. Oderich GS, Sullivan TM, Bower TC, Glovickzi P,
Miller DV, Babovic-Vuksanovic D, Macedo TA,
Stanson A (2007) Vascular abnormalities in patients
with neurofibromatosis syndrome type I: clinical
spectrum, management, and results. J Vasc Surg
46:475–484
S.K. Kannath and T.R. Kapilamoorthy
69. Okendi E, Moghaddam MB, Oderich GS (2010)
Internal carotid artery aneurysms in a patient with
neurofibromatosis type 1. Vasc Endovascular Surg
44:511–514
70. Xu J, Ismat FA, Wang T et al (2007) NF1 regulates a
Ras-dependent vascular smooth muscle proliferative
injury response. Circulation 116:2148–2156
71. Ku YK, Chen HW, Chen HW, Fu CJ, Chin SC, Liu
YC (2008) Giant extracranial aneurysms of both
internal carotid arteries with aberrant jugular veins
in a patient with neurofibromatosis type 1. Am J
Neuroradiol AJNR 29:1750–1752
72. Reubi F (1945) Neurofibromatose et le´sions
vasculaires. Schweiz Med Wochenschr 75:463–465
73. Schievink WI, Riedinger M, Maya MM (2005) Fre-
quency of incidental intracranial aneurysms in neuro-
fibromatosis type 1. Am J Med Genet A 134:45–48
74. You MW, Kim EJ, Choi WS (2011) Intracranial and
extracranial fusiform aneurysms in a patient with
neurofibromatosis type 1: a case report.
Neurointervention 6:34–37
75. Rosser TL, Vezina G, Packer RJ (2005) Cerebrovas-
cular abnormalities in a population of children with
neurofibromatosis type 1. Neurology 64:553–555
76. Patro SN, Gupta AK, Aravinda HR, Jolapara MB,
Saini J (2009) Combined transarterial and percutane-
ous coiling of a spontaneous vertebrovertebral fistula
associated with neurofibromatosis Type 1.
J Neurosurg 111:37–40
77. Deans WR, Bloch S, Leibrock L, Berman BM,
Skultety FM (1982) Arteriovenous fistula in patients
with neurofibromatosis. Radiology 144:103–107
78. Gao P, Chen Y, Zhang H, Ling F (2013) Vertebral
arteriovenous fistulae (AVF) in Neurofibromatosis
type 1: a report of two cases. Turk Neurosurg
23:289–293
79. Baccin CE, Krings T, Alvarez H, Ozanne A,
Lasjaunias PL (2007) A report of two cases with
dolichosegmental intracranial arteries as a new fea-
ture of PHACES syndrome. Childs Nerv Syst
23:559–567
80. Heyer GL, Dowling MM, Licht DJ, Tay SKH,
Morel K, Garzon MC, Meyers P (2008) The cerebral
vasculopathy of PHACES syndrome. Stroke
39:308–316
81. Hess CP, Fullerton HJ, Metry DW, Drolet BA, Siegel
DH, Auguste KI, Gupta N, Haggstrom AN, Dowd CF,
Frieden IJ, Barkovich AJ (2010) Cervical and intra-
cranial arterial anomalies in 70 patients with PHACE
syndrome. AJNR Am J Neuroradiol 31:1980–1986
82. Siegel DH, Tefft KA, Kelly T, Johnson C, Mery D,
Burrows P, Pope E, Cordisco M, Holland KE,
Maheswari M, Keith P, Garzon M, Hess C, Frieden
IJ, Fullerton HJ, Drolet BA (2012) Stroke in children
with posterior fossa brain malformations, hemangi-
omas, arterial anomalies, coarctation of the aorta and
cardiac defects, and eye abnormalities (PHACE) syn-
drome: a systematic review of the literature. Stroke
43:1672–1674
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery
83. Kirkwood ML, Chung J, Timaran CH, Valentine RJ
(2013) Extracranial carotid artery aneurysms in two of
three monozygotic triplets with tuberous sclerosis
complex. J Vasc Surg 57:1120–1122
84. Jurkiewicz E, Jozwiak S (2006) Giant intracranial
aneurysm in a 9-year-old boy with tuberous sclerosis.
Pediatr Radiol 36:463
85. Hung PO, Wang HS, Chou ML, Wong AMC (2008)
Tuberous sclerosis complex with multiple intracranial
aneurysms in an infant. Pediatr Neurol 39:365–367
86. Sayed A, Elwan H, Fouad F, Taha A, Elhindawi K,
Khairy H, Gad A, Eldin HK (2010) Behcet extracra-
nial carotid aneurysms: is there still a role for ligation.
Eur J Vasc Endovasc Surg 39:17–22
87. Albeyoglu S, Cinar B, Eren T, Filizcan U,
Bayserke O, Aslan C (2010) Extracranial carotid
artery aneurysm due to Behcet’s disease. Asian
Cardiovasc Thorac Ann 18:574–576
88. Bouarhroum A, Sedki N, Bouziane Z, Mahi OE,
Idrissi RE, Lahlou Z, Lekehel B, Sefiani Y, Mesnaoui
AE, Benjelloun A, Ammar F, Bensaïd Y (2006)
Extracranial carotid aneurysm in Behçet disease:
report of two new cases. J Vasc Surg 43:627–630
89. Ozveren MF, Matsumoto Y, Kondo R, Takahashi A
(2009) Coil embolization of an unruptured intracra-
nial aneurysm associated with Behcet’s disease.
Neurol Med Chir (Tokyo) 49:471–473
90. Kaku M, Hamada JI, Kuroda JI, Kai Y, Morioka M,
Kuratsu JI (2007) Multiple peripheral middle cerebral
aneurysms associated with Behcet’s disease. Acta
Neurochir (Wien) 149:823–827
91. Garg K, Rockman CB, Lee V, Maldonaldo TS,
Jacobowitz GR, Adelman MA, Mussa FF (2012)
Presentation and management of carotid artery aneu-
rysms and pseudoaneurysms. J Vasc Surg
55:1618–1622
92. Longo GM, Kibbe MR (2005) Aneurysms of the
carotid artery. Semin Vasc Surg 18:178–183
93. Attigah N, Ku¨lkens S, Zausig N, Hansmann J,
Ringleb P, Hakimi M, Eckstein HH, Allenberg JR,
Bo¨ckler D (2009) Surgical therapy of extracranial
carotid artery aneurysms: long-term results over a
24-year period. Eur J Vasc Endovasc Surg
37:127–133
94. Nadig S, Barnwell S, Wax MK (2009)
Pseudoaneurysm of the external carotid artery –
review of literature. Head Neck 31:136–139
95. Cox MW, Whittaker DR, Martinez C, Fox CJ,
Feuerstein IM, Gillespie DL (2007) Traumatic
pseudoaneurysms of the head and neck: early
endovascular intervention. J Vasc Surg 46:1227–1233
96. Maras D, Lioupis C, Magoufis G, Tsamopoulos N,
Moulakakis K, Andrikopoulos V (2006) Covered
stent-graft treatment of traumatic internal carotid
artery pseudoaneurysms: a review. Cardiovasc
Intervent Radiol 29:958–968
97. Welleweerd JC, Moll FL, de Borst GJ (2012) Techni-
cal options for the treatment of extracranial carotid
aneurysms. Expert Rev Cardiovasc Ther 10:925–931
161
98. Langner S, Fleck S, Kirsch M, Petrik M, Hosten N
(2008) Whole-body ct trauma imaging with adapted
and optimized CT angiography of the craniocervical
vessels: do we need an extra screening examination?
AJNR Am J Neuroradiol 29:1902–1907
99. Fleck SK, Langner S, Baldauf J, Kirsch M,
Rosenstengel C, Schroeder HW (2010) Blunt
craniocervical artery injury in cervical spine lesions:
the value of CT angiography. Acta Neurochir
152:1679–1686
100. Ouffiah C, Hall E (2012) Imaging assessment of pen-
etrating injury of the neck and face. Insights Imaging
3:419–431
101. Chang FC, Lirng JF, Luo CB, Wang SJ, Wu HM, Guo
WY, Teng MMH, Chnag CY (2008) Patients with
head and neck cancers and associated postirradiated
carotid blowout syndrome: endovascular therapeutic
methods and outcomes. J Vasc Surg 47:936–945
102. Chang FC, Luo CB, Lirng JF, Lin CJ, Wu HM, Hung
SC, Guo WY, Teng MMH, Chang CY (2013) Evalu-
ation of the outcomes of endovascular management
for patients with head and neck cancers and associated
carotid blowout syndrome of the external carotid
artery. Clin Radiol 68:e561–e569
103. Chaloupka JC, Putman CM, Citardi MJ et al (1996)
Endovascular therapy for the carotid blowout syn-
drome in head and neck surgical patients: diagnostic
and managerial considerations. AJNR Am J
Neuroradiol 17:843–852
104. Lee CW, Yang CY, Chen YF, Huang A, Wang YH,
Liu HM (2014) CT angiography findings in carotid
blowout syndrome and its role as a predictor of 1-year
survival. AJNR Am J Neuroradiol 35:562–567
105. Luo CB, Teng MMH, Chang FC, Chang CY, Guo
WY (2008) Radiation carotid blowout syndrome in
nasopharyngeal carcinoma: angiographic features and
endovascular management. Otolaryngol Head Neck
Surg 138:86–91
106. Yanik B, Keyik B, Conkbayir I, Teber MA (2011)
Carotid blowout syndrome with oronasal hemor-
rhage: magnetic resonance imaging findings. Jpn J
Radiol 29:72–75
107. Headache Classification Committee of the Interna-
tional Headache Society (1988) Classification and
diagnostic criteria for headache disorders, cranial neu-
ralgias and facial pain. Cephalalgia 8(Suppl 7):1–96
108. Headache Classification Subcommittee of the Inter-
national Headache Society (2004) The International
Classification of Headache Disorders: 2nd edn.
Cephalalgia 24(Suppl 1):9–160
109. Kosaka N, Sagoh T, Uematsu H, Kimura H,
Miyayama S, Noguchi M, Itoh H (2007) Imaging by
multiple modalities of patients with a carotidynia
syndrome. Eur Radiol 17:2430–2433
110. Barker R, Massouh H (2009) Carotidynia – a valid
cause of unilateral neck pain. Eur J Radiol Extra 71:
e97–e99
111. Schaumberg J, Eckert B, Michels P (2011)
Carotidynia. Magnetic resonance imaging and
162
ultrasonographic imaging of a self-limiting disease.
Clin Neuroradiol 21:91–94
112. Upton PD, Smith JRG, Charnock DR (2003) Histo-
logic confirmation of carotidynia. Otolaryngol Head
Neck Surg 129:443–444
113. Burton BS, Syms MJ, Petermann GW, Burgess LPA
(2000) MR imaging of patients with carotidynia.
AJNR Am J Neuroradiol 21:766–769
114. Woo JKH, Jhamb A, Heran MKS, Hurley M, Graeb D
(2008) Resolution of existing intimal plaque in a patient
with carotidynia. AJNR Am J Neuroradiol 29:732–733
115. Commacho F, Bottin R, Brescia G, Tsilikas K, Volo T,
Tregnaghi A, Martini A (2012) Carotidynia: new
aspects of a controversial entity. Acta otorhino-
laryngol Ital 32:266–269
116. Taniguchi Y, Horino T, Hashimoto K (2008) Is
carotidynia syndrome a subset of vasculitis?
J Rheumatol 35:1901–1902
117. Amaravadi RR, Behr SC, Kousoubris PD, Raja S
(2008) [18F] Fluorodeoxyglucose positron-emission
tomography-CT imaging of carotidynia. AJNR Am J
Neuroradiol 29:1197–1199
118. Colby CC, Del Gaudio JM (2011) Stylohyoid com-
plex syndrome-a new diagnostic classification. Arch
Otolaryngol Head Neck Surg 137:248–252
119. Fusco DJ, Asteraki S, Spetzler RF (2012) Eagle’s
syndrome: embryology, anatomy, and clinical man-
agement. Acta Neurochir 154:1119–1126
S.K. Kannath and T.R. Kapilamoorthy
120. Piagkou M, Anagnostopoulou S, Kouladouros K,
Piagkos G (2009) Eagle’s syndrome: a review of the
literature. Clin Anat 22:545–558
121. Rodriguez-Vazquez JF, Merida-Velasco JR, Verdugo-
Lopez S, Sanchez-Montesinos I, Merida-Velasco JA
(2006) Morphogenesis of the second pharyngeal arch
cartilage (Reichert’s cartilage) in human embryos.
J Anat 208:179–189
122. Lorman JG, Biggs JR (1983) The Eagle syndrome.
AJR Am J Roentgenol 140:881–882
123. Ghosh LM, Dubey SP (1999) The syndrome of elon-
gated styloid process. Auris Nasus Larynx
26:169–175
124. Dao A, Karnezis S, Lane JS III, Fujitani RM, Saremi
F (2011) Eagle syndrome presenting with external
carotid artery pseudoaneurysm. Emerg Radiol
18:263–265
125. Muthusami P, Kesavadas C, Sylaja PN, Thomas B,
Harsha KJ, Kapilamoorthy TR (2013) Implicating the
long styloid process in cervical carotid artery dissec-
tion. Neuroradiology 55(7):861–867
126. Murtagh RD, Caracciolo RD, Fernandez G (2001) CT
findings associated with Eagle syndrome. AJNR Am J
Neuroradiol 22:1401–1402
127. Nakagawa D, Ota T, Iijima A, Saito N (2011) Diag-
nosis of eagle syndrome with 3-dimensional angiog-
raphy and near-infrared spectroscopy: case report.
Neurosurgery 68:E847–E849
 Imaging of the Pathology of the
Vertebral Arteries 10
David Chiao and Max Wintermark

Contents Subclavian Steal Syndrome . . . . . . . . . . . . . . . . . . . . . . . . 180

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
Embryology and Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . 164
Embryology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
Vascular Territories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
Atherosclerotic Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
Epidemiology and Clinical Findings . . . . . . . . . . . . . . . . 172
Pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
Imaging Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
Work-Up and Management . . . . . . . . . . . . . . . . . . . . . . . . . . 174
Vertebral Artery Dissection . . . . . . . . . . . . . . . . . . . . . . . . 175
Epidemiology and Clinical Findings . . . . . . . . . . . . . . . . 175
Pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Imaging Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Work-Up and Management . . . . . . . . . . . . . . . . . . . . . . . . . . 178
Epidemiology and Clinical Findings . . . . . . . . . . . . . . . . 180
Pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Imaging Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Work-Up and Management . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Bow Hunter’s Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Epidemiology and Clinical Findings . . . . . . . . . . . . . . . . 184
Pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
Imaging Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Work-Up and Management . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Miscellaneous Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Fibromuscular Dysplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Vasculitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188

D. Chiao • M. Wintermark (*)

Department of Radiology, School of Medicine, University
of Virginia, Charlottesville, VA, USA
e-mail: dc.harmonic@gmail.com;
max.wintermark@gmail.com

# Springer Science+Business Media New York 2016 163

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_17
164 D. Chiao and M. Wintermark

encountered vertebral artery diseases, vertebral

Abstract
artery atherosclerosis and vertebral artery dissec-
The vertebral arteries are among the most clin-
tion. This is then followed by discussion of two
ically important arteries in the body, providing
functional disorders, subclavian steal syndrome
blood flow to the posterior circulation of the
and bow hunter’s syndrome. Finally, a brief over-
brain. Both radiologists and clinicians need to
view of miscellaneous conditions, such as
understand the basic anatomy and the pathol-
fibromuscular dysplasia and vasculitis, is covered
ogy that can affect the vertebral arteries in the
in the last section.
acute and non-acute settings. Radiologists, in
particular, need to be cognizant of the imaging
findings of vertebral artery pathology as
misdiagnosis of certain common conditions
Embryology and Anatomy
could be potentially disastrous. In this chapter,
the embryology and anatomy of the vertebral
Embryology
arteries is reviewed, with a focus on normal
variants and vascular territories. Vertebral
Vertebral artery embryology forms a foundation
artery diseases are then discussed, including
for understanding normal and variant anatomy,
atherosclerotic disease, vertebral artery dissec-
which is useful in discriminating normal anatomic
tion, subclavian steal syndrome, and a variety
variation from pathologic processes. Recognition
of other conditions. Each section is organized
of normal variants is also important prior to sur-
by epidemiology and clinical findings, patho-
gical or endovascular intervention.
physiology, imaging findings, and work-up
The vertebral arteries develop from the dorsal
and management.
longitudinal neural arteries, which arise from lon-
gitudinal plexiform anastomoses of the first six
Keywords
cervical intersegmental arteries (Fig. 1) [1]. The
Vertebral artery • Carotid-vertebrobasilar vari-
dorsal longitudinal arteries run parallel to the
ants • Vertebral artery fenestration • Basilar
primitive internal carotid arteries and appear by
artery fenestration • Posterior circulation • Ath-
day 30 of embryonic life. The cranial portions of
erosclerosis • Stroke • Vertebral artery stenosis
the dorsal longitudinal neural arteries gradually
• Vertebral artery dissection • Subclavian steal
appose each other, fusing in the midline around
syndrome • Bow hunter’s syndrome
day 35 to form the basilar artery [1]. As the basilar
artery matures, it begins supplying the posterior
circulation, which is previously supplied by the
Introduction primitive internal carotid arteries via the fetal pos-
terior communicating arteries.
Radiologists are frequently asked to evaluate the Prior to the development of the mature
course, caliber, and integrity of the vertebral arter- vertebrobasilar system, temporary anastomoses
ies. Successful evaluation of the vertebral arteries between the cranial portions of the dorsal longitu-
requires a thorough knowledge of vertebral artery dinal neural arteries and the primitive internal
anatomy and pathology. carotid arteries exist [1]. These carotid-
This chapter begins by reviewing the embryol- vertebrobasilar anastomoses, as well as the cervi-
ogy and anatomy of the normal vertebral arteries. cal intersegmental arteries, eventually regress
Normal variants are discussed in this section, as (with the exception of the seventh cervical
they are commonly seen in clinical practice. This intersegmental arteries which develop into the
is followed by a general discussion of the most subclavian arteries). If these anastomoses fail to
frequently employed imaging modalities used in regress, communication between the
evaluating the vertebral arteries. This is followed vertebrobasilar system and the internal carotid
by discussion of the two most commonly artery persists into adulthood. Four major
10 Imaging of the Pathology of the Vertebral Arteries
Fig. 1 Embryology of the
vertebral arteries. The
vertebral arteries develop Right vertebral artery
from longitudinal plexiform
anastomoses of the first six
Right thyrocervical artery
cervical intersegmental
arteries. The seventh
cervical intersegmental
arteries develop into the
subclavian arteries
Cervical intersegmental arteries
carotid-vertebrobasilar variants have been
described in the literature and are named
according to the cranial nerves that they parallel
(Fig. 2): the persistent trigeminal artery, the per-
sistent otic artery, the persistent hypoglossal
artery, and the proatlantal intersegmental artery
[1–3]. In general, these variants are considered
incidental findings; however, an association with
an increased incidence of aneurysms has been
described [3].
The persistent trigeminal artery is the most
common carotid-vertebrobasilar variant, present
in 0.1–1.0 % of cerebral angiograms and autopsies
[1, 3]. It arises from the cavernous internal carotid
artery and runs across the dorsum sellae to con-
nect with the distal basilar artery (Fig. 3); the
basilar artery proximal to the aberrant connection
is often hypoplastic. On sagittal imaging, the per-
sistent trigeminal artery forms a characteristic
“tau” or “trident” sign (Fig. 4) [5]. The persistent
trigeminal artery can occasionally be mistaken for
an aneurysm on cross-sectional imaging; 3D CTA
or DSA can be helpful in these difficult cases
(Figs. 5 and 6). Patients are typically asymptom-
atic but can uncommonly present with trigeminal
165
Right internal carotid artery
Right external carotid artery
Aorta
1
2
3
4
5
6
7
neuralgia if there is significant mass effect on the
trigeminal nerve.
The persistent hypoglossal artery is the second
most common carotid-vertebrobasilar variant,
present in 0.1–0.25 % of cerebral angiograms
and autopsies [1, 3]. It arises from the distal cer-
vical internal carotid artery (usually between the
C1 and C3 vertebral bodies), passes through an
enlarged hypoglossal canal, and joins the proxi-
mal basilar artery (Figs. 7 and 8). Patients are
typically asymptomatic but can uncommonly pre-
sent with glossopharyngeal neuralgia and hypo-
glossal paralysis if there is significant mass effect
on the hypoglossal nerve.
The persistent otic artery is exceedingly rare,
having only been described in a handful of case
reports [1, 3]. It arises from the petrous internal
carotid artery, exits via the internal acoustic mea-
tus, and joins the proximal basilar artery.
A fourth variant, the proatlantal intersegmental
artery, is similar to the persistent hypoglossal
artery in that it arises from the distal cervical
internal carotid artery and joins the proximal bas-
ilar artery; however, instead of traveling through
the hypoglossal canal, the proatlantal
166
Trigeminal artery
Otic artery
Hypoglossal
artery
Proatlantal
intersegmental
artery
Vertebral
artery
Carotid
artery
Fig. 2 Carotid-vertebrobasilar variants. The persistent tri-
geminal artery is the most common variant, connecting the
cavernous internal carotid artery with the distal basilar
artery. The persistent otic artery connects the petrous inter-
nal carotid artery with the proximal basilar artery via the
internal acoustic meatus. The persistent hypoglossal artery
connects the distal cervical internal carotid artery with the
proximal basilar artery via the hypoglossal canal. The
proatlantal intersegmental artery connects the distal cervi-
cal internal carotid artery with the proximal basilar artery
via the foramen magnum
intersegmental artery travels through the foramen
magnum [1, 5].
Vertebral artery fenestrations occur in 0.2–2.0 %
of cerebral angiograms and autopsies [1]. Extracra-
nial fenestrations are thought to arise due to the
persistence of the cervical intersegmental arteries
while intracranial fenestrations are thought to arise
due to the persistence of basivertebral anastomoses.
On imaging, vertebral artery fenestrations are char-
acterized by a “gap” in the middle of a vertebral
artery; this can be confusing on axial imaging, with
visualization of two apparent lumens, but is nicely
demonstrated on reformats (Fig. 9). Vertebral artery
fenestrations are associated with other vascular
anomalies and predispose to vertebral artery aneu-
rysms [1–3]. Basilar artery fenestration is more
D. Chiao and M. Wintermark
Fig. 3 MRA in a 45-year-old female demonstrating a
persistent trigeminal artery (arrow) which connects the
cavernous internal carotid artery with the distal basilar
artery
common than vertebral artery fenestration, occur-
ring in up to 5 % of cerebral angiograms and
autopsies, and also predisposes to aneurysms
(Fig. 10) [1].
Duplication of the vertebral artery is very rare.
It is thought to be due to failure of fusion of the
cervical intersegmental arteries during
development [1].
Anatomy
The vertebral arteries usually arise from the prox-
imal subclavian arteries. However, in approxi-
mately 5 % of patients, the left vertebral artery
arises directly from the aortic arch (Fig. 11)
[5]. The vertebral arteries are typically asymmet-
ric in size, most commonly with the left vertebral
artery being dominant. When the vertebral arteries
are approximately equal in size, they are consid-
ered to be codominant. The normal luminal diam-
eter of the extracranial vertebral artery is
approximately 3–5 mm.
After arising from the subclavian arteries, the
vertebral arteries ascend in the neck and are
10 Imaging of the Pathology of the Vertebral Arteries 167

Fig. 4 “Tau” sign (arrow)

on MRA sagittal reformat in
a patient with a persistent
trigeminal artery. The
persistent trigeminal artery
is the most common
carotid-vertebrobasilar
variant

foramen, and then exits superiorly through the

C2 transverse foramen to become the V3 segment.
The V2 segment resides inside the vertebral col-
umn (Fig. 13) and is vulnerable to vertebral frac-
tures, especially fractures which violate the
transverse foramina.

Fig. 5 3D carotid angiography nicely demonstrates a per-
sistent trigeminal artery (arrow) which was mistaken for an
aneurysm on prior MRA
categorized into four segments, abbreviated V1,
V2, V3, and V4 (Fig. 12).
V1 Segment
The V1 segment is the first segment of the verte-
bral artery. The V1 segment typically arises from
the subclavian artery, courses posteriorly and
superiorly in the superior mediastinum, and then
enters the C6 transverse foramen to become the
V2 segment.
V2 Segment
The V2 segment courses superiorly through the
C3–C6 transverse foramina, turns superolaterally
through the “inverted L-shaped” C2 transverse
V3 Segment
The V3 segment courses posteriorly and medially
around the atlantooccipital joint (forming a
groove along the posterior ring of the atlas) and
then sharply turns anteriorly and superiorly to
pierce the dura to become the V4 segment. The
V3 segment resides outside the vertebral column
and is vulnerable to rotational forces which can
injure the artery at its dural insertion.
V4 Segment
The V4 segment courses superiorly, anteriorly,
and medially through the foramen magnum and
then superomedially behind the clivus. The V4
segments unite to form the basilar artery near the
pontomedullary junction.
Along its course, the vertebral artery gives rise
to numerous arterial branches: cervical branches,
meningeal branches, and intracranial branches.
These arteries are described below.
Cervical Branches
The cervical branches originate from the extracra-
nial vertebral artery as it courses up the neck. The
cervical branches are categorized into the spinal
branches (which supply the spinal cord and verte-
bral bodies) and the muscular branches (which
supply the deep cervical musculature). The
168
Fig. 6 Right carotid DSA
in the same patient as in
Fig. 5 with frontal (a) and
lateral (b) projections, again
demonstrating a persistent
trigeminal artery (arrow)
Fig. 7 CTA demonstrating a persistent hypoglossal artery
(arrow) traversing through the hypoglossal canal
muscular branches provide an important collateral
pathway in the vertebrobasilar circulation, com-
municating with the occipital branch of the exter-
nal carotid artery.
Meningeal Branches
The meningeal branches originate from the distal
extracranial vertebral artery. The meningeal
branches are categorized into the anterior menin-
geal artery and the posterior meningeal artery. The
anterior meningeal artery arises from V2 and sup-
plies the dura around the foramen magnum. The
posterior meningeal artery arises from V3 and
supplies the falx cerebellum.
D. Chiao and M. Wintermark
Intracranial Branches
The intracranial branches originate from the intra-
cranial vertebral artery (i.e., the V4 segment). The
intracranial branches are categorized into the ante-
rior spinal arteries, the posterior spinal arteries,
the posterior inferior cerebellar arteries, and the
perforating arteries. The anterior spinal arteries
unite in 50 % of cases and course inferiorly in
the anteromedian sulcus of the spinal cord to
supply the anterior spinal cord. The posterior
spinal arteries course inferiorly along the
posterolateral spinal cord to supply the posterior
spinal cord. The posterior inferior cerebellar
arteries course around the medullary tonsils to
supply portions of the medulla, inferior
cerebellum, and choroid plexus of the fourth ven-
tricle. The perforating arteries arise off the intra-
cranial vertebral artery to supply portions of the
medulla, inferior cerebellar peduncles, and
olivary bodies.
Variant anatomy of the posterior inferior cere-
bellar artery is common. Anomalous origin of the
posterior inferior cerebellar arteries from V3
occurs in up to 10 % of cases. Duplication of the
posterior inferior cerebellar arteries is less com-
mon, occurring in approximately 2 % of cases
[4]. Occasionally, the posterior inferior cerebellar
arteries arise from a single common trunk. Occa-
sionally, the vertebral artery terminates in the pos-
terior inferior cerebellar artery; in this situation,
the contralateral vertebral artery supplies the
majority of the posterior circulation blood flow
[4]. Familiarity with these common variants helps
avoid confusion when interpreting the intracranial
branches of the vertebral arteries.
10 Imaging of the Pathology of the Vertebral Arteries 169

Fig. 8 Left carotid DSA in

the same patient as in Fig. 7
with frontal (a) and lateral
(b) projections
demonstrating a persistent
hypoglossal artery (arrow)

Fig. 9 (a) Axial time-of-

flight MRA demonstrates
two apparent lumens
(arrow) of the right
vertebral artery. (b) Coronal
time-of-flight MRA
maximum-intensity
projection reveals that this
represents a focal
fenestration (arrow) of the
V2 segment

Fig. 10 Right vertebral artery DSA (a) and 3D angiogram artery DSA after coil embolization demonstrates no signif-
(b) demonstrating a fenestrated proximal basilar artery icant filling of the aneurysm sac
with an associated saccular aneurysm. (c) Right vertebral

Vascular Territories insufficiency results in neurologic deficits that are

significantly different than those incurred by
The vertebrobasilar system supplies the posterior carotid artery insufficiency. Typical symptoms of
circulation of the brain. As such, vertebrobasilar vertebrobasilar insufficiency include bilateral
170 D. Chiao and M. Wintermark

motor and sensory deficits, nausea, vomiting,
vertigo, ataxia, diplopia, dysarthria, and dysmetria
[4, 5]. The precise deficits can be predicted based
on the vascular territory of the arterial insult.
Cerebellum
The cerebellum is predominantly supplied by
three paired arteries: the superior cerebellar
arteries, the posterior inferior cerebellar arteries,
and the anterior inferior cerebellar arteries. The
superior cerebellar arteries supply the superior
surface of the cerebellum and the upper vermis.
The posterior inferior cerebellar arteries supply
the posterior and inferior surfaces of the cerebel-
lum and inferior vermis. The anterior inferior cer-
ebellar arteries supply the petrosal surface of the
cerebellum. The central portions of the vermis and
cerebellar hemispheres represent a watershed area
with variable contributions from all three vessels.

Brainstem
The brainstem is predominantly supplied by the
basilar artery and its perforating branches. How-
ever, the superior cerebellar arteries supply por-
tions of the superior pons while the distal vertebral
arteries and the posterior inferior cerebellar arter-
ies supply portions of the medulla. Occlusion of
one of the posterior inferior cerebellar arteries can
cause infarction of the lateral medulla, resulting in
Wallenberg syndrome. This syndrome is clini-
cally characterized by contralateral sensory defi-
cits affecting the body and extremities, as well as
ipsilateral sensory deficits affecting the face.

Imaging

Fig. 11 Oblique sagittal reformat NECT demonstrating The vertebral arteries can be evaluated with a
the left vertebral artery arising directly from the aorta variety of imaging modalities, including ultra-
(arrow) sound, MRA, CTA, and angiography. Ultrasound

Fig. 12 3D vessel analysis

in a normal patient with
frontal (a) and oblique (b)
projections demonstrating
the four standard vertebral
artery segments, V1–V4
10 Imaging of the Pathology of the Vertebral Arteries
Fig. 13 3D vessel analysis in a normal patient demon-
strating the relationship of V2 with the bony cervical
vertebral column
with pulsed Doppler using a linear 5.0–7.5 MHz
probe is a relatively inexpensive examination but
is significantly limited by osseous structures, cal-
cified atherosclerosis, patient body habitus, and
operator experience. Thus, cross-sectional imag-
ing is typically performed for full evaluation of
the vertebral arteries.
Two-dimensional time-of-flight MRA typi-
cally provides satisfactory visualization of the
vertebral arteries. A superior saturation band is
commonly applied to eliminate jugular venous
flow. However, it should be noted that a superior
saturation band can potentially hide reversed ver-
tebral artery flow. Thus, in cases of suspected
subclavian steal syndrome or bow hunter’s syn-
drome, a phase contrast sequence can be
performed to unmask reversal of flow. The general
limitations of time-of-flight MRA include
overestimation of stenosis in the setting of vessel
calcification, vessel tortuosity, or slow flow.
Dynamic contrast-enhanced MRA is able to
171
overcome many of these limitations but is still
prone to poor spatial resolution, patient motion,
and metal artifacts.
An alternative to MRA is CTA, which provides
excellent visualization of the vertebral arteries
with superior spatial resolution and shorter imag-
ing times. In addition, CTA is generally offered
day and night at most facilities. The scanning
range typically includes the aortic arch to the
circle of Willis in order to visualize the whole
length of the vertebral arteries. Automatic trigger-
ing or test bolus techniques can be used to time the
contrast injection. An adjunct non-contrast and
post-contrast head CT is often performed with
the CTA. One weakness of CTA is that it tends
to overestimate arterial stenosis in the setting of
heavy vessel calcification. In addition, the direc-
tion of arterial flow cannot be determined with
conventional CT imaging.
Angiography has historically been considered
the “gold standard” in vertebral artery imaging but
is now rarely performed as the initial imaging
examination. In the modern era, angiography is
typically reserved for resolving diagnostic
uncertainties, planning interventions, or
performing therapeutic procedures. Vertebral
artery angiography is typically performed with
rapid sequences during subclavian artery injec-
tions. Selective imaging can also be obtained,
usually in the AP, lateral views, and oblique
views. Angiography carries a small inherent risk
of vascular injury or distal embolization. The risk
of TIA and stroke is approximately 3 % and <1 %
respectively with a cerebral angiogram
examination [4].
The ideal imaging modality for the evaluation
of the vertebral arteries depends on the clinical
context. In cases of suspected vertebral artery
occlusion or dissection, CTA of the head and
neck is a reasonable initial examination, providing
rapid visualization of the vertebral arteries with
excellent spatial resolution, thus potentially min-
imizing the time to revascularization. However, in
cases of subclavian steal syndrome or bow
hunter’s syndrome, dynamic imaging with
contrast-enhanced MRA may be preferred.
Finally, in patients with tenuous renal function,
time-of-flight MRA may be selected in order to
172
avoid contrast agents and the risk of nephrogenic
systemic fibrosis.
Atherosclerotic Disease
Epidemiology and Clinical Findings
The most common nontraumatic condition affect-
ing the extracranial vertebral arteries is atheroscle-
rotic disease [6]. Atherosclerosis is a systemic
arterial disease, prevalent in industrialized
nations. It is responsible for the majority of ische-
mic strokes, resulting in healthcare-associated
costs over $30 billion each year [7]. There are
multiple well-known risk factors for atherosclero-
sis, both modifiable and nonmodifiable. Common
nonmodifiable risk factors include family history,
age, and male gender. Common modifiable risk
factors include age, hypertension, hyperlipidemia,
diabetes, smoking, and renal failure (Table 1) [8].
There is a strong association between vertebral
artery atherosclerosis and carotid artery athero-
sclerosis; however, the vertebral arteries tend to
be less severely and less diffusely affected than
the carotid arteries [6].
Most patients with vertebral artery atheroscle-
rosis are asymptomatic; however, if the degree of
vertebral artery stenosis is significant, patients can
present with vertebrobasilar insufficiency which
can manifest as vertigo, ataxia, imbalance, syn-
cope, and cranial nerve deficits [8]. Patients can
also present with a myriad of non-neurologic
symptoms related to systemic atherosclerosis,
including angina from coronary artery disease
and claudication from peripheral vascular disease.
Table 1 Risk factors for atherosclerotic disease
Nonmodifiable Family history
Age
Male gender
Modifiable Hypertension
Hyperlipidemia
Diabetes
Smoking
Renal failure
D. Chiao and M. Wintermark
Pathophysiology
Atherosclerosis is a complex pathologic process
with endothelial injury playing a central role in its
pathogenesis [8, 9]. Endothelial injury results in
increased permeability of the arterial wall to
lipids, which deposit in the intima resulting in
the fibrofatty plaque. Microscopically, the
fibrofatty plaque is composed of lipids, foam
cells, and connective tissue. A fibrotic cap, com-
posed of smooth muscle and collagen, forms over
the surface of the fibrofatty plaque. Disruption of
the fibrotic cap results in exposure of the
thrombogenic fibrofatty plaque to the blood-
stream, which can result in acute thrombosis
and/or thromboembolus [9]. In addition, as the
fibrofatty plaque enlarges, it progressively nar-
rows the arterial lumen. In general, a 50 %
decrease in luminal diameter is required before
hemodynamic consequences occur, as compensa-
tory physiologic vasodilation mitigates the effects
of mild atherosclerotic disease [9]. It should be
noted that tandem stenoses can be additive in their
hemodynamic effects, much like serial resistances
in a circuit [8]. Ultimately, clinical symptoms
occur when the arterial demand outstrips the arte-
rial supply, resulting in a demand-supply
mismatch and end-organ ischemia. If the
mismatch is transient and infarction has not
occurred, then the clinical manifestations are tran-
sient and the episode is referred to as a transient
ischemic attack. However, if the mismatch is
prolonged and infarction has occurred, then the
clinical manifestations are prolonged and the epi-
sode is referred to as a cerebrovascular accident or
stroke. Patients with prior transient ischemic
attacks have a ten times increased risk of
stroke [9].
Imaging Findings
Vertebral artery stenosis severity is categorized
per the North American Symptomatic Carotid
Endarterectomy Trial (NASCET) method
[10]. The NASCET method defines the percent
stenosis as the ratio of the difference between the
normal and diseased luminal diameter to the
10 Imaging of the Pathology of the Vertebral Arteries 173

Table 2 Grading of atherosclerotic disease

Mild <50 %
Moderate 50–69 %
Severe 70 %

normal luminal diameter (i.e., percent stenosis =

(normal lumen minimal diseased lumen) / (nor-
mal lumen)  100%). Flow-limiting atheroscle-
rotic disease is graded by the percent stenosis,
with mild narrowing <50 %, moderate narrowing
50–70 %, and severe narrowing >70 % (Table 2).
Irregular atheromatous plaques are at an increased
risk of thromboembolism and are considered an
independent risk factor for stroke [4].
Ultrasound with pulsed Doppler has a very
limited role in evaluating vertebral artery athero-
Fig. 14 CTA demonstrating moderate stenosis of the left
sclerosis, largely due to its inability to visualize
vertebral artery origin (arrow) due to ostial calcified ath-
the vertebral artery in its entirety. As with other erosclerotic disease
blood vessels, arterial stenosis is characterized on
ultrasound by increased velocities, turbulence,
high-resistance waveforms, and elevated resistive
indices [6, 8, 11]. The post-stenotic segment is
characterized by decreased velocities, loss of
pulsatility, “parvus-tardus” waveforms, and
decreased resistive indices. Severe stenosis in
one vertebral artery can result in overestimation
in stenosis of the contralateral vertebral artery due
to compensatory increased blood flow [9].
CTA or MRA is often the initial diagnostic
examination for vertebral artery atherosclerosis,
with >90 % sensitivity and >90 % specificity
[5, 8, 9]. Both imaging modalities offer complete
visualization of the vertebral arteries and allow for
routine NASCET characterization of atheroscle-
rotic disease. In general, atherosclerosis is charac-
terized by smooth or irregular plaques resulting in
luminal narrowing, often with concomitant arte-
rial wall calcification [8]. Atherosclerotic disease
is typically most severe at the ostia of the vertebral
arteries (Fig. 14). Multiplanar reformats and/or Fig. 15 CTA demonstrating severe stenosis of the left
extracranial vertebral artery (arrow) due to atherosclerotic
3D vessel analysis can be useful in accurately disease
quantifying the degree of arterial stenosis, espe-
cially in vessel segments which do not run orthog- artery (Figs. 15 and 16 respectively). However,
onal to the axial plane. if the clinical suspicion for significant atheroscle-
Advanced atherosclerotic disease can manifest rosis is low (e.g., young patient without athero-
as near-occlusion or occlusion of the vertebral sclerotic risk factors), other causes of vertebral
174
artery stenosis should be considered, such as ver-
tebral artery dissection, radiation arteritis,
fibromuscular dysplasia, or vasculitis [8]. In
these cases, MRA may provide additional infor-
mation by directly visualizing fibrofatty athero-
sclerotic plaque and excluding other causes of
luminal narrowing (Fig. 17). Lastly, the role of
DSA in evaluating atherosclerotic disease is
Fig. 16 CTA demonstrating complete occlusion of the
right vertebral artery (arrow)
Fig. 17 Axial (a) and
coronal (b) MRA of the
neck demonstrating a
fibrofatty atherosclerotic
plaque (arrows) of the left
vertebral artery
D. Chiao and M. Wintermark
generally reserved for planning for endovascular
interventions (Fig. 18).
Work-Up and Management
The management of vertebral artery atheroscle-
rotic disease has historically been with medical
therapy targeting atherosclerotic risk factors such
as hypertension or hypercholesterolemia. Sur-
gery, while frequently performed for carotid
artery atherosclerosis, is rarely performed for
vertebral artery atherosclerosis [12]. With the
advent of minimally invasive endovascular tech-
niques, angioplasty and stenting now have a role
in patients who fail maximum medical therapy.
Endovascular treatment involves aggressive pro-
cedural anticoagulation and deployment of a self-
expanding stent across the stenotic lesion. The
angiographic success rate of endovascular tech-
niques has been reported to be up to 95 %, with a
stroke rate of <2 % [9]. However, endovascular
stenting suffers from restenosis and neointimal
hyperplasia which often occurs at the margin of
previously placed stents (Fig. 19) [5]. Drug-
eluting stents and other methods of reducing
10 Imaging of the Pathology of the Vertebral Arteries 175

Fig. 18 (a) Right vertebral

artery DSA demonstrates a
high-grade stenosis of V4.
(b) Right vertebral artery
DSA after angioplasty and
placement of a 2  8 mm
chromium cobalt stent
demonstrates an excellent
angiographic result

Fig. 19 (a) Right vertebral

artery DSA 2 years later in
the same patient as in
Fig. 18 demonstrates a >90
% stenosis at the distal
aspect of the previously
placed stent. (b) Right
vertebral artery DSA after
angioplasty and placement
of a second 2  8 mm
chromium cobalt stent
demonstrates an excellent
angiographic result

restenosis and neointimal hyperplasia are cur-
rently being developed and studied.
Vertebral Artery Dissection
Epidemiology and Clinical Findings
Vertebral artery dissection has an annual inci-
dence of 1 per 100,000 patients and is one of the
most common causes of stroke in young adult
patients [4]. Vertebral artery dissection constitutes
approximately 27 % of all craniocervical arterial
dissections and causes up to 40 % of all posterior
fossa ischemic strokes [4, 13]. The mean age at
presentation is approximately 45 years; there is a
slight male predominance [4, 14, 15]. Patients can
present with dizziness/vertigo, headache, neck
pain, ataxia, visual symptoms, nausea/vomiting,
nystagmus, Horner syndrome, sensory deficits,
cranial nerve palsies, dysphagia, and tinnitus
(Table 3) [13, 16]. In a meta-analysis of nearly
2,000 patients, Gottesman et al. found that the
most common symptom was dizziness/vertigo
(58 %), followed by headache (51 %) and neck
pain (48 %) [16]. However, it should be noted that
nearly one in four patients with proven vertebral
artery dissection have no headache or neck pain at
176 D. Chiao and M. Wintermark

Table 3 Signs and symptoms of vertebral artery dissec- Table 5 Risk factors for vertebral artery dissection [16]
tion, sorted by frequency [16]
Trauma
Dizziness/vertigo 58 % Hypertension
Headache 51 % Connective tissue disorders
Neck pain 46 % Fibromuscular dysplasia
Ataxia 38 % Vasculitis
Visual symptoms 36 % Autosomal dominant polycystic kidney disease
Nausea/vomiting 35 % Hyperhomocysteinemia
Nystagmus 29 % Osteogenesis imperfecta
Horner syndrome 22 %
Sensory deficits 21 %
Cranial nerve palsies 21 %
Major trauma (e.g., motor vehicle accident or
Dysphagia 13 %
Tinnitus 7%
penetrating injury) is the single most important
risk factor for vertebral artery dissection. How-
ever, even minor trauma (e.g., abrupt head turn-
ing, sports injury, and chiropractic injury) can
Table 4 Complications of vertebral artery dissection, result in vertebral artery dissection, being reported
sorted by location [16] in 15–70 % of cases [16]. Other risk factors
Stroke Any 60 % include hypertension, connective tissue disorders,
Extracranial 68 % fibromuscular dysplasia, and vasculitis (Table 5)
Intracranial 32 % [17]. When dissection occurs in the absence of
TIA Any 14 % trauma, it is referred to as spontaneous vertebral
Extracranial 21 % artery dissection. While rare, spontaneous verte-
Intracranial 11 %
bral artery dissection is a potentially devastating
SAH Any 35 %
event because delayed diagnosis can result in poor
Extracranial 0%
outcomes.
Intracranial 57 %

Pathophysiology

the time of diagnosis; thus, the absence of these
symptoms do not exclude the diagnosis.
The major complication of vertebral artery dis-
section is cerebral infarction due to vertebral
artery occlusion; this occurs in approximately
60 % of patients [16]. The risk of stroke is highest
in the first few weeks after dissection but can
occur up to a month after dissection [16]. Ischemic
events are particularly common in extracranial
vertebral artery dissection with stroke occurring
in 68 % of patients with extracranial vertebral
artery dissection compared to 32 % in intracranial
vertebral artery dissection [16]. In contrast, sub-
arachnoid hemorrhage occurs almost exclusively
in intracranial vertebral artery dissection with an
incidence of approximately 60 % in patients with
intracranial vertebral artery dissection
(Table 4) [16].
Vertebral artery dissection results from a primary
tear of the intima, which allows blood to travel
from the lumen to the tunica media. The dissec-
tion flap extends cranially (i.e., with the direction
of blood flow) and forms a false lumen which can
compress and obstruct the true lumen resulting in
cerebrovascular ischemia. The dissection flap can
also extend into the adventitia resulting in
pseudoaneurysm formation. In addition, vertebral
artery dissection can result in vascular stasis
and become a nidus for distal thromboembolism
[4, 18].
The extracranial vertebral segments are more
prone to dissection than the intracranial segments
due to their greater mobility and proximity to
adjacent bony structures (Table 6) [16, 19]. How-
ever, the intracranial segments are more prone to
rupture due to their thinner adventitia [4].
10 Imaging of the Pathology of the Vertebral Arteries 177

Table 6 Vertebral artery dissection by vertebral artery Table 8 Outcomes after VAD by MRS [16]
segment [16]
MRS 0–1 67 %
V1 28 % MRS 2–4 18 %
V2 34 % MRS 5–6 10 %
V3 36 % MRS modified Rankin scale, MRS 0 no symptoms, MRS 1
V4 34 % no significant disability, MRS 2 slight disability, MRS 3
moderate disability, MRS 4 moderately severe disability,
MRS 5 severe disability, MRS 6 death

Table 7 Direct and indirect signs of vertebral artery dis-

section [19]
Direct Double lumen 22 %
Intimal flap 21 %
Indirect Arterial stenosis 51 %
String of pearls 48 %
Arterial dilatation 37 %
Arterial occlusion 36 %
Pseudoaneurysm 22 %
Most cases of vertebral artery dissection heal
spontaneously with 90 % of patients having sig-
nificant improvement in the first 2–3 months
[4]. Vertebral artery dissection has an overall
recurrence rate of 15 %, with 50 % of cases
recurring in the first month [4].
Imaging Findings
Given the variable and often nonspecific clinical
presentation of vertebral artery dissection, imag-
ing plays a primary role in diagnosis. Imaging
findings of vertebral artery dissection can be clas-
sified as direct or indirect (Table 7). Direct find-
ings include visualization of a double lumen or an
intimal flap. Indirect findings include arterial ste-
nosis, dilatation, alternating stenosis and dilata-
tion (“string of pearls”), occlusion, or
pseudoaneurysm formation. Overall, the most
common finding in vertebral artery dissection is
vertebral artery stenosis, which is seen in approx-
imately half of patients [19]. While direct signs of
vertebral artery dissection are more intuitive than
indirect signs, they are seen in less than one in four
patients with proven vertebral artery dissection
[19] (Table 8).
Two patterns of vertebral artery dissection
have been described in the literature: the steno-
occlusive pattern and the dissecting aneurysmal
pattern [4, 20]. The steno-occlusive pattern is due
to subintimal dissection resulting in narrowing of
the true lumen and is characterized by a tapered
stenosis or occlusion; long-segment stenosis has
been referred to as the “string” sign. This pattern is
typically seen in extracranial vertebral artery dis-
section and is associated with thromboembolic
phenomenon (Fig. 20) [20]. The dissecting aneu-
rysmal pattern is due to subadventitial extension
of the dissection flap and is characterized by focal
fusiform aneurysmal dilatation with or without
stenosis; alternating stenosis and dilatation has
been referred to as the “string-of-pearls” sign.
This pattern is typically seen in intracranial verte-
bral artery dissection and is associated with sub-
arachnoid hemorrhage (Fig. 21) [4]. A very specific
appearance of vertebral artery dissection is a nar-
row, eccentric lumen surrounded by crescent-
shaped mural thickening [22, 23, 24, 27]. In rare
cases, the lumen is preserved but is surrounded by
hyperdense intramural hematoma, referred to as the
“rind” sign [4, 21].
Conventional angiography has historically
been considered the “gold standard” in the diag-
nosis of vertebral artery dissection, though recent
studies have demonstrated that CTA has a near
100 % sensitivity and specificity when compared
to either conventional angiography or with the
final clinical diagnosis [21]. Today, CTA is typi-
cally the initial diagnostic examination for the
evaluation of acute vertebral artery dissection.
MRA is comparable to CTA but takes longer to
perform and is generally indicated for the
non-emergent evaluation of vertebral artery dis-
section [19]. One advantage of MRA over CTA is
178 D. Chiao and M. Wintermark

Fig. 20 Forty-three-year-old female with vertebral artery projections demonstrating abrupt narrowing and irregular-
dissection status post-motor vehicle collision. CTA of the ity of the right extracranial vertebral artery (arrows). Note
neck with axial (a) and coronal (b) maximum-intensity the adjacent transverse foramina fractures (open arrows)

Fig. 21 Forty-eight-year-
old female with
spontaneous intracranial
vertebral artery dissection
after awakening with severe
headache. (a) NECT
demonstrating diffuse
subarachnoid hemorrhage
and intraventricular blood.
(b) CTA demonstrating
occlusion of the right V4
segment (arrow); the left
vertebral artery is patent
(open arrow)

its superior ability of visualizing an intramural
hematoma, which is characterized as an eccentric
region of high T1 signal intensity, due to methe-
moglobin (Fig. 22) [20]. Another advantage of
MRA is its greater sensitivity for acute infarcts,
the major complication of vertebral artery dissec-
tion (Fig. 23).
The imaging findings of vertebral artery dis-
section on angiography are similar to CTA/MRA.
The characteristic finding is an abrupt caliber
change (Fig. 24): an isolated stenosis is seen in
46 % of cases, a stenosis with a concomitant
aneurysm is seen in 27 % of cases, an isolated
occlusion is seen in 20 % of cases, and an isolated
aneurysm is seen in 7 % of cases [4, 27].
Work-Up and Management
Because patients with vertebral artery dissection
can present with nonspecific symptoms and often
do not have significant risk factors (e.g., major
trauma or connective tissue disorders), a high
level of clinical suspicion is needed for diagnosis.
One possible algorithm for the work-up of vertebral
artery dissection is given below (Fig. 25) [19].
10 Imaging of the Pathology of the Vertebral Arteries 179

Fig. 22 Forty-two-year-
old male with spontaneous
vertebral artery dissection
after “thunderclap”
headache. (a) T1-WI with
fat saturation demonstrating
hyperintense intramural
hematoma (arrow). (b) 3D
contrast-enhanced MRA
demonstrating long-
segment stenosis (“string
sign”) of the left vertebral
artery

Fig. 23 MRI/MRA of the

head in the same patient. (a)
Time-of-flight MRA
maximum-intensity
projection demonstrating an
abrupt cutoff of the left
P1-P2 junction (arrow).
DWI and ADC mapping (b,
c) demonstrating a left
posterior cerebral artery
territory infarct. Vertebral
artery dissection serves as a
nidus for thromboembolism

In this algorithm, patients with normal renal func-
tion and no contrast allergy are evaluated with a
CTA; patients with abnormal renal function or a
documented contrast allergy are evaluated with a
time-of-flight MRA. Angiography is reserved for
troubleshooting in equivocal cases.
Treatment of vertebral artery dissection is con-
troversial because randomized controlled studies
are lacking. Anticoagulation is the most frequent
therapeutic intervention in extracranial dissec-
tions [18]. In intracranial dissections, the risks of
anticoagulation are higher because of the
increased incidence of subarachnoid hemorrhage.
Endovascular or surgical intervention can be con-
sidered in selected patients.
Follow-up imaging is often performed to eval-
uate for stability and/or resolution of vertebral
artery dissection (Fig. 26). On occasion, vertebral
180
Fig. 24 Angiogram pre- (a) and post- (b) coil emboliza-
tion of a right intracranial vertebral artery dissection (open
arrow) which involves the posterior inferior cerebellar
artery. Note the intraluminal filling defect (arrow),
Fig. 25 An imaging work-up algorithm for vertebral
artery dissection
artery dissection can progress and/or convert into
pseudoaneurysms, necessitating endovascular or
surgical intervention. In addition, follow-up imag-
ing is important for evaluating for recurrent ver-
tebral artery dissection or delayed infarcts. Most
recurrences occur within the first month and
involve multiple cervical arteries [28]. After the
first month, the risk of recurrence is only about
1 % per year, and recurrent dissection occurs
D. Chiao and M. Wintermark
representing thrombus formation. After coil embolization
of the right vertebral artery, the posterior circulation is
supplied entirely by the left vertebral artery
mainly in arteries not previously involved by the
dissection [18].
The clinical outcome of vertebral artery dissec-
tion is a function of the degree of cerebral ische-
mia caused by the dissection. Given that the
morbidity and mortality associated with vertebral
artery dissection is typically due to ischemic
infarcts, the modified Rankin scale (MRS) has
been used to categorize patient outcomes
(Table 6). Overall, patients have relatively good
outcomes after vertebral artery dissection, with
two-thirds of patients without significant perma-
nent disability [16].
Subclavian Steal Syndrome
Epidemiology and Clinical Findings
The subclavian steal syndrome refers to the devel-
opment of neurologic symptoms secondary to
retrograde flow in a vertebral artery; it was first
recognized by Contorni in 1960 and Reivich in
1961 [29]. Retrograde flow without neurologic
symptoms has been termed subclavian steal phe-
nomenon. The Joint Study of Extracranial Arterial
Occlusion reported that the incidence of subcla-
vian steal phenomenon is approximately 2.5 %,
with only 5 % of patients with subclavian steal
phenomenon presenting with neurologic
10 Imaging of the Pathology of the Vertebral Arteries 181

Fig. 26 CTA of the neck at

initial presentation (a) and
at 5-month follow-up (b)
after conservative
management of extracranial
vertebral artery dissection.
There is interval resolution
of the vertebral artery
stenosis (arrows)

Percentage of patients with neurologic symptoms

Fig. 27 Prevalence of 40%

neurologic symptoms in
patients with subclavian 35%
steal syndrome by brachial
blood pressure differential 30%

25%

20%

15%

10%

5%

0%
20-30 30-40 40-50 >50
Brachial blood pressure differential (mmHg)

symptoms [30]. The average age at presentation is
60 with a slight male predominance [4].
Vertebrobasilar symptoms arise when there is
inadequate blood flow to the posterior circulation
due to significant retrograde flow. Patients may
experience vertigo, ataxia, imbalance, and syn-
cope; in addition, cranial nerve deficits can occur
resulting in dysarthria, dysphagia, tinnitus, and
visual deficits [31–33]. Subclavian steal syndrome
rarely results in acute infarction because of the
collateral blood supply from the contralateral ver-
tebral artery [4]. However, if the contralateral ver-
tebral artery is significantly diseased, patients are at
increased risk of infarction of the brainstem,
cerebellum, and posterior cerebral hemispheres.
Because there is usually a proximal subclavian
artery stenosis resulting in the subclavian steal
phenomenon, patients can also present with ipsilat-
eral ischemic arm symptoms such as arm claudica-
tion, weakness, and paresthesias [4]. In fact, an
important clinical marker for subclavian steal syn-
drome is asymmetric brachial pressures [33, 34]. A
brachial blood pressure differential of greater than
20 mmHg is considered clinically significant, with
55–85 % of cases demonstrating subclavian steal
phenomenon [33]. Increasing arm blood pressure
differentials correlates with increasing incidence of
subclavian steal syndrome (Fig. 27) [33].
182
Pathophysiology
As discussed above, the subclavian steal phenom-
enon is typically due to a proximal, high-grade
subclavian artery stenosis or occlusion, resulting
in decreased distal subclavian artery pressures.
Vigorous exercise of the arm accentuates this
pressure gradient and provokes retrograde flow
from the ipsilateral vertebral artery, effectively
“stealing” blood flow from the posterior
circulation [4].
The etiology of the inciting subclavian artery
stenosis is most commonly atherosclerosis
[30]. In a study of over 7,000 neck ultrasounds,
Labropoulos et al. demonstrated that more than
80 % of cases of subclavian steal phenomenon
were due to atherosclerosis of the left subclavian
artery [35]. This is thought to be due to the acute
angle that the left subclavian artery takes as it
arises off the aortic arch, predisposing to flow
turbulence and atherosclerotic disease. Rarely,
subclavian steal syndrome can result from vascu-
litis, radiation angiopathy, aortic dissection,
extrinsic compression, or congenital stenosis [4].
A unique scenario can occur in patients with a
history of CABG with LIMA coronary bypass. In
these patients, the proximal subclavian artery ste-
nosis results in vascular steal not only from the
ipsilateral vertebral artery but also from the LIMA
coronary bypass which can result in coronary
ischemia [36]. These patients may present with
angina, especially after vigorous exercise of the
affected arm.
Imaging Findings
Subclavian steal syndrome can be assessed with a
variety of imaging modalities including ultra-
sound, CTA, MRA, and DSA. At some institu-
tions, Doppler ultrasound is used as an initial
screening test due to its availability and low cost
[37]. CTA, MRA, and/or DSA are then performed
as confirmatory studies. The characteristic finding
is reversed flow in a vertebral artery with a prox-
imal subclavian artery stenosis or occlusion.
D. Chiao and M. Wintermark
Table 9 Grading of subclavian steal syndrome [33]
Grade I Reduced antegrade flow
Grade II Alternating antegrade and retrograde flow
Grade III Permanent retrograde flow
Provocative maneuvers, such as ipsilateral arm
exercise or the hyperemia-ischemia cuff test,
may be employed to demonstrate blood flow
reversal.
On Doppler ultrasound, the normal vertebral
arteries demonstrate a low-resistance pattern with
cephalad flow throughout the cardiac cycle. In
patients with subclavian steal syndrome, there is
blunted antegrade or frank retrograde flow on the
affected side, often with compensatory increased
flow (peak systolic velocity >60 cm/s) on the
contralateral side [4, 36, 38, 39–42]. A grading
system based on waveform characteristics has
been developed, with grade I (mild steal) charac-
terized by reduced antegrade flow, grade II (mod-
erate steal) characterized by alternating antegrade
and retrograde flow, and grade III (severe steal)
characterized by permanent retrograde flow
(Table 9). The hyperemia-ischemia cuff test is
often used to provoke subclavian steal. The
hyperemia-ischemia cuff test involves inflating a
blood pressure cuff at least 20 mmHg above the
systolic blood pressure for several minutes and
then rapidly deflating the cuff with real-time
Doppler ultrasound of the ipsilateral vertebral
artery. The rapid cuff deflation leads to transient
increased blood flow to the arm, which can pro-
voke occult subclavian steal [36]. Infrequently,
the proximal subclavian artery stenosis can be
visualized on Doppler ultrasound, characterized
by a high-resistance waveform, increased veloc-
ity, and turbulent flow.
While Doppler ultrasound is excellent at dem-
onstrating the directional flow in the vertebral
arteries, it usually cannot visualize the proximal
subclavian artery stenosis. In these cases, cross-
sectional imaging with MRA or CTA can be
performed. The characteristic finding on cross-
sectional imaging is a high-grade stenosis or
occlusion in the subclavian artery proximal to
10 Imaging of the Pathology of the Vertebral Arteries 183

Fig. 28 Forty-eight-year-
old female with history of
radiation therapy for
non-small cell lung cancer
who presents with
paroxysms of vertigo. (a)
CTA of the head and neck
demonstrating complete
occlusion of the proximal
subclavian artery (arrow).
(b) DSA with injection of
the left subclavian artery
showing the same finding
(arrow)

Fig. 29 Same patient as in

Fig. 28. (a) DSA with
injection of the right
vertebral artery. (b) Delayed
imaging demonstrates
retrograde filling of the left
vertebral artery (arrow)
with opacification of the left
subclavian artery (open
arrow), compatible with
subclavian steal
phenomenon

the takeoff of the vertebral artery (Figs. 28 and 29)
[4]. Fibrofatty or calcified atherosclerotic plaque
is a typical finding and may appear irregular
and/or ulcerated. MRA is generally favored over
CTA because of its ability to visualize proximal
subclavian artery stenosis as well as its ability to
determine the direction of blood flow in the ver-
tebral arteries, especially with a phase contrast
sequence (Figs. 30 and 31).
As with most vascular pathology, angiography
is usually reserved for equivocal cases or prior to
endovascular treatment. The characteristic finding
in subclavian steal syndrome is reversal of flow or
“to-and-fro” flow in the affected vertebral artery,
usually with a high-grade proximal subclavian
artery stenosis (Fig. 32) [4].
Work-Up and Management
Only a small minority (approximately 18 %) of
patients with subclavian steal phenomenon
require therapeutic intervention [35]. Patients
with mild symptoms are generally treated conser-
vatively with medical treatment of the underlying
atherosclerotic risk factors (such as
184
Fig. 30 Fifty-five-year-old female with weakness and
paresthesias in her left hand after prolonged use.
Contrast-enhanced MRA maximum-intensity projection
demonstrates proximal left subclavian artery occlusion
(arrow)
hyperlipidemia, hypertension, diabetes, or
smoking) [43]. Patients with severe symptoms
may require surgical or endovascular treatment
with carotid-subclavian bypass grafting or angio-
plasty/stenting. Surgical bypass has enjoyed wide
success, with resolution of symptoms in the vast
majority of patients and patency rates of up to
95 % at 10 years [44]. More recently,
endovascular techniques have also been
employed successfully, with patency rates of
80 % at 5 years (Fig. 33) [45, 46]. Some authors
argue that endovascular techniques are preferred
over surgical techniques due to their similar effec-
tiveness, lower complication rates, and shorter
hospitalizations; however, long-term patency is
less durable with endovascular therapy, with a
10-year target vessel revascularization rate of
47 % compared to 6 % for bypass [47].
D. Chiao and M. Wintermark
Fig. 31 Eighty-three-year-old male with lower extremity
ataxia. 3D contrast-enhanced MRA demonstrates left sub-
clavian artery occlusion (arrow)
Bow Hunter’s Syndrome
Epidemiology and Clinical Findings
Bow hunter’s syndrome, also known as positional
or rotational occlusion of the vertebral artery, is a
rare cause of vertebrobasilar insufficiency. The
syndrome was originally described by Dr. Bruce
Sorensen in a case report in Neurosurgery in 1978
[48]. Sorensen described a 39-year-old man who
developed persistent right-sided weakness and
left-sided numbness during archery practice.
Based on clinical findings, the patient was thought
to have right medullary infarction resulting in
modified Wallenberg syndrome. However, the
patient had no significant atherosclerotic risk fac-
tors. On angiography, the patient was found to
have significant right vertebral artery spasm.
Sorensen reasoned that the patient’s condition
10 Imaging of the Pathology of the Vertebral Arteries
Fig. 32 Sixty-year-old male with intermittent episodes of
diplopia and dizziness. (a) DSA with injection of the aortic
arch. The right subclavian artery is not seen. (b) Subse-
quent imaging demonstrates retrograde filling of the right
Fig. 33 Same patient as in Fig. 32. DSA with injection of
the innominate artery after endovascular stent placement
(arrow) demonstrating antegrade flow in the right vertebral
(open arrow) and subclavian arteries
stemmed from positional occlusion of the verte-
bral artery during archery practice (i.e., turning
the head sharply to the left will occlude the right
vertebral artery and vice versa). In archers and
bow hunters, who stand perpendicular to a target
185
vertebral artery (arrow) with delayed opacification of the
right subclavian artery (open arrow). Right-sided subcla-
vian steal syndrome is less common than left-sided subcla-
vian steal syndrome
with their heads turned to the left (if right-
handed), the phenomenon of positional vertebral
artery occlusion is accentuated and could poten-
tially result in ischemia or infarction [48].
Since Sorensen’s original case report, at least
100 cases of bow hunter’s syndrome have been
reported in the literature [49]. The age at presen-
tation ranges widely, with most patients in their
40–60s [49, 50]. Patients typically present with
signs and symptoms of vertebrobasilar insuffi-
ciency, such as weakness, ataxia, numbness, tin-
gling, dysphagia, dysarthria, tinnitus, diplopia,
and nystagmus [4, 51]. Patients can also present
with nonspecific symptoms such as headache,
dizziness, nausea, and syncope [4, 52]. The key
finding is that clinical symptoms occur after a
sharp turn of the head.
Pathophysiology
Bow hunter’s syndrome is thought to arise due to
positional occlusion of the contralateral vertebral
artery after sharply turning the head. In general,
transient positional vertebral artery occlusion is
considered physiologic and is asymptomatic due
to sufficient contralateral vertebral artery flow or
collateral flow via the posterior communicating
arteries [50]. However, if the contralateral
186
Fig. 34 DWI (a) and ADC
mapping (b) demonstrating
an area of restricted
diffusion in the right
thalamus, concerning for an
acute infarct
vertebral artery is hypoplastic or significantly dis-
eased, bow hunter’s syndrome is more likely to
occur [50, 51]. Bow hunter’s syndrome is also
more likely if the affected vertebral artery origi-
nates the dominant posterior inferior cerebellar
artery [52].
Most cases of bow hunter’s syndrome occur at
the atlantoaxial level, which is especially prone to
dynamic injury [4]. During neck rotation, the con-
tralateral atlantoaxial joint rotates asymmetrically
forward and downward, potentially stretching the
vertebral artery between the C1 and C2 foramina
or under the atlantooccipital ligament [50]. Cervi-
cal spondylosis, herniated disks, and atlantoaxial
instability can predispose to vertebral artery com-
pression [50, 52]. In addition, cases of vertebral
artery compression by the longus colli and scalene
muscles have been reported [49].
Imaging Findings
Evaluation of bow hunter’s syndrome relies on
dynamic imaging. Imaging is typically performed
in the neutral, left, and right positions. The key
imaging finding is occlusion or stenosis of the
vertebral artery with the head turned to the left
or right. Particular attention should be made to the
V3 segment of the vertebral artery (i.e., at the
atlantoaxial level), which is the most commonly
affected segment.
Imaging is typically with MRA, CTA, or angi-
ography. Cross-sectional imaging with MRA or
D. Chiao and M. Wintermark
CTA can visualize areas of stenosis or occlusion.
In addition, cross-sectional imaging has the
advantage of evaluating extravascular pathology
which may explain or contribute to the patient’s
symptoms. Complications of bow hunter’s syn-
drome, such as acute stroke, can also be visualized
(Figs. 34 and 35).
The “gold standard” for diagnosis remains
dynamic angiography with head turning
[52]. Angiography has the advantage of real-
time visualization of stenosis or occlusion and
can demonstrate direction of flow in collateral
vessels (Fig. 36). Angiography is also frequently
performed to demonstrate angiographic improve-
ment after an intervention is performed [53].
Work-Up and Management
While bow hunter’s syndrome is rare, it can result
in repeated vertebral artery ischemia and/or per-
manent neurologic deficits if not diagnosed and
treated appropriately. Work-up for bow hunter’s
syndrome begins with a complete history and
physical examination. In particular, a thorough
description of the patient’s symptoms and exacer-
bating factors should be elicited, especially in
patients with evidence of vertebrobasilar insuffi-
ciency. In particular, reproducibility of symptoms
with head turning is highly suggestive of bow
hunter’s syndrome [49]. Occasionally, brainstem
auditory-evoked response testing is performed to
demonstrate increased latencies [51]. Dynamic
10 Imaging of the Pathology of the Vertebral Arteries 187

Fig. 35 CTA in the same

patient as in Fig. 34 with the
head turned to the left (a)
and with the head turned to
the right (b) demonstrating
positional narrowing of the
right vertebral artery
(arrow). Cervical
spondylosis, herniated
disks, and atlantoaxial
instability can predispose to
vertebral artery
compression

Fig. 36 DSA of the right

vertebral artery in the same
patient as in Fig. 34 in the
neutral position (a) and with
head turned to the left (b)
demonstrating positional
occlusion of the vertebral
artery (arrow). The V3
segment is the most
commonly affected
segment in bow hunter’s
syndrome

imaging studies can then be obtained to confirm
the diagnosis.
Patients with bow hunter’s syndrome are ini-
tially managed with a cervical brace to restrict
neck rotation. Historically, medical treatment
with anticoagulation was performed in mild
cases and surgical decompression and/or cervical
spinal fusion was performed in severe cases.
However, conservative management is effective
in preventing stroke in only half of affected
patients, and surgical management, while effec-
tive, can result in permanent restriction of neck
rotation and can be complicated by a number of
operative risks including infection, paralysis, and
death [49]. Recently, endovascular therapy with
self-expanding stents has been found to be an
effective, minimally invasive treatment modality
[52]. The downside to endovascular stenting is the
potential for restenosis; data on long-term patency
is unfortunately lacking. Nevertheless,
188
endovascular therapy remains a treatment option
and may supplant surgical management in
selected patients.
Miscellaneous Conditions
Fibromuscular Dysplasia
Fibromuscular dysplasia is a noninflammatory
condition, typically affecting middle-aged
women. It most commonly affects the renal,
carotid, or iliac arteries but can also affect the
vertebral arteries in 10–40 % of cervical vascular
cases. It is bilateral in two-thirds of patients.
While there are several types of fibromuscular
dysplasia, the most common type is
medial fibroplasia characterized by a “string-of-
pearls” appearance. Fibromuscular dysplasia
predisposes to ischemia and infarct, likely due to
the increased incidence of thromboembolism in
the affected arteries. It also predisposes to
spontaneous vertebral artery dissection. The
most sensitive examination for fibromuscular dys-
plasia is angiography, which is able to detect
subtle cases. In symptomatic cases, treatment
with balloon angioplasty with or without stenting
is warranted. Only rarely is surgical management
needed.
Vasculitis
Idiopathic vasculitis, including Takayasu’s arteri-
tis and giant cell arteritis, can occasionally affect
the vertebral arteries. Vasculitis is characterized
by thickening of the arterial wall with long,
smooth stenosis and/or “sausage-link” lesions.
Treatment is with steroids to reduce inflammatory
injury.
Radiation vasculitis can also affect the verte-
bral arteries and can occur months to years
after radiation therapy. It is characterized by irreg-
ular, calcified arteriopathy confined to the radia-
tion portal. Treatment is challenging, as irradiated
vessels do not respond well to angioplasty or
surgery.
D. Chiao and M. Wintermark
Summary
The vertebral arteries are clinically important
blood vessels. Given the relatively high incidence
of vertebral artery pathology, thorough interroga-
tion of the vertebral arteries with diagnostic imag-
ing is essential, with accurate interpretation
guiding management decisions. Collaboration
among the radiology, neurology, and neurosur-
gery teams is of utmost importance. Clinical con-
text greatly aids in the selection of the most
appropriate imaging modality. New and improved
anatomic imaging is resulting in improved sensi-
tivity and specificity in diagnosing vertebral artery
diseases. In addition, dynamic imaging with or
without provocative maneuvers is able to demon-
strate abnormal blood flow in functional vertebral
artery disorders. Finally, interventional neuroradi-
ology is becoming increasingly prevalent in ther-
apeutic management, with a variety of minimally
invasive endovascular techniques being used in
the treatment of a number of vertebral artery
diseases.
References
1. Okahara M, Kiyosue H, Mori H, Tanoue S, Sainou M,
Nagatomi H (2002) Anatomic variations of the cerebral
arteries and their embryology: a pictorial review. Eur
Radiol 12(10):2548–2561. doi:10.1007/s00330-001-
1286-x
2. Dimmick SJ, Faulder KC (2009) Normal variants of
the cerebral circulation at multidetector CT angiogra-
phy. Radiographics 29(4):1027–1043. doi:10.1148/
rg.294085730
3. Luh GY, Dean BL, Tomsick TA, Wallace RC (1999)
The persistent fetal carotid-vertebrobasilar anastomo-
ses. AJR Am J Roentgenol 172(5):1427–1432.
doi:10.2214/ajr.172.5.10227532
4. Osborn AG, Salzman KL, Barkovich AJ (2009) Diag-
nostic imaging. Brain, 2nd edn. Amirsys, Salt Lake
City
5. Osborn AG, Jacobs JM, Osborn AG, Ralph Erskine
Conrad Memorial Fund (1999) Diagnostic cerebral
angiography, 2nd edn. Lippincott Willims & Wilkins,
Philadelphia, p 462
6. Tay KY, U-King-Im JM, Trivedi RA et al (2005) Imag-
ing the vertebral artery. Eur Radiol 15(7):1329–1343.
doi:10.1007/s00330-005-2679-z
7. Go AS, Mozaffarian D, Roger VL et al (2013) Execu-
tive summary: heart disease and stroke statistics – 2013
update: a report from the american heart association.
10 Imaging of the Pathology of the Vertebral Arteries
Circulation 127(1):143–152. doi:10.1161/CIR.
0b013e318282ab8f
8. Osborn AG (1994) Diagnostic neuroradiology. Mosby,
St. Louis, p 936
9. Kaufman JA, Lee MJ (2004) Vascular and interven-
tional radiology: the requisites. Mosby, Philadelphia,
p 705
10. Ferguson GG, Eliasziw M, Barr HW et al (1999) The
north american symptomatic carotid endarterectomy
trial: surgical results in 1415 patients. Stroke
30(9):1751–1758
11. Vilela P, Goulao A (2005) Ischemic stroke: carotid and
vertebral artery disease. Eur Radiol 15(3):427–433.
doi:10.1007/s00330-004-2632-6
12. Stayman A, Nogueira RG, Gupta R (2013) Diagnosis
and management of vertebrobasilar insufficiency. Curr
Treat Options Cardiovasc Med 15(2):240–251.
doi:10.1007/s11936-013-0228-7
13. Arnold M, Bousser MG, Fahrni G et al (2006) Verte-
bral artery dissection: presenting findings and predic-
tors of outcome. Stroke 37(10):2499–2503.
doi:10.1161/01.STR.0000240493.88473.39
14. Schievink WI, Roiter V (2005) Epidemiology of cer-
vical artery dissection. Front Neurol Neurosci
20:12–15. doi:10.1159/000088125
15. Kristensen B, Malm J, Carlberg B et al (1997) Epide-
miology and etiology of ischemic stroke in young
adults aged 18 to 44 years in northern sweden. Stroke
28(9):1702–1709
16. Gottesman RF, Sharma P, Robinson KA et al (2012)
Clinical characteristics of symptomatic vertebral artery
dissection: a systematic review. Neurologist
18(5):245–254. doi:10.1097/NRL.0b013e31826754e1
17. Schievink WI (2001) Spontaneous dissection of the
carotid and vertebral arteries. N Engl J Med 344
(12):898–906. doi:10.1056/NEJM200103223441206
18. Rodallec MH, Marteau V, Gerber S, Desmottes L, Zins
M (2008) Craniocervical arterial dissection: spectrum
of imaging findings and differential diagnosis. Radio-
graphics 28(6):1711–1728. doi:10.1148/rg.286085512
19. Gottesman RF, Sharma P, Robinson KA et al (2012)
Imaging characteristics of symptomatic vertebral artery
dissection: a systematic review. Neurologist
18(5):255–260. doi:10.1097/NRL.0b013e3182675511
20. Shin JH, Suh DC, Choi CG, Leei HK (2000) Vertebral
artery dissection: spectrum of imaging findings with
emphasis on angiography and correlation with clinical
presentation. Radiographics 20(6):1687–1696
21. Chen CJ, Tseng YC, Lee TH, Hsu HL, See LC (2004)
Multisection CT angiography compared with catheter
angiography in diagnosing vertebral artery dissection.
AJNR Am J Neuroradiol 25(5):769–774
22. Bartels E, Flugel KA (1996) Evaluation of extracranial
vertebral artery dissection with duplex color-flow
imaging. Stroke 27(2):290–295
23. Sturzenegger M, Mattle HP, Rivoir A, Baumgartner
RW (1995) Ultrasound findings in carotid artery dis-
section: analysis of 43 patients. Neurology
45(4):691–698
189
24. Leclerc X, Godefroy O, Salhi A, Lucas C, Leys D,
Pruvo JP (1996) Helical CT for the diagnosis of extra-
cranial internal carotid artery dissection. Stroke
27(3):461–466
25. Levy C, Laissy JP, Raveau V et al (1994) Carotid
and vertebral artery dissections: three-dimensional
time-of-flight MR angiography and MR imaging ver-
sus conventional angiography. Radiology 190(1):
97–103
26. Mascalchi M, Bianchi MC, Mangiafico S et al (1997)
MRI and MR angiography of vertebral artery dissec-
tion. Neuroradiology 39(5):329–340
27. Houser OW, Mokri B, Sundt TM Jr, Baker HL Jr,
Reese DF (1984) Spontaneous cervical cephalic arte-
rial dissection and its residuum: angiographic spec-
trum. AJNR Am J Neuroradiol 5(1):27–34
28. Dittrich R, Nassenstein I, Bachmann R et al (2007)
Polyarterial clustered recurrence of cervical artery dis-
section seems to be the rule. Neurology
69(2):180–186. doi:10.1212/01.wnl.0000265595.
50915.1e
29. Reivich M, Holling HE, Roberts B, Toole JF (1961)
Reversal of blood flow through the vertebral
artery and its effect on cerebral circulation. N Engl J
Med 265:878–885. doi:10.1056/NEJM1961110226
51804
30. Fields WS, Lemak NA (1972) Joint study of extracra-
nial arterial occlusion. VII. subclavian steal – a review
of 168 cases. JAMA 222(9):1139–1143
31. Gosselin C, Walker PM (1996) Subclavian steal syn-
drome: existence, clinical features, diagnosis and man-
agement. Semin Vasc Surg 9(2):93–97
32. Herring M (1977) The subclavian steal syndrome: a
review. Am Surg 43(4):220–228
33. Osiro S, Zurada A, Gielecki J, Shoja MM, Tubbs RS,
Loukas M (2012) A review of subclavian steal syn-
drome with clinical correlation. Med Sci Monit 18(5):
RA57–RA63
34. Tan TY, Schminke U, Lien LM, Tegeler CH (2002)
Subclavian steal syndrome: can the blood pressure
difference between arms predict the severity of steal?
J Neuroimaging 12(2):131–135
35. Labropoulos N, Nandivada P, Bekelis K (2010) Prev-
alence and impact of the subclavian steal syndrome.
Ann Surg 252(1):166–170. doi:10.1097/SLA.
0b013e3181e3375a
36. Kliewer MA, Hertzberg BS, Kim DH, Bowie JD,
Courneya DL, Carroll BA (2000) Vertebral artery
doppler waveform changes indicating subclavian steal
physiology. AJR Am J Roentgenol 174(3):815–819.
doi:10.2214/ajr.174.3.1740815
37. Vecera J, Vojtisek P, Varvarovsky I, Lojik M,
Masova K, Kvasnicka J (2010) Non-invasive diagnosis
of coronary-subclavian steal: role of the doppler ultra-
sound. Eur J Echocardiogr 11(9):E34. doi:10.1093/
ejechocard/jeq068
38. Buckenham TM, Wright IA (2004) Ultrasound of the
extracranial vertebral artery. Br J Radiol 77(913):
15–20
190
39. Flynn PD, Delany DJ, Gray HH (1993) Magnetic res-
onance angiography in subclavian steal syndrome.
Br Heart J 70(2):193–194
40. Van Grimberge F, Dymarkowski S, Budts W, Bogaert J
(2000) Role of magnetic resonance in the diagnosis of
subclavian steal syndrome. J Magn Reson Imaging
12(2):339–342
41. Virmani R, Carroll TJ, Hung J, Hopkins J, Diniz L,
Carr J (2008) Diagnosis of subclavian steal syndrome
using dynamic time-resolved magnetic resonance angi-
ography: a technical note. Magn Reson Imaging
26(2):287–292. doi:10.1016/j.mri.2007.05.005
42. Akin K, Kosehan D, Kirbas I, Yildirim M, Koktener A
(2011) Diagnosis and percutaneous treatment of partial
subclavian steal: doppler ultrasonography and phase
contrast magnetic resonance angiography findings
and a brief review of the literature. Jpn J Radiol
29(3):207–211. doi:10.1007/s11604-010-0521-2
43. Olsen KG, Lund C (2006) Subclavian steal syndrome.
Tidsskr Nor Laegeforen 126(24):3259–3262
44. Qi L, Gu Y, Zhang J et al (2010) Surgical treatment of
subclavian artery occlusion. Zhongguo Xiu Fu Chong
Jian Wai Ke Za Zhi 24(9):1030–1032
45. De Vries JP, Jager LC, Van den Berg JC et al (2005)
Durability of percutaneous transluminal angioplasty
for obstructive lesions of proximal subclavian artery:
long-term results. J Vasc Surg 41(1):19–23.
doi:10.1016/j.jvs.2004.09.030
46. Palchik E, Bakken AM, Wolford HY, Saad WE, Davies
MG (2008) Subclavian artery revascularization: an
outcome analysis based on mode of therapy and
presenting symptoms. Ann Vasc Surg 22(1):70–78.
doi:10.1016/j.avsg.2007.07.020
D. Chiao and M. Wintermark
47. Song L, Zhang J, Li J et al (2012) Endovascular
stenting vs. extrathoracic surgical bypass for symptom-
atic subclavian steal syndrome. J Endovasc Ther
19(1):44–51. doi:10.1583/11-3692.1
48. Sorensen BF (1978) Bow hunter’s stroke. Neurosur-
gery 2(3):259–261
49. Kuether TA, Nesbit GM, Clark WM, Barnwell SL
(1997) Rotational vertebral artery occlusion: a mecha-
nism of vertebrobasilar insufficiency. Neurosurgery
41(2):427–432, discussion 432-3
50. Netuka D, Benes V, Mikulik R, Kuba R (2005) Symp-
tomatic rotational occlusion of the vertebral artery –
case report and review of the literature. Zentralbl
Neurochir 66(4):217–222. doi:10.1055/s-2005-
836600
51. Horowitz M, Jovin T, Balzar J, Welch W, Kassam A
(2002) Bow hunter’s syndrome in the setting of
contralateral vertebral artery stenosis: evaluation
and treatment options. Spine (Phila Pa 1976)
27(23):E495–E498. doi:10.1097/01.BRS.0000035308.
10464.37
52. Darkhabani MZ, Thompson MC, Lazzaro MA,
Taqi MA, Zaidat OO (2012) Vertebral artery
stenting for the treatment of bow hunter’s syndrome:
report of 4 cases. J Stroke Cerebrovasc Dis 21(8):908.
e1-908.e5. doi: 10.1016/j.jstrokecerebrovasdis.2011.
09.006
53. Velat GJ, Reavey-Cantwell JF, Ulm AJ, Lewis SB
(2006) Intraoperative dynamic angiography to detect
resolution of bow hunter’s syndrome: technical case
report. Surg Neurol 66(4):420–423; discussion 423.
doi: 10.1016/j.surneu.2006.03.040
 Carotid Artery Surgery
11
Roberto Montisci and Luca Saba

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191 Carotid endarterectomy is at present the most
frequent intervention on arterial vessels
Preoperative Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
performed in vascular surgery units. Currently
Surgical Technique of Endarterectomy . . . . . . . . . . . 196 the surgical technique is significantly
Carotid Body Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198 improved compared to 10 years ago with a
consequent reduction in terms of mortality
Aneurysms of the Extracranial Carotid
Artery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 and morbidity. This chapter will cover the sur-
gical approaches as well as the most important
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
indications.
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
Keywords
Carotid Endarterectomy • Carotid Artery •
Carotid Artery Surgery

Introduction

Carotid endarterectomy is at present the most

frequent intervention on arterial vessels
performed in vascular surgery units. Current
wide development of this technique of reconstruc-
tion for occlusive carotid disease is a consequence
of a long arduous surgical course that had a deci-
sive impulse in the early 1950s with first success-
ful operation performed by DeBakey in 1953 on a
R. Montisci (*) 53-year-old patient with “intermittent episodes of
Department of Vascular Surgery, AOU of Cagliari, weakness of the right arm and leg, hesitancy and
Monserrato, Cagliari, Italy
difficulty in speaking, and difficulty in writing
e-mail: Roberto.montisci@iol.it
clearly” [1]. In the same period, other important
L. Saba
steps in carotid revascularization were climbed
Department of Radiology, AOU of Cagliari, Monserrato,
Cagliari, Italy through different methods like resection of the
e-mail: lucasaba@tiscali.it; lucasabamd@gmail.com segment of the internal carotid artery narrowed
# Springer Science+Business Media New York 2016 191
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_51
192
and restoration of carotid circulation by direct
end-to-end anastomosis between the common
carotid artery and the distal internal carotid artery
or using a graft after resection of a segment of the
internal carotid artery (Eastcott 1954) [2]. In 1956
Lyons performed a subclavian to common carotid
graft bypass using a nylon prosthesis for a proxi-
mal common carotid occlusion [3]. In the same
year Cooley reported the first implant of a tempo-
rary shunt during carotid endarterectomy [4]. Suc-
cessively, the selective use of a temporary shunt
was recommended by Moore in 1969 on the basis
of the measurement of the stump pressure in the
internal carotid artery after proximal cross-
clamping and on the basis of EEG monitoring by
Collow in 1980 [5, 6].
Also anesthesia techniques shown a specific
evolution from the simple general anesthesia
toward the current spectrum of possibilities that
include local anesthesia via cervical plexus block
or various intermediate techniques like conscious
sedation during regional anesthesia [7] or pre-
served consciousness in general anesthesia [8].
In the late 1990s, the results of some large
prospective randomized trials heavily increased
the leaning to endarterectomy for carotid artery
occlusive disease [9]. The North American Symp-
tomatic Carotid Endarterectomy Trial (NASCET)
[10] and the European Carotid Surgery Trial
(ECST) [11] demonstrated significative outcome
improving in symptomatic patients with high-
grade (70–99 %) stenosis of the internal carotid
artery. A third large trial, the Asymptomatic
Carotid Atherosclerosis Study (ACAS) in 1995,
showed a more modest benefit of carotid surgical
treatment also in asymptomatic patients [12].
In 1977 Mathias reported the first percutaneous
transluminal angioplasty of the internal carotid
artery (Mathias 1977). In 1994 Marks reported
the treatment of spontaneous dissection of the
internal carotid artery using the Palmaz stent. In
the first period of use of carotid artery stenting
(CAS), the rate of neurologic intra- or post-
procedural neurologic complications was very
high so that the first randomized trial of carotid
artery stenting vs carotid endarterectomy was
stopped (Naylor 1998). Subsequently, technical
and material evolution, especially with the
R. Montisci and L. Saba
introduction of embolic protection devices,
improved the outcome of carotid stenting. In
2010, the CREST trial assessed that “among
patients with symptomatic or asymptomatic
carotid stenosis, the risk of the composite primary
outcome of stroke, myocardial infarction, or death
did not differ significantly in the group undergo-
ing carotid-artery stenting and the group undergo-
ing carotid endarterectomy” [13]. On this point,
however, the debate is still open, and many cen-
ters continue to reserve CAS to selected patients
with high risk for open surgery (hostile neck, post-
surgical restenosis, active coronary artery disease,
or congestive heart failure).
Preoperative Study
Vascular surgeons require from the vascular labo-
ratory a large series of answers. In particular, a
preoperative study of the patient with symptoms
of cerebral ischemia or suspected cerebral vascular
disease should be focused on identification of the
carotid plaque and its extension on the vessel,
evaluation of the myointimal thickness, the per-
centage of stenosis, the classification of the
plaque’s morphology, the localization of the carotid
artery and its bifurcation and its relationship with
other neck anatomical structures, the possible pres-
ence of multiple stenotic lesions (tandem stenosis)
both in extracranial and intracranial tracts, the pos-
sible presence of fresh intraluminal thrombus or
plaque ulceration, the status of vertebral arteries
and of the circle of Willis, and the presence of
cerebral ischemic lesions. These information can
bring to modify the surgical or anesthesiologic
strategy or also induce to avoid surgery.
Most of these information can be acquired with
duplex scan. According to the Society of Radiol-
ogists in Ultrasound Consensus statement, Dopp-
ler sonography is increasingly becoming the sole
imaging technique used for the evaluation for
carotid stenosis before surgery [14]. Ultrasound
evaluation localizes the common, internal, and
external carotid artery and identifies the level of
the bifurcation and the myointimal thickness. In
most of the cases, the presence of an atheroscle-
rotic plaque can be detected and the percentage of
11 Carotid Artery Surgery
Fig. 1 Schematic drawing that shows the measurements
for NASCET and ECST criteria
stenosis calculated. The main limit of this tech-
nique is represented by calcification of the arterial
wall that hinders the ultrasound passage produc-
ing an “umbra” image.
One of the most important topics in Duplex
scan examination is the calculation of the percent-
age of stenosis. This feature is extremely impor-
tant because it is crucial for the surgical decision.
However, there are various methods used for this
calculation. The two most used methods derive
from the NASCET and ECST trials (Fig. 1) and
are deeply different (Table 1). The NASCET and
ECST criteria were originally established using
contrast angiography. Afterward the same dimen-
sional criteria can be applied using ultrasound
assessment.
It is widely accepted that average Doppler
velocity rises in direct proportion to the degree
of stenosis, Therefore, the internal carotid artery
peak systolic velocity (PSV) and the end-diastolic
velocity (EDV) can be used to identify hemody-
namically significant carotid stenosis [15]. Maxi-
mal PSV and EDV are taken within the narrowing.
193
Table 1 Angiographic criteria
NASCET (A–B)
—————————  100 % stenosis
A
ECST (C–B)
—————————  100 % stenosis
C
Table 2 Hemodynamic criteria
NASCET > = 70 % stenosis PSV > 220 cm/s
EDV > 80 cm s
ECST > = 70 % stenosis PSV > 190 cm/s
EDV > 65 cm s
The cutoff for the hemodynamically significative
stenosis is expressed in Table 2. There are some
differences between various authors in the defini-
tion of the cutoff threshold, ranging between
220 and 283 for NASCET criteria [14, 15],
Heijenbrok-Kal (2005), [16].
It’s evident that, in a minor number of patients,
the alternative use of NASCET or ECST criteria
can produce a different surgical decision. NASCET
criteria are just based on the narrowing of the
hematic column in the carotid axis, and a stenosis
is detected when there is a reduction of this column
diameter compared to the distal internal carotid
artery. ECST criteria instead compare the maxi-
mum narrowing with the entire internal carotid
diameter at the same level. Paradoxically, consid-
ering the normal anatomical dilatation of the origin
of the internal carotid artery (carotid bulb), a plaque
that intersects just the dilated part of the bulb with-
out reducing the lumen with respect to the distal
internal carotid artery determines a 0 % stenosis
using NASCET criteria, while using ECST criteria
the stenosis is detected (Fig. 2). If we consider that
the risk of cerebral ischemia is consequent at the
embolic potential of the plaque, in some cases
NASCET criteria could underestimate the real sit-
uation. Due to the different systems of calculating,
the percentage of stenosis results generally higher
using ECST method.
In the last years the role of plaque’s morphol-
ogy is emerging in the evaluation of embolic risk
(Fig. 3). The assessment of a stenosis is just one of
the elements useful in the prediction of the clinical
194
Fig. 2 Different criteria
can produce different
surgical decision
Fig. 3 Macroscopic view of the carotid specimen showing
a typical feature of vulnerable plaque: the intra-plaque
hemorrhage (IPH)
outcome. Ouhlous [17] and Nandalur [18] in 2005
pointed out the relationship between the compo-
sition of the carotid plaque and the risk of cere-
brovascular ischemic events. Ohara in 2008
demonstrated that eccentric stenosis was associ-
ated with a significant increase of ipsilateral ische-
mic events compared with concentric stenosis.
The “vulnerable” carotid plaque is characterized
R. Montisci and L. Saba
by a higher tendency to rupture, potentially
resulting in embolization and consequent cerebral
ischemia [19].
Therefore, characterization of the plaque mor-
phology can be important for the surgical decision,
mostly in patients with borderline level of stenosis.
Ultrasound morphologic evaluation of the
carotid plaque identifies the various plaque compo-
nents. In particular, anechogenic plaques have a
higher component of soft tissue, such as lipid or
intra-plaque hemorrhage, while echogenic plaques
are composed mainly of fibrous tissue. The Gray-
Weale’s scale modified by Geroulakos can identify
five types of carotid plaques: type 1 (anechogenic
with echogenic fibrous cap), type 2 (predominantly
anechogenic but with echogenic areas representing
less than 25 % of the plaque), type 3 (predominantly
hyperechogenic but with anechogenic areas
representing less than 25 % of the plaque), type
4 (echogenic and homogeneous plaque), and type
5 (unclassified plaques reflecting calcified plaques
with areas of acoustic shadowing which hide the
deeper part of the arterial layers).
Another aspect of plaque morphology is
represented by the analysis of the plaque surface.
11 Carotid Artery Surgery
Fig. 4 Longitudinal US view of an ulcerated (white
arrow) carotid artery plaque at the level of carotid
bifurcation
A smooth surface indicates a plain interface
between the plaque and the lumen. The presence
of plaque surface irregularities is related with an
increased risk of ischemic events [20, 21]. But the
most relevant surface alteration is the ulceration,
defined as an intimal defect larger than 1 mm that
expose the necrotic core of the atheromatous plaque
[22]. The presence of plaque irregularities or ulcer-
ation (Fig. 4) exposes to major risk of embolization
due to possible platelet aggregation and thrombosis
or embolic release of the lipidic plaque core. There-
fore, the presence of a “vulnerable plaque” should
be carefully evaluated for surgical treatment also in
patients with noncritical stenosis.
Multi-detector computed tomography angiog-
raphy (MDCTA) is a reliable tool for the evalua-
tion of carotid atherosclerosis (Fig. 5). It’s
generally used after ultrasound analysis.
In addition to the calculation of the stenosis
percentage, it’s useful in the detection of plaque
irregularities and especially of ulceration with
diagnostic accuracy higher than US-ECD [23].
Moreover, measuring Hounsfield unit (HU) atten-
uation in the carotid plaque is useful to categorize
the plaque, also in case of calcification.
In alternative to CT, MR, with different acqui-
sition techniques, is useful to obtain information
about the percentage of stenosis and plaque
195
Fig. 5 MDCTA axial view of an ulceration involving the
ICA of the left side
characteristics. Extending the study to the
encephalic district, it’s possible to evaluate the
status of the circle of Willis. The presence of two
or more interruption in the circle of Willis appears
to be associated to high risk of intolerance to
cross-clamping [24] (Fig. 6).
CT and MR are also useful in the evaluation of
cerebral parenchyma. Sometimes also in clinically
asymptomatic patients, these examination can
reveal areas of previous ischemic cerebral lesions.
CT and MR provide the surgeon very important
information about overall anatomy of the area.
Carotid endarterectomy can be quite simple or
extremely difficult due to the wide variability of
anatomical disposition. In particular, a “high”
bifurcation with a longitudinal extended vulnera-
ble plaque in a deep internal carotid artery can
represent a very challenging struggle. Consider-
ing that carotid endarterectomy is a preventive
intervention, often performed in asymptomatic
patients, the availability of the maximum level
of information is extremely useful in order to
minimize the risk of intraprocedural
complications.
196
Fig. 6 MRA TOF (time-of-flight sequence) in the axial
and coronal views that show the circle of Willis configu-
ration (BA basilar artery, MCA middle cerebral artery, ACA
anterior cerebral artery, PCA posterior cerebral artery). In
this patient there is no A1 segment of the left ACA (yellow
Surgical Technique of Endarterectomy
The intervention begins with an incision of
10–12 cm along the anterior border of the
sternocleidomastoid muscle, centered at the level
of carotid bifurcation. Cutting the platysma, it
exposes the sternocleidomastoid muscle whose
anterior margin is freed and retracted posteriorly.
The vascular phase begins; hence, it is useful to
magnify in order to identify the thin nerve struc-
tures that with great variability cross and surround
the vascular structures and that must be mobilized
and resected. The internal jugular vein is found
and the common facial vein is isolated, ligated,
and divided. At this point the common carotid
artery is located by palpation. After cautious inci-
sion of the vascular sheath, the adventitial plan is
joined, and the common carotid artery is freed and
passed with a vascular loop. The opening of the
vascular sheath is progressively extended toward
the bifurcation that is widely exposed; then the
external carotid artery is isolated and encircled
R. Montisci and L. Saba
arrowhead) and both the PcoA (yellow open arrowheads).
Moreover, there is no flow in the left PCA. This patient
showed cross-clamping intolerance during the carotid end-
arterectomy procedure requiring the shunt positioning dur-
ing the surgery
with a vascular loop. Sometimes the superior thy-
roid artery – normally the first branch of the exter-
nal carotid artery – originates directly from the
common carotid artery at the bifurcation level and
needs to be isolated and passed with a loop. With
gentle traction on the external carotid artery, the
internal carotid is exposed and can be freed along
the adventitial plan, avoiding damage to the ner-
vous structures (Fig. 7) that lie near this vessel
(in particular the hypoglossal nerve with its
descending branch and the vagus nerve). The
internal carotid artery needs to be prepared exten-
sively until the end of the carotid plaque, with at
least 1 cm of healthy distal carotid for a safe
clamping.
In most of the cases, the end of the plaque can
be assessed both with visualization and gentle
palpation. Preoperative imaging can be very help-
ful on this point, in order to avoid hazardous
handling of the carotid plaque. Also the internal
carotid artery is passed with a vascular loop. In
case of “high” bifurcation, with a plaque very
longitudinally extended, the exposure of the
11 Carotid Artery Surgery
Fig. 7 Surgical exposition
of the carotid bifurcation
during the carotid
endarterectomy procedure.
The CCA, ICA, and ECA
are visible together with the
hypoglossal nerve (white
arrowheads), the ansa
cervicalis (white open
arrow), and the transverse
cervical nerve (white
arrow)
internal carotid artery can require complex
maneuvers like mobilization of the hypoglossal
nerve, division of the digastric and stylohyoid
muscle, retraction of the parotid gland, subluxa-
tion of the mandibular condyle, and others skill
that can require a multidisciplinary approach. For-
tunately, these cases are very rare. The exposition
concluded that the patient is heparinized.
At this point, the internal, external, and com-
mon carotid arteries are sequentially clamped.
After cross-clamping, it’s mandatory to acquire
information about tolerance to carotid cross-
clamping. In fact, about 10–20 % of the patients
don’t tolerate the interruption of the flow and
show symptoms of cerebral ischemia. In the
patient operated under local anesthesia, symptoms
are quickly detected: in most of the cases there is
loss of consciousness, mental confusion, and
inability to squeeze a rattle in his or her contralat-
eral hand. In the patient operated under general
anesthesia, information about cross-clamping
intolerance is obtained with continuous
intraoperative electroencephalographic monitor-
ing or with somatosensory-evoked potential mon-
itoring. Information about intracranial circulation
can also be obtained both intraoperatively via
transcranial Doppler [25] and with near-infrared
spectroscopy [26].
The onset of intolerance of carotid cross-
clamping requires immediate positioning of a
temporary shunt in order to ensure an adequate
197
vascular flow avoiding cerebral damages. In that
case, after a wide arteriotomy that extended from
the distal common carotid artery up to the end of
the plaque in the internal carotid artery, the shunt
(usually the Pruitt-Inahara shunt or the Javid
shunt) is inserted – previous rapid removing of
the clamp – before in the internal carotid artery
and then in the common carotid artery. Both ends
of the shunt are stabilized (the Pruitt-Inahara shunt
inflating the balloons near the extremities, the
Javid shunt with tourniquets). In most of the
cases, the shunt positioning is followed by rapid
restoring of the cerebral functions, easily recog-
nizable with clinical (local anesthesia) or instru-
mental (general anesthesia) signs.
After cross-clamping, a longitudinal
arteriotomy is performed on the posterolateral
distal common carotid artery and extended toward
the internal carotid artery. For direct endarterec-
tomy the extension of the arteriotomy regards the
entire length of the plaque. With the use of a
dissector, the surgeon identifies the plan between
the atherosclerotic plaque and the residual arterial
wall. This plan can be found between the media
and the external elastic lamina or between the
intima and the internal elastic lamina. In my expe-
rience, the media is often involved, especially in
calcific plaques. In most of the cases, the plan of
dissection of the plaque is easily identifiable. The
plaque is circumferentially detached before in the
common carotid artery’s side and, then,
198
continuing the dissection toward the external
carotid artery and lastly toward the internal carotid
artery. Distally, the plaque usually separates from
the normal intima but in some cases can be nec-
essary to incise the distal intima. However, once
the plaque has been removed, an accurate check of
the end point, i.e., the transition line between the
endarterectomized area and the distal normal
intima is mandatory. In fact, if the distal intima
is partially separated from the underlying plan,
restoring the blood flow can totally dissect the
distal intima at the downstream edge, causing
acute occlusion of the internal carotid artery. In
case of moderate instability of the distal intima,
this can be stabilized with some stitches (Kunlin
stitches), ensuring firm adhesion to the medio-
adventitial plan. The internal surface of the inter-
nal wall is then irrigated with heparinized saline
solution, and any debris or microflap floating is
removed. All these phases of the endarterectomy
can be performed also when a shunt has been
inserted, although this result can be quite
cumbersome.
The suture begins at the internal carotid artery
and runs down toward the center of the
arteriotomy. A second suture begins at the com-
mon carotid artery toward the first suture. When
the two sutures meet, before tying, a backflow
from the internal carotid artery is flushed out via
arteriotomy in order to purge air bubbles and
microdebris from the lumen. After closure, we
prefer to re-clamp the internal carotid artery at its
origin and declamp at the same time the common
and the external carotid artery for about 10 s,
sending any possible microdebris toward the
external carotid artery’s district. Then the internal
carotid clamp is removed. Just when the circum-
ference of the internal carotid artery is small, the
arteriotomy can be closed using a patch (autoge-
nous vein or polyester or polyurethane or bovine
pericardium patch) in order to avoid a further
reduction of the lumen. Generally it is not
necessary.
The regular flow can be assessed via
intraoperative ultrasonography.
Restoring the flow, careful hemostasis is
achieved. If necessary, heparinization can be
reversed with protamine infusion. A 4  10
R. Montisci and L. Saba
Jackson-Pratt drain is placed near the arterial
plan and brought out with a little incision. The
sternocleidomastoid muscle is accosted to the
median cervical fascia and the platysma is
reconstructed. The cutaneous plan is sutured
with staples. Generally, the drain is removed in
the first post-op day.
As alternative to direct endarterectomy, the
eversion endarterectomy requires the transection
of the internal carotid artery at its origin. Then the
plaque is grasped and the vessel wall is progres-
sively peeled back and everted. The plaque is
completely removed and after meticulous
checking of the “end point,” the internal carotid
artery is reimplanted at the bifurcation with a
circular anastomosis. This technique is used
often in case of stenosis associated to kinking or
coiling of the internal carotid artery: this tech-
nique allows a simple elimination of the redun-
dancy of the vessel that is shortened before the
reimplantation.
Carotid Body Tumors
Carotid body tumors are rare, generally benign,
and occasionally malignant (5–7 %) tumors
(paraganglioma) of the carotid body, a little
organ highly vascularized and placed in the saddle
between external and internal carotid arteries. Its
frequency is about 1 in 30,000 [27, 28]. The tumor
presents as painless neck mass under the mandib-
ular angle. Color Doppler sonography can reveal
the origin of the mass from the carotid body tumor
due to the splaying of the bifurcation and
hypervascularity. MDCTA shows the relationship
between the mass and the surrounding tissues and
consents to classify the tumor according to the
classical Shamblin’s classification (Fig. 8):
type I, tumor that grows splaying the two arteries
(“lyre sign”) (Fig. 9) but without close contact
with them; type II, tumor that begins to incorpo-
rate the two vessels but not completely (Fig. 10);
and type III, tumor that has incorporated one or
both the vessels, sometimes together with nerve
structures.
This classification is extremely important in
the evaluation of the surgical risk. The contrast
11 Carotid Artery Surgery 199

Fig. 8 Schematic drawing ICA

of the Shamblin ECA
N,X
classification. See the other sup, laryn. n.
details in the text

N,XII
N,XII

sup.
laryngeal
nerve

N,x -

TYPE I TYPE II TYPE III

tumor presents a very rich vascularization, and

until the main peduncle is ligated, the risk of
blood loss is high. OctreoScan scintigraphy is
useful to assess the neuroendocrine character of
the mass and to localize any multicenter or meta-
static lesion. Selective angiography is possible to
treat preoperatively the mass with ultraselective
embolization of the feeding vessels.
Surgical excision is the treatment of choice.
Radiation therapy or embolization is considered
for patients that presents a high risk for surgery
both for local or general conditions [29].

Aneurysms of the Extracranial

Fig. 9 DSA view of a case with Shamblin type 1 tumor
(“lyre sign”): the tumor grows splaying the ICA and ECA
enhancement is useful in differentiation from
other mass and, in the very early phase, can help
to identify the main vascular peduncle. This infor-
mation can simplify the surgical dissection. The
Carotid Artery
They are rare and more often of atherosclerotic
origin than traumatic (also iatrogenic). Other very
rare causes are infections and fibromuscular dys-
plasia [30, 31]. The clinical main sign for common
carotid artery aneurysm is a laterocervical
200
Fig. 10 CTA shows a Shamblin type 2 tumor (white
arrow); the tumor shows partial encasement of the ICA
and ECA (white arrowhead and white arrow, respectively)
pulsatile mass, while internal carotid artery aneu-
rysms often present in the pharynx. Nerve com-
pression of branches of cranial nerves can cause
pain (V, VII) or paralysis (V, VI, VII, IX, X, XI,
XII). Thromboembolism from the aneurysm can
cause transient ischemic attacks or stroke. Rup-
ture is a rare complication [32].
In case of very high localization, Duplex scan
can fail to identify the aneurysm. MDCTA is the
gold standard, identifying the aneurysm, its exten-
sion and relation with the surrounding structures,
the presence of intraluminal thrombus or dissec-
tion, and all information very useful for surgical
planning.
Whereas at the end of the twentieth century the
aneurysms involving the internal carotid artery
above the line between the angle of the mandible
and the tip of the mastoid process (Blaisdell line)
were considered inaccessible [32], at present the
development of reconstructive vascular surgery
with multidisciplinary approach allows surgical
treatment of these lesions, also adjoining the base
of the skull [33]. Considering the complexity of the
open surgical treatment of such cases that require a
very extensive and invasive exposure of the base of
the skull, with risks for nerve integrity, the
endovascular approach with covered stenting is
R. Montisci and L. Saba
conquering an increasing popularity [34, 35, 36].
However, though an anticoagulation therapy,
about 6 % of thrombosis and occlusion of the
stent occur.
Conclusion
Nowadays the carotid endartectomy can be con-
sidered as an excellent modality to eliminate the
carotid artery plaque and to treat other pathologi-
cal conditions involving the carotid arteries.
References
1. DeBakey ME (1975) Successful carotid endarterec-
tomy for cerebrovascular insufficiency. JAMA
233:1083.rG
2. Lin P, Javid H, Doyle E (1956) Partial internal carotid
artery occlusion treated by primary resection and vein
graft. J Neurosurg 13:650
3. Lyons C, Galbraith G (1957) Surgical treatment of
atherosclerotic occlusion of the internal carotid artery.
Ann Surg 146:487
4. Cooley D, Al-Naaman Y, Carton C (1956) Surgical
treatment of arteriosclerotic occlusion of common
carotid artery. J Neurosurg 13:500
5. Moore O, Karlan M, Sigler L (1969) Factors
influencing the safety of carotid ligation. Am J Surg
118:666
6. Callow A, David M (1980) Hume memorial lecture, an
overview of the stroke problem in carotid territory. Am
J Surg 140:181
7. Marrocco-Trischitta MM, Bandiera G, Camilli S,
Sillo F, Cirielli C, Guerrini P (2001) Remifentanil
conscious sedation during regional anaesthesia for
carotid endarterectomy: rationale and safety. Eur J
Vasc Endovasc Surg 22:405
8. Marcucci G, Siani A, Accrocca F, Gabrielli R,
Giordano A, Antonelli R et al (2011) Preserved con-
sciousness in general anesthesia during carotid endar-
terectomy: a six-year experience. Interact Cardiovasc
Thorac Surg 13:601
9. Robicsek F, Roush TS, Cook JW, Reames MK (2004)
From Hippocrates to Palmaz Schatz, the history of
carotid surgery. Eur J Vasc Endovasc Surg 27:389
10. The North American Symptomatic Carotid Endarter-
ectomy Trial Collaborators (1991) Beneficial effect of
carotid endarterectomy in symptomatic patients with
high grade stenosis. N Engl J Med 325:445
11. European Carotid Surgery Trialists’s Collaborative
Group (1998) Randomised trial of endarterectomy for
recently symptomatic carotid stenosis: final result of
the MRC European Carotid Surgery Trial (ECST).
Lancet 351:1379
11 Carotid Artery Surgery
12. Executive Committee for the Asymptomatic Carotid
Atherosclerosis Study (1995) Endarterectomy for
asymptomatic carotid artery stenosis. JAMA 273:1421
13. Brott TG, Hobson RW 2nd, Howard G, Roubin GS,
Clark WM, Brooks W, Mackey A, Hill MD,
Leimgruber PP, Sheffet AJ, Howard VJ, Moore WS,
Voeks JH, Hopkins LN, Cutlip DE, Cohen DJ, Popma
JJ, Ferguson RD, Cohen SN, Blackshear JL, Silver
FL, Mohr JP, Lal BK, Meschia JF, CREST Investiga-
tors (2010) Stenting versus endarterectomy for treat-
ment of carotid-artery stenosis. N Engl J Med
363(1):11–23
14. Grant EG, Benson CB, Moneta GL et al (2003) Carotid
artery stenosis: gray-scale and Doppler US diagnosis –
society of radiologists in ultrasound consensus confer-
ence. Radiology 229:340
15. Staikov IN, Nedeltchev K, Arnold M, Remonda L,
Schroth G, Sturzenegger M, Herrmann C, Rivoir A,
Mattle HP (2002) Duplex sonographic criteria for mea-
suring carotid stenoses. J Clin Ultrasound 30:275
16. Saba L, Sanfilippo M, Montisci R, Mallarini G (2010)
Correlation between US-PSV and MDCTA in the
quantification of carotid artery stenosis. Eur J Radiol
74:99
17. Ouhlous M, Flach HZ, de Weert TT et al (2005) Carotid
plaque composition and cerebral infarction: MR imag-
ing study. AJNR Am J Neuroradiol 26:1044
18. Nandalur KR, Baslurt E, Hagspiel K, Douglas
Philips C, Kramer CM (2005) Calcified carotid athero-
sclerotic plaque is associated less with ischemic symp-
toms than is non calcified plaque on MDTC. Am J
Roentgenol 184:295
19. Saba L, Sanfilippo R, Montisci R, Atzeni M,
Ribuffo D, Mallarini G (2011) Vulnerable plaque:
detection of agreement between multidetector-row CT
angiography and US-ECD. Eur J Radiol 77:509
20. Kitamura A, Iso H, Imano H et al (2004) Carotid
intima-media thickness and plaque characteristics as a
risk factor for stroke in Japanese elderly men. Stroke
35:2788
21. Prabhakaran S, Rundek T, Ramas R et al (2006)
Carotid plaque surface irregularity predicts ischemic
stroke: the northern Manhattan study. Stroke 37:2696
22. Sitzer M, Muller W, Siebler M et al (1995) Plaque
ulceration and lumen thrombus are the main sources
of cerebral microemboli in high-grade internal carotid
artery stenosis. Stroke 26:1231
23. Saba L, Caddeo G, Sanfilippo R, Montisci R, Mallarini
G (2007) CT and US in the study of ulcerated carotid
plaque compared with surgical results. Advantages of
201
multi-detector –row CT angiography. AJNR Am J
Neuroradiol 28:1061
24. Montisci R, Sanfilippo R, Bura R, Branca C, Piga M,
Saba L (2013) Status of the circle of Willis and intol-
erance to carotid cross-clamping during carotid endar-
terectomy. Eur J Vasc Endovasc Surg 45:107
25. Stejskal L, Kramar F, Ostry S, Benes V, Mohapl M,
Limberk B (2007) Experience of 500 cases of neuro-
physiological monitoring in carotid endarterectomy.
Acta Neurochir 149:681
26. Pennekamp CWA, Immink RV, den Ruijter HM,
Kappelle LJ, Bots ML, Buhre WF, Moll FL, de Borst
GJ (2013) Near-infrared spectroscopy to indicate selec-
tive shunt use during carotid endarterectomy. Eur J
Vasc Endovasc Surg 46:397
27. Sajid MS, Hamilton G, Baker DM (2007) Joint vascu-
lar research group. A multicenter review of carotid
body tumour management. Eur J Vasc Endovasc Surg
34(2):127–30
28. Ma D, Liu M, Yang H, Ma X, Zhang C (2010) Diag-
nosis and surgical treatment of carotid body tumor: a
report of 18 cases. J Cardiovasc Dis Res 1:122
29. Davidovic LB, Djukic VB, Vasic DM, Sindjelic RP,
Duvnjak SN (2005) Diagnosis and treatment of carotid
body paraganglioma: 21 years of experience at a clin-
ical center of Serbia. World J Surg Oncol 3:10
30. Schechter DC (1979) Cervical carotid aneurysm. N Y
State J Med 79:892
31. McCann RL (1990) Basic data related to peripheral
artery aneurysms. Ann Vasc Surg 4:411
32. Stoney RJ, Qvarfordt PG (1987) Accessible and inac-
cessible aneurysm of the extracranial carotid artery. In:
Moore WS (ed) Surgery for cerebrovascular disease.
Churchill Livingstone, New York
33. Rosset E, Albertini JN, Magnan PE, Ede B, Thomassin
JM, Branchereau A (2000) Surgical treatment of extra-
cranial internal carotid artery aneurysms. J Vasc Surg
31:713
34. Li Z, Chang G, Yao C, Guo L, Liu Y, Wang M, Liu D,
Wang S (2011) Endovascular stenting of extracranial
carotid artery aneurysm: a systematic review. Eur J
Vasc Endovasc Surg 42(4):419–26
35. Pickering GW, Rob CG (1954) Reconstruction of inter-
nal carotid artery in a patient with intermittent attacks
of hemiplegia. Lancet 264:994
36. Ohara T, Toyoda K, Otsubo R et al (2008) Eccentric
stenosis of the carotid artery associated with ipsilateral
cerebrovascular events. AJNR Am J Neuroradiol
29:1200
 Part III
The Basics of Intracranial Arterial
Circulation
 Intracranial Atherosclerosis
12
Xinyi Leng and David S. Liebeskind

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206 Atherosclerotic and arteriolosclerotic diseases
Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206 of the cervicocerebral arteries are prevalent
Pathology and Pathogenesis . . . . . . . . . . . . . . . . . . . . . . . . . 206 across the world, respectively, affecting arter-
Recurrent Risk in Symptomatic ICAS and Related ies from large to small and distal arterioles.
Risk Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Research Gaps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207 This chapter is focused on atherosclerotic dis-
ease that affects large intracranial arteries, an
Neuroimaging Methods for Diagnosing
and Evaluating Symptomatic ICAS . . . . . . . . . . . . . . . 208
important cause of ischemic stroke and tran-
Introduction on Routine and Novel Imaging sient ischemic attack worldwide that requires
Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208 further understanding. An introduction covers
Routine Neuroimaging Methods to Evaluate the epidemiology and pathology of intracranial
ICAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
Novel Neuroimaging Methods to Evaluate
atherosclerosis (ICAS), including both asymp-
ICAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218 tomatic and symptomatic manifestations, and
the recurrent risk of symptomatic lesions. This
Treatment Strategies for Patients with ICAS . . . . 227
is followed by a detailed discussion on routine
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 and novel imaging methods to evaluate the
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230 different aspects of symptomatic ICAS and
their roles in stratifying affected patients,
followed by a brief summary of recommenda-
tions and evidence for treatment of such
lesions based on current guidelines.
Imaging methods for diagnosing and eval-
uating symptomatic ICAS reviewed in this
chapter include noninvasive and invasive
methods to gauge the severity of luminal ste-
X. Leng nosis and direct and indirect methods to define
Department of Medicine and Therapeutics, The Chinese the extent and function of collateral circulation.
University of Hong Kong, Prince of Wales Hospital, We also reviewed routine and novel methods to
Hong Kong, SAR, China
reveal the status of the overall and territorial
e-mail: lengxinyi@gmail.com
tissue perfusion and to detect the salvageable
D.S. Liebeskind (*)
penumbra. Moreover, there is a discussion on
UCLA Department of Neurology, UCLA, Los Angeles,
CA, USA the emerging method of plaque imaging to
e-mail: davidliebeskind@yahoo.com delineate plaque morphology and constituents
# Springer Science+Business Media New York 2016 205
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_39
206
and rising methods to directly and indirectly
quantify the hemodynamic effects of such
lesions. For the particularly high risk of recur-
rent stroke in symptomatic ICAS, not only the
severity of luminal stenosis has been identified
as an independent predictor, other characteris-
tics of the lesion revealed by imaging methods
as discussed in this chapter may also affect the
recurrent risk. Appropriate application and
interpretation of these imaging methods in
cases suspected of symptomatic ICAS may
facilitate reasonable clinical decision making
for acute treatment and secondary prevention
of such lesions.
Keywords
Intracranial atherosclerosis • Intracranial arte-
rial stenosis • Symptomatic • Asymptomatic •
Cerebrovascular disease • Neuroimaging •
Ischemic stroke • Transient ischemic attack •
Recurrence • Risk stratification • Collateral
circulation • Perfusion imaging • Plaque imag-
ing • Fractional flow • Cerebral hemodynamics
Introduction
Epidemiology
Intracranial atherosclerosis (ICAS) is of high prev-
alence throughout the world, particularly in Asian
populations. Symptomatic ICAS has been identi-
fied as the leading cause for ischemic stroke and
transient ischemic attack (TIA) in Asian
populations, which accounts for around 30–50 %
of ischemic stroke and around 50 % of TIA in
Asians. ICAS is also predominant in stroke patients
with Hispanic and African origins. Although it is
relatively uncommon in Caucasians, ICAS has
been found responsible for around 10 % of ische-
mic stroke and TIA in them. As for asymptomatic
ICAS, data are relatively scarce. Based on limited
data, asymptomatic ICAS affects about 7–13 % of
community-dwelling Chinese aged over 40 years
and those without histories of stroke or TIA, while
it has been detected in 13 % of predominantly
Caucasian patients referred for carotid Doppler in
a clinic-based [1].
X. Leng and D.S. Liebeskind
As summarized in a previous review article,
risk factors associated with ICAS can be catego-
rized as modifiable, nonmodifiable, and less well-
documented risk factors [1]. Among all the
nonmodifiable risk factors, age, sex, and race are
probably the most widely established. Based on
previous studies, there are several well-
documented modifiable risk factors for ICAS,
including smoking, hypertension, dyslipidemia,
diabetes, and metabolic syndrome, which are
among the main target risk factors to be controlled
in secondary prevention of stroke by current
guidelines [2, 3]. There are also some less well-
documented factors that might be correlated to
ICAS, such as atherosclerosis of aortic, coronary
and carotid arteries, head and neck radiotherapy,
and Alzheimer disease, which mandate further
investigation [1].
Pathology and Pathogenesis
Atherosclerosis is a systemic disorder, involving
the coronary, carotid, intracranial, renal, and
peripheral arteries. Atherosclerosis progresses
with increasing age, which is a complex interaction
affected by variable risk factors, mediated by cel-
lular, molecular, and hemodynamic factors [4, 5].
In systemic atherosclerosis, atherogenesis of intra-
cranial arteries may be different between
populations of different racial origins. Previous
studies carried out in Asian populations have
found that carotid atherosclerosis might be more
resembling to that of coronary atherosclerosis but
not ICAS. The correlation between ICAS and
carotid atherosclerosis in Caucasians may be stron-
ger than that in Asians. Intracranial atherosclerosis
probably occurs earlier than carotid atherosclerosis
in Asians, while on the contrary, ICAS may be a
more severe process than carotid atherosclerosis in
Caucasians [6].
Pathology reveals that fatty streaks are the
earliest visible atherosclerotic lesions, which
later progress to fibrous plaques and then
advanced or complicated plaques [7]. Progression
of atherosclerotic plaques could lead to the
narrowing of the vessel lumen, which may cause
turbulence or slow flow, in turn further promoting
12 Intracranial Atherosclerosis
atherosclerosis by activating platelets and other
pathways. Complicated plaque constituents and
focal hemodynamic and cellular factors may also
alter the vulnerability of plaques [5]. Although
relevant data and studies are limited due to the
difficulty in performing biopsy and autopsy stud-
ies of intracranial arteries, investigating patholog-
ical characteristics of ICAS lesions will facilitate
understanding of related mechanisms of stroke,
such as acute thrombosis, hypoperfusion, and
artery-to-artery embolization, which needs to be
addressed in future studies [8].
Recurrent Risk in Symptomatic ICAS
and Related Risk Factors
Among all principal causes of ischemic stroke or
TIA, including intracranial and extracranial large-
artery disease, cardioembolism, and small vessel
disease, those due to ICAS are at particularly high
risk of recurrence, despite medical and/or inter-
ventional treatment and risk factor modification
[9–11]. Until now, there have been several large
clinical trials focusing on ischemic stroke or TIA
patients with symptomatic ICAS, such as the
Warfarin-Aspirin Symptomatic Intracranial Dis-
ease (WASID) trial and the Stenting and Aggres-
sive Medical Management for Preventing
Recurrent Stroke in Intracranial Stenosis
(SAMMPRIS) trial that were performed predom-
inantly in Caucasians and the Trial of cilOstazol in
Symptomatic intracranial arterial Stenosis
(TOSS) and TOSS 2 performed in Asians
[9–12]. According to the results from these clini-
cal trials, the risk of stroke recurrence in ischemic
stroke or TIA patients with a symptomatic ICAS
of 50–99 % luminal stenosis could be as high as
15–20 % at 1 year after ictus, with more than half
occurring in the territory of the index stenotic
artery [9, 11, 12].
The recurrent stroke risk of symptomatic ICAS is
related to several factors, including nonmodifiable
and modifiable demographic and clinical factors, as
well as imaging manifestations directly and indi-
rectly related to the ICAS lesion. Age has been
established as a predictor for stroke recurrent
in this patient subset in several studies [13].
207
As for modifiable factors, elevated blood pressure
and cholesterol levels, and the presence of metabolic
syndrome and diabetes, have been found to increase
the risk of stroke recurrence and other major vascu-
lar events in patients with symptomatic ICAS [2].
Furthermore, the severity of presenting stroke or
ischemia, number of stenotic cervicocerebral arter-
ies, and number of ischemic lesions at baseline may
also be associated with recurrent stroke in this
patient subset.
Moreover, characteristics of the ICAS lesion
as revealed by different imaging modalities and
methods could also affect subsequent risk of
stroke recurrence, which yield potential indica-
tors for risk stratification of symptomatic ICAS.
First of all, the severity of luminal stenosis or the
percent stenosis in ICAS has been established as
an independent predictor for stroke recurrence in
ischemic stroke or TIA due to ICAS, which has
been widely used to define a “severe” or “mod-
erate” ICAS lesion from both the clinical and
imaging points of view [2, 14]. However, during
the past few years, increasing evidence has
emerged, challenging the role of percent luminal
stenosis in predicting the risk of stroke recur-
rence in symptomatic ICAS. For instance, factors
including but not limited to collateral status,
plaque morphology and components, perfusion
status, and hemodynamic characteristics may
also affect the recurrent risk of symptomatic
ICAS [15]. Considering the high risk of recur-
rence, appropriate evaluation and stratification of
symptomatic ICAS is required for accurate clin-
ical decision making. Routine and novel imaging
methods for the evaluation of such lesions are
discussed in details hereinafter.
Research Gaps
As an important cause of ischemic stroke or TIA,
with high risk of stroke recurrence worldwide,
symptomatic ICAS has not been adequately
investigated to date. Concerning the imaging eval-
uation of symptomatic ICAS, the percentage of
luminal stenosis has been designated as almost the
only indicator for scaling the severity of such
lesions for years. Increasing evidence supports
208
the potential clinical relevance of collateral circu-
lation, perfusion status, plaque characteristics,
and other features of the lesion in stratifying this
patient subset and guiding clinical decisions.
However, most of these findings have not been
well established, partly due to the lack of a more
wide-scale attention on this particular disorder,
the misleading focus on the severity of luminal
stenosis for diagnosis of this disorder, and proba-
bly the heterogeneity of imaging methods to eval-
uate particular aspects of such lesions. Therefore,
broad consensus is in great demand regarding a
practical paradigm for the assessment of selected
aspects of symptomatic ICAS and comprehensive
interpretation of relevant imaging findings. Only
in this way the ultimate values of different aspects
of such lesions in stratifying affected patients and
guiding clinical decisions could be systemically
investigated in future studies.
Neuroimaging Methods
for Diagnosing and Evaluating
Symptomatic ICAS
Introduction on Routine and Novel
Imaging Methods
Considering the high risk of stroke recurrence in
symptomatic ICAS, accurate and timely risk strat-
ification of the lesions is crucial for the sake of
reducing the recurrent risk and consequent dis-
ability and mortality, which requires a reasonable
application of imaging methods to diagnose and
evaluate ICAS and comprehensive interpretation
of the imaging results. Currently, there are several
routine imaging methods to evaluate the severity
of luminal stenosis, cerebral blood flow, and col-
lateral and perfusion status in the case of ICAS.
There are also some novel methods to more accu-
rately or less invasively delineate different aspects
of ICAS, including the above aspects that could be
explored by routine modalities as well, and also
other applications that could not be achieved by
using routine methods only such as plaque imag-
ing and selective perfusion territory visualization.
Routine imaging methods most commonly used
in clinical examination of ICAS, and novel or
X. Leng and D.S. Liebeskind
rising imaging methods to assess ICAS in clinical
scenarios or the research area, is reviewed below
in this chapter.
Routine Neuroimaging Methods
to Evaluate ICAS
Currently, several neuroimaging methods, includ-
ing routine and novel methods, are used in clinical
practice and the research area to evaluate ICAS
lesions. Common applications of routine imaging
methods in the evaluation of ICAS lesions are as
follows: transcranial Doppler imaging (TCDI) for
diagnosing the severity of stenosis mainly based
on blood flow velocity, determining flow direc-
tion, evaluating cerebral autoregulation, detecting
microemboli, and monitoring recanalization dur-
ing reperfusion therapy, of which the diagnostic
ability has been enhanced by the application of
transcranial color-coded duplex (TCCD) in recent
years [16]; different techniques of magnetic reso-
nance angiography (MRA) for diagnosing the
degree of stenosis, assessing the collaterals, and
quantifying flow velocities and flow volumes
[17]; computed tomography angiography (CTA)
for diagnosing the percentage of luminal stenosis,
detecting primary and secondary collaterals, and
assessing the thrombus burden [17, 18]; digital
subtraction angiography (DSA), commonly
regarded as the gold standard for quantifying the
luminal stenosis and the collateral status [14, 18];
and CT perfusion (CTP) and dynamic susceptibil-
ity contrast perfusion-weighted MR imaging
(DSC-PWI) for evaluating perfusion and collat-
eral status and for determining salvageable tissues
for reperfusion therapies in acute cases. Each of
the routine imaging methods has its unique advan-
tages and limitations, which are discussed in
details below.
TCDI/TCCD
TCDI, as a safe and noninvasive method to diag-
nose ICAS, has become widely available through-
out the world during the past three decades
[16]. Although the accuracy of TCDI to diagnose
luminal stenosis in ICAS was reported to be infe-
rior to that of MRA or CTA, its unique advantage
12 Intracranial Atherosclerosis
lies in its ability to provide real-time flow infor-
mation, reveal blood flow velocities at different
vessel segments, evaluate cerebral autoregulation,
detect microembolic signals (MES), and provide
evidence for flow direction, collateral compensa-
tion, and steal phenomenon. And by combining
with carotid duplex ultrasound, it could more
accurately depict the cervicocerebral vasculature
and reveal the hemodynamic impact of associated
arterial lesions [14, 16].
According to the Stroke Outcomes and Neuro-
imaging of Intracranial Atherosclerosis (SONIA)
trial, the largest prospective study so far to system-
ically examine the diagnostic accuracies of nonin-
vasive imaging methods TCDI, MRA, and CTA to
scale the severity of luminal stenosis in ICAS as
compared with DSA, TCDI could reliably rule out
50 % luminal stenosis, with negative predictive
values (NPV) of higher than 80 % for both the
prospectively tested cut points and the cut points
modified to maximize the positive predictive value
(PPV). However, abnormal findings in intracranial
arteries suggested by TCDI with regard to luminal
stenosis (Fig. 1) may require further confirmation
by other imaging methods, since the PPV of TCDI
was found to be lower than 50 % [14].
Besides the most common application of TCDI
in grading luminal stenosis of ICAS, it could also
be used in the following manners. Quantified
vasomotor reactivity by TCDI could represent
the status of cerebral autoregulation, which prob-
ably is globally impaired in patients with symp-
tomatic ICAS, and also a risk factor for stroke
[16]. MES revealed by TCDI supports the theory
that artery-to-artery embolization may be a poten-
tial mechanism of ICAS-related ischemic stroke,
and it has been found to be related to the worsen-
ing of neurological deficits in acute ischemic
stroke and also independently correlated to recur-
rent events in ischemic stroke due to ICAS [16]. In
addition, recent studies suggested that the steal
phenomenon termed “reversed Robin Hood syn-
drome” revealed by TCDI, referring to arterial
blood flow stolen from ischemic to non-affected
brain tissues, could be a possible mechanism for
an early deterioration of acute ischemic stroke and
correlated with the risk of stroke recurrence inde-
pendent of stroke subtypes [19].
209
In recent years, TCCD (Fig. 2) has become a
standard diagnostic technique to evaluate intracra-
nial arteries in ischemic stroke patients at some
centers and has been increasingly used in guiding
thrombolysis of acute stroke [20]. The combina-
tion of flow and anatomic information of intracra-
nial arteries and adjacent structures enhanced the
diagnostic accuracy of the severity of luminal
stenosis in ICAS, as compared with routine
TCDI. Moreover, the use of contrast in TCCD
yields higher diagnostic value than TCDI in
patients with insufficient temporal acoustic
window [20].
Real-time flow monitoring by TCDI or TCCD
yields an economic and effective method for
detecting the recanalization of occluded arteries
in acute ischemic stroke (Fig. 3) and is possibly
related to stroke severity and mortality [16]. In the
past few years, TCDI and TCCD have also been
used to enhance the thrombolytic activity of intra-
venous tissue plasminogen activator (tPA), which
was found to be efficient and safe. As compared
with tPA alone, tPA plus high-frequency TCDI/
TCCD may correlate with higher rates of com-
plete recanalization, without increasing the risk of
symptomatic intracranial hemorrhage [21].
TCDI, as a noninvasive and relatively inexpen-
sive imaging method, is widely available through-
out the world. With its unique advantages as
mentioned above, TCDI could be a useful tool to
evaluate hemodynamic characteristics of symp-
tomatic ICAS, and some of the characteristics
could be potential predictors for stroke severity
and the risk of recurrence in patients with symp-
tomatic ICAS. However, TCDI is a highly
operator-dependent examination; inconsistency
of experience and skills of different operators
may impact on its diagnostic accuracy. Also,
insufficiency of temporal windows, especially in
elder females, leads to underestimation of dis-
eased intracranial arteries [16].
MRA
Several different techniques or sequences of
MRA, a noninvasive imaging modality, have
been applied in clinical practice to evaluate the
cervicocerebral vasculature during the past
decades [22]. Among all these techniques,
210 X. Leng and D.S. Liebeskind

a 30 30
40 40
50 50
mm

mm
60 60
70 70
80 80

100 140
80 120
60 100
cm/s

cm/s
40 80
20 60
0 40
–20 20

b 60 60

70 70

80 80
mm

mm
90 90

100 100

110 110

–160 –220
–140 –200
–120 –180
–100 –160
–80 –140

cm/s
cm/s

–60 –120
–40 –100
–20 –80
0 –60
20 –40
40 –20
60 0

c 60 60

70 70

80 80
mm
mm

90 90

100 100

110 110

–100
–80
–90
–70
–80
–60
–70
–50
cm/s

–60
–40
cm/s

–50
–30
–40
–20
–30
–10
–20
0
–10
10
0
20
10

Fig. 1 Intracranial atherosclerotic stenoses diagnosed by
transcranial Doppler imaging (TCDI) and digital subtrac-
tion angiography (DSA). (a) DSA (center), 58 % stenosis
of the M1 segment of the middle cerebral artery (MCA);
TCDI, normal mean flow velocity (MFV) in the MCA at a
depth of 60 mm (insert on the left) and high MFV in the
mid-MCA at a depth of 49 mm (insert on the right),
indicating a 50 % MCA stenosis. (SPR in the figure
indicates stenotic-to-prestenotic ratio). (b) DSA (center),
71 % mid-basilar artery (BA) stenosis; TCDI, normal MFV
in the BA at a depth of 85 mm (insert on the left) and high
MFV in the BA at a depth of 94 mm (insert on the right),
indicating a high-grade (70 %) mid-BA stenosis. (c)
DSA (center), 84 % proximal BA stenosis; TCDI, delayed
systolic upstroke in both proximal and mid-BA with MFVs
of 20 and 23 cm/s, indicating a most severe or elongated
stenosis (Adapted from Zhao L, Barlinn K, Sharma VK,
Tsivgoulis G, Cava LF, Vasdekis SN, Teoh HL,
Triantafyllou N, Chan BP, Sharma A, Voumvourakis K,
Stamboulis E, Saqqur M, Harrigan MR, Albright KC,
Alexandrov AV (2011) Velocity criteria for intracranial
stenosis revisited: an international multicenter study of
transcranial Doppler and digital subtraction angiography.
Stroke 42: 3429–3434. doi: 10.1161/strokeaha.111.
621235; with permission)

time-of-flight MRA (TOF-MRA) is the most com-
monly used MRA technique in clinical investiga-
tion of intracranial arteries. In addition, phase-
contrast MRA (PC-MRA), contrast-enhanced
MRA (CE-MRA), and quantitative MRA
(QMRA) could also be used to reveal anatomic
severity of luminal stenosis as well as flow infor-
mation in the case of ICAS. Besides depicting
vessel structures, most of these MRA techniques
bear flow information as well, which is a
12 Intracranial Atherosclerosis
MCA stenosis
30 30
40 40
mm
50 50
60 60
70 70
80 80
Fig. 2 Appearance of a critical right middle cerebral
artery (MCA) stem stenosis (white star) on transcranial
color-coded duplex (TCCD) (upper left) and power
motion-mode (PMD) transcranial Doppler imaging
(TCDI) (lower left). Flow disturbance is easily discernible
on both TCCD and PMD. TCDI waveforms of the diseased
right MCA and other cerebral arteries are displayed with a
corresponding magnetic resonance angiography image
particular advantage of this imaging modality
over others. However, in the case of symptomatic
ICAS, MRA techniques may not be available in
acute settings, at least at primary or secondary
medical centers, which limit its values in acute
evaluation of such lesions.
TOF-MRA (Fig. 4) is an angiographic method
without using any contrast agents, which could be
widely used in patients with contraindications to
contrast agents, such as renal failure and relevant
allergic reactions. According to the SONIA trial,
the positive and negative predictive values for
TOF-MRA to diagnose a 50–99 % intracranial
stenosis were 59 % (95 % CI: 54–65 %) and
91 % (95 % CI: 89–93 %), with DSA being the
reference standard [14]. Thus, TOF-MRA could
be a reliable tool to rule out an arterial stenosis in
ICAS, but positive diagnosis of luminal stenosis
by TOF-MRA needs to be confirmed by other
imaging methods, which is similar to that of
TCDI as mentioned above. This could be due to
211
-210
-160
180
(right). There is significantly increased mean flow velocity
(MFV, 220 cm/s) at the stenosis region and decreased MFV
(40 cm/s) and delayed upstroke (0.25 s) at the post-stenotic
region (Adapted from Topcuoglu MA (2012) Transcranial
Doppler ultrasound in neurovascular diseases: diagnostic
and therapeutic aspects. J Neurochem 123 Suppl 2: 39–51.
doi: 10.1111/j.1471-4159.2012.07942.x; with permission)
the signal contrast mechanism of TOF-MRA,
flow-related enhancement, which may lead to an
overestimation of the severity of luminal stenosis
in lesions with slow or minimal flow. Although
the flow-dependent nature of TOF-MRA may
hinder its ability to scale the luminal stenosis in
the case of ICAS, it could to a certain extent reveal
the hemodynamic impact of an ICAS lesion. So
the inherent flow information yields a unique
advantage of this imaging method over some of
the others, which is further discussed below [15].
Like TOF-MRA, PC-MRA does not need con-
trast agents either, though the latter technique is
not as commonly used as the former one in clin-
ical practice to evaluate intracranial atheroscle-
rotic disease. By using PC-MRA, not only the
arterial structure could be revealed, flow velocities
and cerebral blood flow could also be quantified,
based on phase shifts of flowing protons
within the vessel lumen. By using TOF-MRA to
facilitate three-dimensional vessel localization,

Fig. 3 Description and illustration of the thrombolysis in brain ischemia (TIBI)
transcranial Doppler imaging (TCDI) flow grades (Grades 0–5) to classify residual
flow through an occluded intracranial artery. Emergent TCDI TIBI classification was
found to be correlated with initial stroke severity, clinical recovery, and mortality in
stroke patients treated with intravenously administered tissue plasminogen activator
212
(Adapted from Demchuk AM, Burgin WS, Christou I, Felberg RA, Barber PA, Hill
MD, Alexandrov AV (2001) Thrombolysis in Brain Ischemia (TIBI) transcranial
Doppler flow grades predict clinical severity, early recovery, and mortality in patients
treated with intravenous tissue plasminogen activator. Stroke 32: 89–93. doi: 10.1161/
X. Leng and D.S. Liebeskind
01.STR.32.1.89; with permission)
12 Intracranial Atherosclerosis
Fig. 4 Intracranial atherosclerotic lesions at the left middle cerebral artery (a, arrow), right middle cerebral artery (b,
arrow), and basilar artery (c, arrow), as revealed by time-of-flight magnetic resonance angiography
PC-MRA-based QMRA could significantly
improve the accuracy of PC-MRA to quantify
cerebral blood flow, which is detailed below [23].
In addition to the contrast agent-independent
MRA techniques as above, CE-MRA is also fre-
quently used in clinical practice to simulta-
neously visualize the entire supra-aortic
vasculature, including extracranial and intracra-
nial arteries. With contrast injection, CE-MRA
could better delineate vessel structures as com-
pared with TOF-MRA, but the diagnostic accu-
racy of arterial stenoses by CE-MRA was lower
for intracranial arteries than that for extracranial
arteries. Further, CE-MRA does not bear flow
information in the evaluation of ICAS, unlike
TOF- and PC-MRA.
CTA
Routine single-phase CTA has been increasingly
used in the clinical evaluation of ICAS during the
past few years (Fig. 5). As a minimally invasive
imaging modality, routine CTA could yield a use-
ful tool for scaling the severity of luminal stenosis
in ICAS. By using intravenous contrast, CTA
could better depict the vessel geometry as com-
pared with MRA, but delineation of some seg-
ments of the intracranial vasculature, such as the
petrous and cavernous portions of internal carotid
artery, and the distal vertebral artery, may be
affected by surrounding bones, though the associ-
ated artifacts could be improved by proper post-
processing techniques.
Regarding the diagnostic ability of CTA for
defining the severity of luminal stenosis in
ICAS, there have been no large prospective stud-
ies. According to some small studies, CTA
213
probably has higher diagnostic accuracy than
MRA to define an ICAS lesion of 50–99 % lumi-
nal stenosis. However, based on results from the
SONIA trial, the positive predictive value of CTA
to define a 50–99 % intracranial stenosis was low,
almost comparable to that of MRA and TCDI
[24]. The SONIA trial suggested that, like MRA
or TCDI, CTA was also a useful tool to rule out the
presence of arterial stenosis due to ICAS but not
as much reliable in scaling the severity of luminal
stenosis, with DSA as the reference standard,
while the results may be biased by the fact that
only a small part of subjects analyzed in the
SONIA trial underwent CTA examinations.
Thus, further prospective large studies should be
performed to validate the diagnostic accuracy of
CTA for detecting arterial stenosis by ICAS.
Additionally, CTA could also be used to eval-
uate the thrombus burden in intracranial athero-
sclerotic disease, for instance, the clot burden
score (CBS), which has been found to be able to
predict functional outcome, final infarct size, and
the parenchymal hematoma risk [25]. Further,
CTA source images (CTA-SI) are very useful in
the assessment of tissue status in acute ischemic
stroke. Ischemic brain tissues with low cerebral
blood flow are revealed as regions without the
enhancement on CTA-SI, which could be differ-
entiated from brain tissues with normal cerebral
blood flow. By using the Alberta Stroke Program
Early CT Score (ASPECTS) on CTA-SI, acute
ischemic stroke within the middle cerebral arterial
territory could be more reliably identified, and the
final infarct volume and clinical outcomes could
be more accurately predicted, as compared with
non-contrast CT [26].
214
Fig. 5 Intracranial atherosclerosis on computed tomogra-
phy angiography (CTA). This patient presented with mul-
tiple acute ischemic lesions in the left hemisphere as shown
in the diffusion-weighted imaging (DWI) and apparent
diffusion coefficient (ADC) maps. Notice the old right
X. Leng and D.S. Liebeskind
parietal infarct. The CTA demonstrates a tight stenosis at
the left M1 segment of the middle cerebral artery. The 3D
reconstruction better delineates the finding and the vessel
morphology
12 Intracranial Atherosclerosis
Fig. 6 Axial (a) and coronal
(b) maximum intensity
projections of computed
tomography angiography
demonstrating
leptomeningeal collateral
blood flow from the posterior
cerebral artery (arrows) to
distal segments of an
occluded middle cerebral
artery (From Liebeskind DS
(2003) Collateral circulation.
Stroke 34: 2279–2284.
doi: 10.1161/01.STR.
0000086465.41263.06)
Moreover, CTA could be used to evaluate pri-
mary and secondary collaterals (Fig. 6). For
instance, there have been several methods to eval-
uate leptomeningeal collaterals by CTA-SI or
reconstructed CTA, the interobserver agreement
of which methods ranged from moderate to excel-
lent [18]. Multiple studies have demonstrated the
protective role of good collaterals as revealed by
CTA in clinical outcomes of stroke patients
[26]. However, till now, there have been no com-
monly accepted methods using CTA to evaluate
collateral circulation in the case of symptomatic
ICAS that has been validated in large prospective
studies. The inconsistency of these methods used
for evaluating collaterals may hinder the general-
ized use of this minimally invasive imaging
modality in multicenter studies to fully appreciate
the impact of collateral flow in the case of symp-
tomatic ICAS.
Although CTA has multiple applications in
the evaluation of ICAS, it tends to eliminate
any temporal flow information through the
lesions, which by itself is not a good tool for
evaluating hemodynamics of ICAS lesions. But
it may be a good source for the reconstruction of
blood flow models of ICAS by using the compu-
tational fluid dynamics (CFD) technique, which
is discussed below [15]. Another limitation of
this imaging method is that it is contrast agent
dependent, which is contraindicated in patients
with renal failure or allergic reactions to the
contrast agent.
215
DSA
DSA (Fig. 7), with its superb spatial resolution, is
regarded as the gold standard to diagnose the
severity of luminal stenosis, flow direction, and
collaterals in ICAS. It has been used in enormous
studies as the reference standard to validate the
diagnostic accuracies of other imaging modalities
to assess the intracranial vasculature, for instance,
in the SONIA trial [14]. It has also been used in
large clinical trials to determine the severity of
luminal stenosis by ICAS, for the sake of patient
selection for the investigation of certain treatment
strategies, for example, the WASID and the
SAMMPRIS trials [9, 10]. The WASID criteria
for measuring the percentage of luminal stenosis
in ICAS are now widely used in clinical practice
and relevant research areas. Further, although
there are still controversies in applying interven-
tional therapy in symptomatic ICAS, when
employed, the severity of luminal stenosis defined
by DSA is usually used to guide clinical decisions
or to guide patient selection in the setting of clin-
ical trials [27].
Moreover, DSA may be used to evaluate and
grade collateral status in symptomatic ICAS.
There have been multiple methods using DSA to
evaluate the collateral circulation, including col-
laterals of the Circle of Willis and leptomeningeal
collaterals. For instance, the American Society of
Interventional and Therapeutic Neuroradiology/
Society of Interventional Radiology (ASITN/
SIR) Collateral Flow Grading System (Table 1)
216
Fig. 7 Serial angiographic images from a selective left internal carotid artery injection in a patient who has middle
cerebral artery occlusion, illustrating retrograde collateral flow. (# David S. Liebeskind, MD.)
Table 1 The ASITN/SIR collateral flow grading system
Grade 0 No collaterals visible to the ischemic site
Grade 1 Slow collaterals to the periphery of the
ischemic site with persistence of some of the
defect
Grade 2 Rapid collaterals to the periphery of ischemic
site with persistence of some of the defect and
to only a portion of the ischemic territory
Grade 3 Collaterals with slow but complete
angiographic blood flow of the ischemic bed
by the late venous phase
Grade 4 Complete and rapid collateral blood flow to
the vascular bed in the entire ischemic
territory by retrograde perfusion
Adapted from Higashida RT, Furlan AJ, for the Technol-
ogy Assessment Committees of the American Society of
Interventional and Therapeutic Neuroradiology and the
Society of Interventional Radiology (2003) Trial design
and reporting standards for intra-arterial cerebral throm-
bolysis for acute ischemic stroke. Stroke 34: E109-E137.
doi: 10.1161/01.str.0000082721.62796.09; with permis-
sion. ASITN/SIR indicates American Society of Interven-
tional and Therapeutic Neuroradiology/Society of
Interventional Radiology
is one of the most commonly used grading
methods to gauge the function of collateral circu-
lation. However, there were few studies investi-
gating the interobserver reproducibility of
methods utilizing DSA to grade collateral circula-
tion in the case of ICAS. According to the limited
existing data, several of these methods were of
good and very good interobserver agreement [18].
X. Leng and D.S. Liebeskind
Collateral status graded by DSA has been found to
be able to alter the risk of stroke recurrence in
ischemic stroke or TIA due to ICAS, in the
WASID cohort and other studies as well
[15]. With the collateral status identified as an
independent predictor for recurrent stroke, proba-
bly surpassing the role of percent stenosis, in
patients with symptomatic ICAS [28], there is a
great need for the establishment and validation of
generalizable grading methods, so that the role of
collateral circulation could be further explored in
larger, multicenter clinical studies.
DSA may differentiate antegrade and retro-
grade flow by directly revealing the flow direc-
tions. However, with its invasive nature, DSA
could result in periprocedural complications,
including transient or even permanent neurologi-
cal deficits in rare cases, risk of which may be
operator dependent. Also, the DSA procedure is
relatively more expensive than other imaging
methods. Thus, DSA could not be widely used
in clinical scenarios, especially in primary and
secondary medical centers [29]. Moreover, the
superb resolution of DSA to depict intracranial
arteries may be affected by slow flow in severely
stenosed arteries that prevent or slow the contrast
agents to fill in the vessel, which sometimes may
lead to the overestimation of the severity of lumi-
nal stenosis or the hemodynamic impact of an
ICAS lesion.
12 Intracranial Atherosclerosis
Fig. 8 MR perfusion/diffusion mismatch in a patient with
left middle cerebral artery occlusion. Note the small area of
infarct core (arrow) in the diffusion-weighted image (DWI;
a) in comparison with the large area of perfusion abnormal-
ity (arrows) in mean transit time (MTT; b) resulting in
Routine CT and MR Perfusion Imaging
In the presence of an occluded intracranial artery,
downstream cerebral blood flow may fall below
the threshold of 20 mL/100 g tissue/min, under
which the function of neurons is impaired. If the
cerebral blood flow falls below 12 mL/100 g tis-
sue/min, irreversible damages could happen to
neuronal cells [30]. Thus, hypoperfusion has
been identified as an important mechanism for
ischemia in patients with symptomatic ICAS,
with affected brain tissues becoming the infarct
core or ischemic penumbra, depending on the
extent of hypoperfusion in different ischemic
areas. The proportion of salvageable tissues, or
the ischemic penumbra, and the infarct core, in the
case of acute ischemic stroke, to a great extent
depends on the balance between the residual
antegrade flow across the occlusive artery and
retrograde flow via collaterals. Therefore, routine
perfusion imaging methods CTP and DSC-PWI,
respectively, based on the imaging modalities of
CT and MRI, is valuable in evaluating the perfu-
sion status of focal and downstream brain tissues,
and hence reflecting the collateral status, in
patients with acute symptomatic ICAS [17].
The potentially salvageable ischemic penum-
bra is an ideal target for reperfusion therapies
which require proper patient selection, while the
217
positive mismatch with a large penumbra (Adapted from
Kloska SP, Wintermark M, Engelhorn T, Fiebach JB (2010)
Acute stroke magnetic resonance imaging: current status and
future perspective. Neuroradiology 52: 189–201. doi:
10.1007/s00234-009-0637-1; with permission)
mismatch between the hypoperfused area as
revealed by CT and MR perfusion imaging and
the acute infarct core serves as a reliable tool for
the identification of the penumbra [27]. The
mismatch between hypoperfused area on
DSC-PWI and the ischemic area on diffusion-
weighted imaging (DWI) (Fig. 8), and the mean
transit time-cerebral blood volume mismatched by
CTP (Fig. 9), has been tested in several large clin-
ical trials to guide selection of patients indicated for
thrombolysis beyond the 3-h time window.
Although definite clinical benefits of imaging-
guided thrombolysis in acute ischemic stroke
have not yet been demonstrated in these trials,
there was a trend that imaging-detected mismatch
might be correlated to increased reperfusion
[27]. According to the most up-to-date guideline,
CTP or DSC-PWI may be considered for the selec-
tion of patients for acute reperfusion therapy
beyond the time windows for intravenous fibrino-
lysis [27]. However, substantial heterogeneities in
the acquisition and post-processing protocols for
CT and MR perfusion imaging in exiting studies
may limit comparisons between studies and gener-
alization of the study results. Thus, further studies
are warranted and needed to validate the role of
these two perfusion imaging methods in defining
the penumbra for thrombolytic therapy.
218
Fig. 9 Mean transit time (MTT, e)-cerebral blood volume
(CBV, d) mismatch in a patient with acute ischemic stroke,
approximately 1 h after symptom onset. The upper row and
lower rows, respectively, show the ischemic signs in the
left middle cerebral territory at the basal ganglia level and
the supraganglionic level. (a–e), respectively, represent
non-contrast CT (a), source images of CT angiography
(b), penumbra (green) and infarct (red) maps by CT perfu-
sion superimposed on CTA source image (c), CBV maps
In addition to its application in detecting
mismatch, perfusion imaging adds additional
information to angiographic investigations in the
evaluation of collateral circulation in symptom-
atic ICAS, due to the reason that downstream
perfusion status in the case of intracranial arterial
occlusion could be partly determined by the col-
lateral compensating ability. Till now, there have
been several methods using perfusion imaging
alone, or by combining with angiographic imag-
ing, to evaluate the extent of collateralization,
some of which suggested beneficial effects of
good collateral flow in acute ischemic stroke [18].
As compared with MR perfusion, CTP is more
accessible and less time-consuming, which may
be more feasible to be applied in the hyperacute
phase of a symptomatic ICAS in clinical practice.
However, the need for contrast agents may limit
the use of routine CT and MR perfusion imaging
in patients with contraindications and may also
lead to contrast-induced complications. More
recently, arterial spin-labeling perfusion-weighted
X. Leng and D.S. Liebeskind
(d), and MTT maps (e). doi:10.1371/journal.
pone.0075615.g001 (Adapted from van Seeters T, Biessels
GJ, Niesten JM, van der Schaaf IC, Dankbaar JW, Horsch
AD, Mali WPTM, Kappelle LJ, van der Graaf Y, Velthuis
BK, on behalf of the Dust Investigators (2013) Reliability
of visual assessment of non-contrast CT, CT angiography
source images and CT perfusion in patients with suspected
ischemic stroke. PLoS ONE 8: e75615. doi: 10.1371/jour-
nal.pone.0075615. Copyright: # 2013 van Seeters et al.)
MR imaging (ASL-PWI) has been used to mea-
sure cerebral blood flow and reveal perfusion
status of individual cerebral arteries [22]. In
some small studies, the ability of ASL-PWI has
been tested to detect the ischemic penumbra so as
to guide patient selection for intravenous throm-
bolytic therapy, which may be an alternative for
patients with contraindications to the use of con-
trast agents. The application of ASL-PWI in
ischemic stroke due to ICAS is further discussed
below.
Novel Neuroimaging Methods
to Evaluate ICAS
Besides of the routine methods as mentioned
above, there are several novel imaging methods
increasingly applied in the assessment of ICAS,
which may facilitate more comprehensive assess-
ment and more efficient risk stratification of such
lesions. For instance, high-resolution magnetic
12 Intracranial Atherosclerosis
resonance imaging (HR-MRI) is increasingly
used to evaluate constituents and morphology of
intracranial plaques and the vessel wall, which is
promising in detecting vulnerable plaques and
revealing mechanisms for ICAS-related stroke
[31]; QMRA to quantify cerebral blood flow,
which could simultaneously visualize vascular
anatomy and evaluate the hemodynamic impact
of ICAS lesions [32]; ASL-PWI, with no need to
employ any contrast agents, to delineate perfusion
status and detect mismatch for reperfusion thera-
pies in acute stroke cases, to reveal perfusion
territory and quantify cerebral blood flow of indi-
vidual intracranial arteries, and to reveal and
quantify collateral functions [22]; and four-
dimensional CTA (4D CTA) or time-resolved
CTA or multiphase CTA to provide angiographic
and perfusion information based on CTP without
adding additional single-phase CTA scan, which
could improve the imaging quality as compared
with routine single-phase CTA and could reveal
flow directions and collateral status [33, 34].
Moreover, there are another two novel methods
to evaluate the hemodynamic impact of symptom-
atic ICAS, paralleling to the evaluation of frac-
tional flow in coronary artery disease, respectively
based on two routine imaging methods
TOF-MRA and CTA [35]. The MRA-based
method is a newly developed index termed signal
intensity ratio (SIR), representing the relative
change of signal intensities (SI) across a symp-
tomatic ICAS lesion on TOF-MRA images; while
the CTA-based method utilizes the CFD tech-
nique to simulate blood flow, so as to depict the
hemodynamic characteristics of symptomatic
ICAS lesions [35]. All of these novel methods
are of potential values in facilitating more com-
prehensive assessment and more accurate risk
stratification of ICAS lesions, discussed in
detailed below, which need further investigation
and validation in the near future.
HR-MRI
HR-MRI has been used to delineate coronary and
carotid plaques in enormous studies, both athero-
sclerotic and non-atherosclerotic. Morphology
and components of atherosclerotic plaques in
these arteries revealed by HR-MRI have been
219
validated with histological findings in both animal
models and humans, which reinforced the role of
this technique to be the leading noninvasive
in vivo imaging modality to characterize athero-
sclerotic plaques in these arteries [36]. Only in
recent years has HR-MRI been applied in the
characterization of intracranial plaques and the
adjacent vessel wall (Fig. 10). In recent relevant
studies, HR-MRI were usually performed under
the field strength of 1.5 or 3.0 T, with higher field
strength providing images of increased signal-to-
noise ratio and higher quality, which could
improve delineation of complex intracranial
plaques [31].
In several studies using HR-MRI to depict
coronary and carotid plaques, it has been demon-
strated feasible and reliable to quantify the plaque
size, to characterize the vessel wall, and to define
the plaque vulnerability in these arteries, while its
application in intracranial plaques has not been
fully appreciated [31, 36]. Based on limited data,
intracranial arterial stenosis due to atherosclerosis
could be differentiated from that of other causes
by using HR-MRI, and intraplaque hemorrhage
(Fig. 11), lipid core, and fibrous cap in ICAS
could also be detected. But components of intra-
cranial plaques as defined by HR-MRI have not
been widely validated with histopathological find-
ings [31]. According to a small postmortem,
in vitro study, HR-MRI at a high field strength of
7 T had substantial accuracy in detecting intracra-
nial plaque and its components [37].
In symptomatic carotid atherosclerosis,
intraplaque hemorrhage revealed by HR-MRI
has been correlated with increased plaque vulner-
ability and the risk of stroke recurrence, while
such characteristic of intracranial plaques has not
been well studied. According to the existing evi-
dence, intraplaque hemorrhage may be of higher
prevalence in symptomatic ICAS lesions than
asymptomatic lesions, but the correlation between
intraplaque hemorrhage in symptomatic ICAS
and the risk of stroke recurrence has not been
demonstrated. As for other constituents of athero-
sclerotic plaques, the presence and extent of lipid
core and the thickness of fibrous cap have also
been well correlated with the stroke risk in other
arterial beds, while the prognostic values of these
220
Fig. 10 High-resolution magnetic resonance imaging
(HR-MRI) images (3.0 T) of a symptomatic ICAS lesion
at the left middle cerebral artery (MCA). A 74-year-old
male had multiple acute infarcts (arrow, a) in the distribu-
tion of stenotic left MCA (arrow, b). On consecutive
sagittal T2-weighted image slices, a plaque with a band
of high signal (long arrows, c and d) is seen in the stenotic
characteristics in intracranial plaques need to be
further investigated [31].
Although intracranial arteries differ with
carotid and coronary arteries in the structure of
vessel walls, findings of the carotid and coronary
plaque and vessel wall imaging by HR-MRI shed
light on corresponding studies in intracranial ath-
erosclerotic plaques. Investigation of plaque mor-
phology and components in ICAS by HR-MRI
will help reveal underlying mechanisms of related
ischemic stroke or TIA, identify lesions prone to
rupture, and stratify patients with higher risk of
recurrence. Moreover, this technique may also be
useful in investigating the response of intracranial
plaques to medical therapies, such as antiplatelets
and statins, as well as interventional therapies,
which may guide individual treatment of patients
with such lesions [31].
X. Leng and D.S. Liebeskind
segment. The reference segment (a relatively normal seg-
ment) is also shown (e) (Adapted from Xu WH, Li ML,
Gao S, Ni J, Zhou LX, Yao M, Peng B, Feng F, Jin ZY, Cui
LY (2010) In vivo high-resolution MR imaging of symp-
tomatic and asymptomatic middle cerebral artery athero-
sclerotic stenosis. Atherosclerosis 212: 507–511. doi:
10.1016/j.atherosclerosis.2010.06.035; with permission)
QMRA
QMRA, based on PC-MRA, is a noninvasive tool
to measure blood flow in intracranial arteries,
which has been used to quantify cerebral blood
flow. By combining with TOF-MRA to facilitate
three-dimensional vessel localization, QMRA sig-
nificantly improves the accuracy of PC-MRA to
quantify cerebral blood flow [23]. Till now,
QMRA has been applied in the quantification of
total cerebral blood flow and regional blood flow
supplied by specific intracranial arteries in relevant
studies. QMRA can quantify blood flow through
major intracranial arteries including the anterior,
middle, and posterior cerebral arteries, and the bas-
ilar artery, as well as collaterals within the Circle of
Willis, so that it could provide information of terri-
torial blood flow and the hemodynamic impact of
individual atherosclerotic lesions [38]. QMRA has
12 Intracranial Atherosclerosis 221

Fig. 11 In a patient with

(a) multiple acute ischemic
lesions on diffusion-
weighted images located in
the left temporal lobe, a
magnetic resonance
angiography shows (b) a
high-grade atherosclerotic
stenosis of the left middle
cerebral artery. On high-
resolution magnetic
resonance imaging (3.0 T),
an eccentric plaque is
identified on T2-weighted
imaging (arrow in c), and
T1-weighted fat-suppressed
imaging shows intraplaque
hemorrhage (arrow in d)
(Adapted from Xu W-H, Li
M-L, Gao S, Ni J, Yao M,
Zhou L-X, Peng B, Feng F,
Jin Z-Y, Cui L-Y (2012)
Middle cerebral artery
intraplaque hemorrhage:
Prevalence and Clinical
Relevance. Ann Neurol 71:
195–198. doi: 10.1002/
ana.22626; with
permission)

also been found to be able to depict arterial wave-
forms during systolic and diastolic phases of the
cardiac cycle, which provided evidence of the
dampening of the arterial waveforms distal to
symptomatic ICAS lesions. As compared with
TCDI examinations for arterial waveforms,
QMRA is not dependent on operators or limited
by the absence of temporal window, and it could
simultaneously provide anatomic and hemody-
namic information of intracranial arteries affected
by ICAS lesions [38].
QMRA may not only directly reveal the direc-
tion of blood flow via the collaterals within the
Circle of Willis and quantify the compensating
blood flow volume, but could also indirectly assess
the presence and function of leptomeningeal col-
laterals [18, 38, 39]. According to a preliminary
study, significantly increased flow rates were
observed in ipsilateral posterior cerebral artery in
some cases with a flow-limiting ICAS lesion in the
middle cerebral artery or internal carotid artery,
which indirectly reflected collateral flow via
leptomeningeal anastomoses between the posterior
and anterior circulations [18, 39]. Thus, QMRA is a
promising tool to noninvasively and quantitatively
evaluate the collateral circulation in patients with
symptomatic ICAS.
In addition, QMRA can visualize the anatomic
and hemodynamic effects of interventional therapy
of affected arteries in the case of symptomatic
ICAS. As previously reported, QMRA could
reveal restoration of distal arterial waveforms and
enhancement of distal cerebral blood flow in symp-
tomatic cervical arterial stenosis treated by stenting
therapy. Though there were few data concerning of
such application of QMRA in symptomatic ICAS
lesions, there were small studies indicating that
QMRA could help assess intracranial in-stent ste-
nosis (Fig. 12). Therefore, this technique may facil-
itate noninvasive investigation and surveillance of
different responses of patients with symptomatic
ICAS lesions to the stenting therapy and the related
factors, so that it may help identify those who will
benefit from this therapy.
Furthermore, low flow distal to symptomatic
ICAS lesions of the vertebrobasilar arteries as
222
Fig. 12 Detection of an in-stent stenosis at the
supraclinoid right internal carotid artery (ICA) by quanti-
tative magnetic resonance angiography (QMRA). (a) Con-
ventional digital subtraction angiography (DSA)
demonstrating an ICAS lesion of 76 % luminal stenosis
at the supraclinoid right ICA before treatment. (b) Three-
dimensional surface rendering of the vasculature by the
Noninvasive Optimal Vessel Analysis (NOVA) system
(VasSol Inc., Chicago, Ill) demonstrating the scan line on
the ICA proximal to the region of disease (arrow). QMRA
measured a flow of 139 mL/min before treatment. (c) DSA
demonstrating mild residual stenosis (arrow) immediately
after angioplasty and wingspan stent placement. QMRA
quantified by QMRA was found to be related to a
higher risk of recurrence, which needs to be
further validated in prospective studies
[40]. Therefore, the noninvasive imaging method
of QMRA, by combining TOF- and PC-MRA
techniques, could simultaneously visualize vas-
cular anatomy and measure blood flow in the
presence of symptomatic ICAS lesions, assess
the presence and functions of primary and
X. Leng and D.S. Liebeskind
flow improved to 367 mL/min. (d) DSA demonstrating
in-stent restenosis with 83 % recurrent stenosis (arrow) at
4 months of follow-up. QMRA flow decreased to
94 mL/min. Two weeks later, the patient had a right hemi-
spheric stroke. (e) DSA demonstrating progressive in-stent
stenosis of >90 % after the stroke, with further decrease in
QMRA flow to 33 mL/min (Adapted from Amin-Hanjani-
S, Alaraj A, Calderon-Arnulphi M, Aletich VA, Thulborn
KR, Charbel FT (2010) Detection of intracranial in-stent
restenosis using quantitative magnetic resonance angiog-
raphy. Stroke 41: 2534–2538. doi: 10.1161/
strokeaha.110.594739; with permission)
secondary collaterals, and evaluate the structural
and hemodynamic effects of stenting therapy of
affected arteries, and it may also play a role in
risk stratification of patients with symptomatic
ICAS. But these applications of QMRA in symp-
tomatic ICAS were almost all based on small
studies, which need to be further verified in
larger studies, by validating against better-
established imaging modalities.
12 Intracranial Atherosclerosis
ASL-PWI
As mentioned above, hypoperfusion is probably
an important mechanism in ischemic stroke due to
ICAS, and the evaluation of perfusion status in
areas surrounding the infarct in acute ischemic
stroke is useful in identifying salvageable brain
tissues for thrombolytic therapy. Unlike routine
perfusion imaging based on CTP and DSC-PWI,
ASL-PWI relying on magnetic labeling of arterial
blood does not require the use of contrast agents,
which makes it a potential alternative for the eval-
uation of perfusion status in stroke patients with
contraindications to contrast agents. Also, this
contrast agent-free imaging method could speed
up screening and evaluation of all acute stroke
patients indicated for perfusion status examina-
tion, by eliminating the need for performing
renal function tests [17, 22].
During the past few years, ASL-PWI has
become a popular imaging method to evaluate
the cerebral hemodynamics in ischemic cerebro-
vascular disease. The accuracy of ASL-PWI for
evaluating cerebral perfusion status in ischemic
stroke has been validated with DSC-PWI in sev-
eral small studies [41]. Overall, the results
suggested that ASL- and DSC-PWI provided
basically consistent perfusion map in acute ische-
mic stroke, but ASL may overestimate the
hypoperfused volume as compared with
DSC-PWI in some cases. As for the ability of
ASL-PWI in detecting the ischemic penumbra,
referring to the mismatch between hypoperfused
region on ASL and acute ischemic lesion on DWI,
a small retrospective study reported the moderate
inter-modality agreement of ASL- and DSC-PWI
to categorize mismatch status, which also found
an overestimation of PWI lesion size by ASL in
some cases. Therefore, as a promising alternative
for contrast agent-dependent perfusion imaging,
practical application of ASL in assessing the
ischemic penumbra and guiding clinical decisions
in acute ischemic stroke due to ICAS could prob-
ably be pursued by further refinement of the
detecting threshold for hypoperfusion and
improvement of the technique [41].
As an advantage over routine perfusion imag-
ing methods, the recently emerging selective ASL
could encode blood flow of individual feeding
223
arteries, so that it could reveal perfusion territories
of individual cerebral arteries, commonly the left
and right internal carotid arteries and the basilar
artery (Fig. 13). This could facilitate more accu-
rate identification of perfusion territories of differ-
ent cerebral arteries in patients with symptomatic
ICAS lesions, especially in those with anatomic
variation of intracranial arteries [42]. Further, this
technique also enables noninvasive quantification
of regional cerebral blood flow of individual feed-
ing cerebral arteries, which leads to the identifica-
tion of hemodynamically significant ICAS lesions
[42]. Also, selective ASL was found to be able to
quantitatively assess impaired cerebrovascular
reactivity at the brain-tissue level, in patients
with symptomatic internal carotid artery occlusive
disease versus healthy controls [43].
Moreover, both nonselective and selective
ASLs could be used to evaluate the collateral
circulation in the presence of symptomatic ICAS
[18, 42]. Similar to DSC-PWI, nonselective ASL
could depict the collateral circulation in symptom-
atic ICAS by revealing the brain perfusion status,
only that the latter does not require application of
contrast agents. Compared with nonselective
ASL, selective ASL could provide more informa-
tion regarding the collateral status, including the
presence, the origin, and the function of collateral
flow. Collateral flow defined by selective ASL
was found to be substantially consistent with that
confirmed by DSA in small studies [22]. More-
over, selective ASL could reveal direct correla-
tions between collateral flow and the
corresponding compensated perfusion territories,
which could not be achieved by DSA examina-
tion. Therefore, selective ASL is a promising tool
to noninvasively and quantitatively depict territo-
rial perfusion status and collateral circulation in
clinical scenarios of symptomatic ICAS.
Four-Dimensional CTA
As mentioned above, routine CTA, or single-
phase CTA, is used to evaluate the severity of
luminal stenosis, the thrombus burden, and the
collateral status in the case of symptomatic
ICAS, and CTP is used to assess the perfusion
status, to detect the salvageable penumbra, and to
indirectly reveal the collateral circulation in the
224
Fig. 13 Illustration for selective arterial spin labeling
(ASL). Transverse (a), sagittal (b), and coronal (c) maxi-
mum intensity projections of the circle of Willis show the
planning of oblique labeling planes for territorial ASL
imaging of the right internal carotid artery (ICA) (red),
left ICA (green), and the posterior circulation (blue). Ter-
ritorial ASL images (d–f) from caudal (d) to cranial (f) of
one patient with an ischemic lesion in the posterior circu-
lation (e, star) due to a flow-limited arterial stenosis at the
presence of such lesions. According to a study that
systematically compared the diagnostic accura-
cies of CTP and single-phase CTA in the assess-
ment of the site of arterial occlusion, infarct core,
salvageable brain tissue, and collateral circulation
in patients with or without symptomatic ICAS
lesions, the combination of these two techniques
was the most accurate assessment method
[44]. With the recent advances in CT scanning
technique, high-quality CTA data could be
derived from whole-brain CTP source images,
which was referred to as 4D CTA, or time-
resolved CTA, or multiphase CTA, as compared
to the routine single-phase CTA [33, 34]. Unlike
routine CTA images, which are reconstructed only
based on a short interval of the arterial phase after
contrast injection, 4D CTA derived from CTP
X. Leng and D.S. Liebeskind
right posterior cerebral artery (a, star). The perfusion terri-
tory images of the separate labeling cycles are
superimposed and color coded according to the labeling
planes above (Adapted from Hartkamp NS, Petersen ET,
De Vis JB, Bokkers RPH, Hendrikse J (2013) Mapping of
cerebral perfusion territories using territorial arterial spin
labeling: techniques and clinical application. NMR
Biomed 26: 901–912. doi: 10.1002/nbm.2836; with
permission)
contains hemodynamic information and delin-
eates the vascular geometry based on sequential
images acquired at multiple phases [33].
Four-dimensional CTA has recently been used
in the evaluation of intracranial vasculature. The-
oretically, 4D CTA images, based on perfusion
CT images acquired over time and covering the
whole period of contrast inflow into the brain,
could improve the imaging quality as compared
with routine CTA, in which depiction of intracra-
nial arteries could be suboptimal due to the differ-
ences in the contrast material arrival in different
areas of the brain. According to available data
from clinical studies, 4D CTA based on CTP did
have superior or at least comparable image quality
to detect intracranial arterial stenosis, as compared
with that of routine CTA, while the application of
12 Intracranial Atherosclerosis
Fig. 14 Antegrade flow across incomplete vessel occlu-
sion of the right middle cerebral artery (MCA) revealed by
four-dimensional computed tomographic angiography
(CTA), as confirmed by digital subtraction angiography
(DSA). (a) Coronal maximum intensity projection from
single-phase CTA in a patient with left-side hemiplegia
demonstrates short-segment occlusion of the right MCA.
(b) Early to late four-dimensional CTA (left to right) dem-
onstrates early opacification distal to the occlusion,
extending more distally in an antegrade fashion (arrows).
the former technique could reduce the total radia-
tion dose and the amount of contrast agents, by
performing a CTP examination only without
adding an additional single-phase CTA scan, in
the circumstances when both angiographic and
perfusion information are required [33].
4D CTA could be used to assess an important
determinator of the prognosis of stroke patients
with symptomatic ICAS lesions as mentioned
above, which is the collateral circulation. Though
relevant data were limited, 4D CTA could reflect
the extent and function of collateral flow in the
presence of such lesions, with better collateral
compensation correlated with smaller acute ische-
mic lesions on DWI images [45]. Moreover, in a
225
(c) Serial right internal carotid artery injection DSA dem-
onstrates slow antegrade opacification distal to the occlu-
sion (arrows), confirming an incomplete vessel occlusion
(Adapted from Frolich AMJ, Psychogios MN, Klotz E,
Schramm R, Knauth M, Schramm P (2012) Antegrade
flow across incomplete vessel occlusions can be distin-
guished from retrograde collateral flow using
4-dimensional computed tomographic angiography. Stroke
43: 2974–2979. doi: 10.1161/strokeaha.112.668889; with
permission)
small study, it was suggested of high sensitivity
and specificity, as well as substantial interobserver
agreement, to differentiate antegrade residual flow
through incompletely occlusive vessels from ret-
rograde collateral flow in the case of symptomatic
ICAS (Fig. 14), in which residual antegrade flow
as revealed by 4D CTA was associated with an
increased chance of early vessel recanalization in
those treated with thrombolytic therapy [34].
A Novel Method to Gauge Hemodynamic
Impact of ICAS Based on TOF-MRA
As mentioned above, TOF-MRA is a noninvasive
imaging method widely used in clinical practice to
depict the intracranial vasculature. Although it is
226
Fig. 15 The method for measurement of signal intensity
ratio (SIR) of an intracranial atherosclerotic lesion on a
magnetic resonance angiography maximum intensity pro-
jection. SIR of the lesion at the right middle cerebral artery
is calculated as the ratio of mean signal intensities distal
(1039.6) and proximal (1340.3) to the lesion, adjusted by
probably of relatively low positive predictive
value in quantifying the severity of luminal steno-
sis of cerebral arteries [14], TOF-MRA yields a
potential imaging modality to noninvasively
gauge the hemodynamic impact of ICAS, based
on its unique contrast mechanism of inflow-
enhancement effect, also known as flow-related
enhancement. TOF-MRA using gradient-echo
sequences allows stationary tissues to become
saturated and produce minimal background sig-
nals, whereas flowing blood to produce higher
SIs. The degree of enhancement of flowing
blood, referred to as SI on TOF-MRA images,
nonlinearly increases with increasing absolute
flow velocity, with other parameters being con-
stant. Laminar flow in normal and non-tortuous
arteries produces relatively uniform signal inten-
sities along the longitudinal axis of a specific
vessel segment and appears bright on maximum
intensity projections (MIP) of TOF-MRA. While
in the presence of flow-limited ICAS lesions, the
turbulence or slow flow might cause signal
X. Leng and D.S. Liebeskind
the mean background signal intensity (401.1; mean of
409.5 and 392.6), which is (1039.6 401.1)/
(1340.3 401.1) = 0.68 (From Leng X, Wong KS,
Liebeskind DS (2014) Evaluating intracranial atheroscle-
rosis rather than intracranial stenosis. Stroke 45: 645–651.
doi: 10.1161/STROKEAHA.113.002491)
reduction or complete signal loss in situ and distal
to the stenosis. Therefore, changes of SIs across
an ICAS on MIPs of TOF-MRA could to a certain
extent represent the hemodynamic significance of
the lesion. Thus, a novel term SIR as
abovementioned has been developed to quantify
the hemodynamic impact of symptomatic ICAS,
which represented the ratio of post- and
pre-lesional signal intensities on MRA MIP
adjusted by the background SI (Fig. 15) [15, 35].
The existing evidence suggested that the index
of SIR was feasible and easy-to-perform and had
substantial intra- and interobserver reproducibility
as well [15, 46]. Moreover, there were also studies
suggesting the clinical relevance of this index to
define the hemodynamic impact of symptomatic
ICAS lesions. For instance, significant linear cor-
relation was identified between the SIR values of
symptomatic ICAS lesions and the infarct vol-
umes on DWI in the acute phase of ICAS-related
ischemic stroke, which indicated that SIR of an
ICAS as evaluated on TOF-MRA might reflect its
12 Intracranial Atherosclerosis
functional and hemodynamic significance
[46]. Further, dichotomized SIR values by the
median have been found to be independently cor-
related to recurrent ischemic stroke in the WASID
cohort [15]. Therefore, SIR of symptomatic ICAS
could be a potential indicator to predict the recur-
rent risk of affected patients, which may yield a
novel noninvasive and easy-to-perform method to
stratify symptomatic ICAS lesions. However, all
of the existing studies were retrospective, and
some were of small sample sizes. Also, till now,
its ability to gauge the hemodynamic impact of
ICAS has not been validated. Therefore, future
prospective studies are needed to validate this
index against other imaging modalities capable
of evaluating the hemodynamic impact of ICAS
lesions, as well as to investigate its ultimate value
in risk stratification of affected patients.
CFD Modeling of ICAS to Reveal
the Hemodynamic Characteristics
In the past decade, the CFD technique has been
used in coronary and carotid artery diseases to
reveal hemodynamic characteristics of atheroscle-
rotic stenoses [35]. As for coronary artery disease,
the CFD technique has been applied in the simu-
lation of blood flow through the lesion based on
coronary CTA images. The noninvasive fractional
flow reserve derived from CTA (FFRCT) assessed
by using this technique has been found of sub-
stantial diagnostic accuracy of the hemodynamic
significance of coronary artery disease, with inva-
sive fractional flow reserve (FFR) obtained during
the performance of percutaneous coronary angi-
ography as the reference standard [47]. Although
cerebral and coronary arteries differ in anatomic,
physiological, and other aspects, advanced use of
the CFD modeling technique in the cardiology
field could inspirit similar application of this tech-
nique in ICAS, to facilitate direct evaluation of the
hemodynamic characteristics of such lesions.
CFD modeling of ICAS lesions could be
achieved based on different imaging modalities
providing angiographic information, such as
DSA, CTA, and MRA, while CTA may be a better
source for this purpose. Although it has been
demonstrated feasible to construct CFD models
based on biplane DSA images, CTA-based blood
227
flow simulation could eliminate the invasive pro-
cedure of DSA and the consequent low but unde-
niable risk of periprocedural complications.
Further, since the simulation of blood flow by
the CFD technique is based on the reconstructed
vessel geometry and that CTA could better depict
the anatomy of intracranial arteries as compared
with MRA, CTA may also be superior to MRA as
the source for CFD modeling of ICAS. CFD
modeling of symptomatic ICAS lesions based on
routinely obtained CTA source images has been
demonstrated feasible, according to the limited
number of relevant studies [15]. According to
the limited data, CFD models could directly reveal
the hemodynamic characteristics of such lesions,
such as pressure gradient across the lesion, and
turbulent flow and increased shear stress at the site
of arterial stenosis and downstream (Fig. 16).
CFD models could also directly depict the hemo-
dynamic impact of interventional therapy of the
ICAS lesions [48, 49]. However, till now, the few
studies utilizing the CFD technique to investigate
intracranial atherosclerotic disease were all of
extremely small sample sizes, and there is still a
long way to go before an extensive application of
this technique in clinical evaluation of ICAS. But
encouraging findings regarding this technique in
the field of cardiovascular diseases imply poten-
tial values of its use in cerebrovascular diseases.
Future studies exploring the clinical relevance and
possible prognostic values of hemodynamic char-
acteristics of symptomatic ICAS lesions as delin-
eated by CFD models could shed new light on the
assessment and diagnosis of such lesions.
Treatment Strategies for Patients
with ICAS
Primary prevention of intracranial atherosclerotic
disease lies in the control of associated modifiable
risk factors, such as cigarette smoking, hyperten-
sion, dyslipidemia, diabetes, and metabolic syn-
drome, while the application of aspirin for
primary prevention of ICAS in general
populations has not been well established
[50]. Regarding the acute treatment and secondary
prevention of patients with symptomatic ICAS, so
228 X. Leng and D.S. Liebeskind

Fig. 16 Computational a
fluid dynamics model Pressure
reconstructed based on
1.196e+002
computed tomography
angiography source images
showing the hemodynamic 9.660e+001
characteristics of a
symptomatic ICAS lesion at
the left middle cerebral 7.363e+001
artery: decreased pressure
(a, arrows), and increased
flow velocity (b, arrows) 5.066e+001
and wall shear stress (c,
arrows) at the site of the 2.769e+001
lesion and downstream [mm Hg]

b
Velocity
4.268e+000

3.202e+000

2.135e+000

1.069e+000

2.004e-003
[m s^-1]

c
Wall Shear
1.396e+002

1.047e+002

6.978e+001

3.489e+001

0.000e+000
[Pa]

far there have been no specific guidelines, which
were usually incorporated in guidelines on ische-
mic stroke or TIA due to all causes [2, 3, 27].
According to the current guidelines in the
United States and Europe, patients with acute
ischemic symptoms due to ICAS identified by
imaging examinations that have not resolved are
recommended to receive intravenous fibrinolytic
therapy, if within the 3-h time window and no
contraindications exist. For those who can be
treated within 3–4.5 h after ictus, intravenous
fibrinolytic therapy is probably reasonable for
more selective patients, with additional exclusion
criteria. Multimodal imaging investigations pro-
viding information regarding the infarct core and
the penumbra may be considered to select eligible
12 Intracranial Atherosclerosis
patients for this therapy who are beyond the time
windows. For patients with acute symptomatic
ICAS who are ineligible for or irresponsive to
intravenous thrombolysis, other treatment strate-
gies may be considered for carefully selected indi-
viduals, such as intra-arterial thrombolysis and
mechanical thrombectomy, which procedures are
not routinely applied in clinical practice and need
to be further validated in randomized clinical tri-
als, while clinical application of emergent intra-
cranial angioplasty and/or stenting for the
responsible artery has not been well established
for acute symptomatic ICAS [3, 27].
For the secondary prevention of patients with
ICAS-related ischemic stroke or TIA, the two
large, prospective, randomized clinical trials,
WASID and SAMMPRIS, provided the evidence
for clinical decision making concerning
antiplatelet, anticoagulant, or interventional ther-
apies for affected patients [2, 3]. According to the
WASID trial, aspirin was superior to warfarin in
secondary prevention of ischemic stroke or TIA
due to an ICAS of 50–99 % luminal stenosis, in
the way that aspirin was as effective as warfarin in
reducing the risks of stroke recurrence and other
vascular events while at the mean time safer than
warfarin by significantly decreasing the risks of
adverse events including death and major hemor-
rhage [9]. However, although several large ran-
domized clinical trials have been investigating the
efficacy and safety of dual versus mono-
antiplatelet therapies in the secondary prevention
of ischemic stroke and TIA patients, few data
existed specifically on that of ICAS-related
stroke. Till now, there has been no agreement on
the application of dual antiplatelet therapy in the
secondary prevention of symptomatic ICAS.
Besides the antiplatelet therapy, a risk factor con-
trol should also be included in the medical treat-
ment of ICAS-related ischemic stroke or TIA,
such as management of hypertension, diabetes,
and dyslipidemia [2, 3].
Concerning the intracranial angioplasty and
stenting therapy, the lately released final results
of the SAMMPRIS trial indicated early and last-
ing benefits of aggressive medical treatment over
stenting for those with a symptomatic ICAS of
70–99 % luminal stenosis, who were traditionally
229
identified as of higher risk than those with a
symptomatic ICAS of 50–69 % luminal stenosis
[10, 11]. As compared with medical treatment,
extracranial-intracranial bypass surgery has been
demonstrated to be associated with worse out-
comes for patients with symptomatic ICAS
lesions based on existing evidence, which is not
recommended for such patients [2].
Despite of all these treatment strategies, the
risk of stroke recurrence remains particularly
high in patients with symptomatic ICAS, which
is up to 15–20 % at 1 year after ictus in lesions
resulting in 50–99 % luminal stenosis
[9–11]. Therefore, accurate and timely risk strati-
fication of these patients based on their demo-
graphic and clinical features, and especially
findings from appropriate imaging investigations
for different aspects of ICAS lesions as discussed
above, could help identify patients truly at high
risk and facilitate clinical decision making.
Summary
Intracranial atherosclerosis is a disease of consid-
erable prevalence throughout the world, espe-
cially in Asia. Among the ischemic stroke of
different causes, those due to ICAS are at partic-
ularly high risk of recurrence. Currently, the eval-
uation of symptomatic ICAS is usually based on
the severity of luminal stenosis, which has been
found to be an independent predictor for the recur-
rent risk. However, with the emerging evidence on
the role of collateral circulation in risk stratifica-
tion of symptomatic ICAS, increasing attention
has been attracted to imaging methods capable
of revealing the collateral status. In addition,
plaque imaging and novel perfusion and angio-
graphic imaging methods have been rising in the
field of ICAS during the past few years to delin-
eate different features of ICAS lesions. Moreover,
inspired by the advances in studies of the hemo-
dynamic impact of coronary artery disease, there
have been preliminary studies to investigate that
of symptomatic ICAS respectively based on
TOF-MRA and CTA. In the near future, ICAS
should be paid more attention in clinical practice
and relevant studies. By the reasonable
230
application of imaging methods and comprehen-
sive interpretation of the imaging results, affected
patients could be timely and accurately stratified
by the risk of recurrence, and treatment strategies
could be properly determined. This in the long run
may reduce the recurrent risk of such lesions.
References
1. Qureshi AI, Feldmann E, Gomez CR, Johnston SC,
Kasner SE, Quick DC, Rasmussen PA, Suri MF, Taylor
RA, Zaidat OO (2009) Intracranial atherosclerotic dis-
ease: an update. Ann Neurol 66:730–738. doi:10.1002/
ana.21768
2. Furie KL, Kasner SE, Adams RJ, Albers GW, Bush
RL, Fagan SC, Halperin JL, Johnston SC, Katzan I,
Kernan WN, Mitchell PH, Ovbiagele B, Palesch YY,
Sacco RL, Schwamm LH, Wassertheil-Smoller S,
Turan TN, Wentworth D, American Heart Association
Stroke Council, Council on Cardiovascular Nursing,
Council on Clinical Cardiology, Interdisciplinary
Council on Quality of Care and Outcomes Research
(2011) Guidelines for the prevention of stroke in
patients with stroke or transient ischemic attack: a
guideline for healthcare professionals from the Amer-
ican Heart Association/American Stroke Association.
Stroke 42:227–276. doi:10.1161/STR.0b013e3
181f7d043
3. The European Stroke Organisation Executive Commit-
tee and the ESO Writing Committee (2008) Guidelines
for management of ischaemic stroke and transient
ischaemic attack 2008. Cerebrovasc Dis 25:457–507.
doi:10.1159/000131083
4. Moore KJ, Tabas I (2011) Macrophages in the patho-
genesis of atherosclerosis. Cell 145:341–355.
doi:10.1016/j.cell.2011.04.005
5. Wentzel JJ, Chatzizisis YS, Gijsen FJH, Giannoglou
GD, Feldman CL, Stone PH (2012) Endothelial shear
stress in the evolution of coronary atherosclerotic
plaque and vascular remodelling: current understand-
ing and remaining questions. Cardiovasc Res
96:234–243. doi:10.1093/cvr/cvs217
6. Leng XY, Chen XY, Chook P, Xiong L, Lin WH, Liu
JY, Tomlinson B, Thomas GN, Lam TH, Lam KSL,
Cheung BMY, Wong KS (2013) Correlation of large
artery intracranial occlusive disease with carotid
intima-media thickness and presence of carotid plaque.
Stroke 44:68–72. doi:10.1161/strokeaha.112.675652
7. Consigny PM (1995) Pathogenesis of atherosclerosis.
Am J Roentgenol 164:553–558. doi:10.2214/
ajr.164.3.7863871
8. Gao S, Wang YJ, Xu AD, Li YS, Wang DZ (2011)
Chinese ischemic stroke subclassification. Front
Neurol 2:6. doi:10.3389/fneur.2011.00006
9. Chimowitz MI, Lynn MJ, Howlett-Smith H, Stern BJ,
Hertzberg VS, Frankel MR, Levine SR, Chaturvedi S,
X. Leng and D.S. Liebeskind
Kasner SE, Benesch CG, Sila CA, Jovin TG, Romano
JG (2005) Comparison of warfarin and aspirin for
symptomatic intracranial arterial stenosis. N Engl J
Med 352:1305–1316. doi:10.1056/NEJMoa043033
10. Chimowitz MI, Lynn MJ, Derdeyn CP, Turan TN,
Fiorella D, Lane BF, Janis LS, Lutsep HL, Barnwell
SL, Waters MF, Hoh BL, Hourihane JM, Levy EI,
Alexandrov AV, Harrigan MR, Chiu D, Klucznik RP,
Clark JM, McDougall CG, Johnson MD, Pride GL Jr,
Torbey MT, Zaidat OO, Rumboldt Z, Cloft HJ (2011)
Stenting versus aggressive medical therapy for intra-
cranial arterial stenosis. N Engl J Med 365:993–1003.
doi:10.1056/NEJMoa1105335
11. Derdeyn CP, Chimowitz MI, Lynn MJ, Fiorella D,
Turan TN, Janis LS, Montgomery J, Nizam A, Lane
BF, Lutsep HL, Barnwell SL, Waters MF, Hoh BL,
Hourihane JM, Levy EI, Alexandrov AV, Harrigan
MR, Chiu D, Klucznik RP, Clark JM, McDougall
CG, Johnson MD, Pride GL Jr, Lynch JR, Zaidat OO,
Rumboldt Z, Cloft HJ (2014) Aggressive medical treat-
ment with or without stenting in high-risk patients with
intracranial artery stenosis (SAMMPRIS): the final
results of a randomised trial. Lancet 383:333–341.
doi:10.1016/s0140-6736(13)62038-3
12. Kwon SU, Hong KS, Kang DW, Park JM, Lee JH, Cho
YJ, Yu KH, Koo JS, Wong KSL, Lee SH, Lee KB, Kim
DE, Jeong SW, Bae HJ, Lee BC, Han MK, Rha JH, Kim
HY, Mok VC, Lee YS, Kim GM, Suwanwela NC, Yun
SC, Nah HW, Kim JS (2011) Efficacy and safety of
combination antiplatelet therapies in patients with symp-
tomatic intracranial atherosclerotic stenosis. Stroke
42:2883–2890. doi:10.1161/strokeaha.110.609370
13. Jung JM, Kang DW, Yu KH, Koo JS, Lee JH, Park JM,
Hong KS, Cho YJ, Kim JS, Kwon SU, Toss-2 Investiga-
tors (2012) Predictors of recurrent stroke in patients with
symptomatic intracranial arterial stenosis. Stroke
43:2785–2787. doi:10.1161/STROKEAHA.112.659185
14. Feldmann E, Wilterdink JL, Kosinski A, Lynn M,
Chimowitz MI, Sarafin J, Smith HH, Nichols F,
Rogg J, Cloft HJ, Wechsler L, Saver J, Levine SR,
Tegeler C, Adams R, Sloan M, The SONIATrial Inves-
tigators (2007) The Stroke Outcomes and Neuroimag-
ing of Intracranial Atherosclerosis (SONIA) trial.
Neurology 68:2099–2106. doi:10.1212/01.
wnl.0000261488.05906.c1
15. Leng X, Wong KS, Liebeskind DS (2014) Evaluating
intracranial atherosclerosis rather than intracranial
stenosis. Stroke 45:645–651. doi:10.1161/
STROKEAHA. 113.002491
16. Alexandrov AV, Sloan MA, Tegeler CH, Newell DN,
Lumsden A, Garami Z, Levy CR, Wong LK,
Douville C, Kaps M, Tsivgoulis G (2012) Practice
standards for transcranial Doppler (TCD) ultrasound.
Part II. Clinical indications and expected outcomes. J
Neuroimaging 22:215–224. doi:10.1111/j.1552-
6569.2010.00523.x
17. Khan R, Nael K, Erly W (2013) Acute stroke imaging:
what clinicians need to know. Am J Med 126:379–386.
doi:10.1016/j.amjmed.2012.11.014
12 Intracranial Atherosclerosis
18. McVerry F, Liebeskind DS, Muir KW (2012) System-
atic review of methods for assessing leptomeningeal
collateral flow. AJNR Am J Neuroradiol 33:576–582.
doi:10.3174/ajnr.A2794
19. Palazzo P, Balucani C, Barlinn K, Tsivgoulis G,
Zhang Y, Zhao L, Dewolfe J, Toaldo B,
Stamboulis E, Vernieri F, Rossini PM, Alexandrov
AV (2010) Association of reversed Robin Hood syn-
drome with risk of stroke recurrence. Neurology
75:2003–2008. doi:10.1212/WNL.0b013e3181ffe4e4
20. Nedelmann M, Stolz E, Gerriets T, Baumgartner RW,
Malferrari G, Seidel G, Kaps M (2009) Consensus
recommendations for transcranial color-coded duplex
sonography for the assessment of intracranial arteries
in clinical trials on acute stroke. Stroke 40:3238–3244.
doi:10.1161/strokeaha.109.555169
21. Tsivgoulis G, Eggers J, Ribo M, Perren F, Saqqur M,
Rubiera M, Sergentanis TN, Vadikolias K, Larrue V,
Molina CA, Alexandrov AV (2010) Safety and efficacy
of ultrasound-enhanced thrombolysis: a comprehen-
sive review and meta-analysis of randomized and
nonrandomized studies. Stroke 41:280–287.
doi:10.1161/strokeaha. 109.563304
22. Donahue MJ, Strother MK, Hendrikse J (2012) Novel
MRI approaches for assessing cerebral hemodynamics
in ischemic cerebrovascular disease. Stroke
43:903–915. doi:10.1161/STROKEAHA.111.635995
23. Zhao M, Charbel FT, Alperin N, Loth F, Clark ME (2000)
Improved phase-contrast flow quantification by three-
dimensional vessel localization. Magn Reson Imaging
18:697–706. doi:10.1016/S0730-725X(00)00157-0
24. Liebeskind DS, Kosinski AS, Saver JL, Feldmann E,
SONIA Investigators (2007) CT angiography in the
Stroke Outcomes and Neuroimaging of Intracranial
Atherosclerosis (SONIA) study. Stroke 32:477
25. Puetz V, Dzialowski I, Hill MD, Subramaniam S,
Sylaja PN, Krol A, O’Reilly C, Hudon ME, Hu WY,
Coutts SB, Barber PA, Watson T, Roy J, Demchuk AM,
Calgary CTASG (2008) Intracranial thrombus extent
predicts clinical outcome, final infarct size and hemor-
rhagic transformation in ischemic stroke: the clot bur-
den score. Int J Stroke 3:230–236. doi:10.1111/j.1747-
4949.2008.00221.x
26. Mortimer AM, Simpson E, Bradley MD, Renowden
SA (2013) Computed tomography angiography in
hyperacute ischemic stroke: prognostic implications
and role in decision-making. Stroke 44:1480–1488.
doi:10.1161/strokeaha.111.679522
27. Jauch EC, Saver JL, Adams HP, Bruno A, Connors JJ,
Demaerschalk BM, Khatri P, McMullan PW, Qureshi
AI, Rosenfield K, Scott PA, Summers DR, Wang DZ,
Wintermark M, Yonas H, American Heart Association
Stroke, Council CoCN, Council on Peripheral Vascular
Disease, Council on Clinical Cardiology (2013) Guide-
lines for the early management of patients with acute
ischemic stroke: a guideline for healthcare profes-
sionals from the American Heart Association/Ameri-
can Stroke Association. Stroke 44:870–947.
doi:10.1161/STR. 0b013e318284056a
231
28. Liebeskind DS, Cotsonis GA, Saver JL, Lynn MJ,
Turan TN, Cloft HJ, Chimowitz MI (2011) Collaterals
dramatically alter stroke risk in intracranial atheroscle-
rosis. Ann Neurol 69:963–974. doi:10.1002/ana.22354
29. Willinsky RA, Taylor SM, terBrugge K, Farb RI,
Tomlinson G, Montanera W (2003) Neurologic com-
plications of cerebral angiography: prospective analy-
sis of 2,899 procedures and review of the literature.
Radiology 227:522–528. doi:10.1148/radiol.
2272012071
30. Kloska SP, Wintermark M, Engelhorn T, Fiebach JB
(2010) Acute stroke magnetic resonance imaging: cur-
rent status and future perspective. Neuroradiology
52:189–201. doi:10.1007/s00234-009-0637-1
31. Bodle JD, Feldmann E, Swartz RH, Rumboldt Z,
Brown T, Turan TN (2013) High-resolution magnetic
resonance imaging: an emerging tool for evaluating
intracranial arterial disease. Stroke 44:287–292.
doi:10.1161/STROKEAHA.112.664680
32. Prabhakaran S, Warrior L, Wells KR, Jhaveri MD,
Chen M, Lopes DK (2009) The utility of quantitative
magnetic resonance angiography in the assessment of
intracranial in-stent stenosis. Stroke 40:991–993.
doi:10.1161/strokeaha.108. 522391
33. Smit EJ, Vonken EJ, van der Schaaf IC, Mendrik AM,
Dankbaar JW, Horsch AD, van Seeters T, van
Ginneken B, Prokop M (2012) Timing-invariant recon-
struction for deriving high-quality CT angiographic
data from cerebral CT perfusion data. Radiology
263:216–225. doi:10.1148/radiol.11111068
34. Frolich AMJ, Psychogios MN, Klotz E, Schramm R,
Knauth M, Schramm P (2012) Antegrade flow across
incomplete vessel occlusions can be distinguished
from retrograde collateral flow using 4-dimensional
computed tomographic angiography. Stroke
43:2974–2979. doi:10.1161/strokeaha.112.668889
35. Liebeskind DS, Feldmann E (2013) Fractional flow in
cerebrovascular disorders. Interv Neurol 1:87–99.
doi:10.1159/000346803
36. Corti R, Fuster V (2011) Imaging of atherosclerosis:
magnetic resonance imaging. Eur Heart J
32:1709–1719b. doi:10.1093/eurheartj/ehr068
37. Majidi S, Sein J, Watanabe M, Hassan AE, Van de
Moortele PF, Suri MF, Clark HB, Qureshi AI (2013)
Intracranial-derived atherosclerosis assessment: an
in vitro comparison between virtual histology by intra-
vascular ultrasonography, 7 T MRI, and histopatho-
logic findings. AJNR Am J Neuroradiol
34:2259–2264. doi:10.3174/ajnr.A3631
38. Brisman JL, Pile-Spellman J, Konstas AA (2012) Clin-
ical utility of quantitative magnetic resonance angiog-
raphy in the assessment of the underlying
pathophysiology in a variety of cerebrovascular disor-
ders. Eur J Radiol 81:298–302. doi:10.1016/j.
ejrad.2010.12.079
39. Ruland S, Ahmed A, Thomas K, Zhao MD, Amin-
Hanjani S, Du XJ, Charbel FT (2009) Leptomeningeal
collateral volume flow assessed by quantitative mag-
netic resonance angiography in large-vessel
232
cerebrovascular disease. J Neuroimaging 19:27–30.
doi:10.1111/j.1552-6569.2008.00249.x
40. Amin-Hanjani S, Du XJ, Zhao MD, Walsh K, Malisch
TW, Charbel FT (2005) Use of quantitative magnetic
resonance angiography to stratify stroke risk in symp-
tomatic vertebrobasilar disease. Stroke 36:1140–1145.
doi:10.1161/01.STR.0000166195. 63276.7c
41. Nael K, Meshksar A, Liebeskind DS, Coull BM,
Krupinski EA, Villablanca JP (2013) Quantitative anal-
ysis of hypoperfusion in acute stroke: arterial spin
labeling versus dynamic susceptibility contrast. Stroke
44:3090–3096. doi:10.1161/strokeaha. 113.002377
42. van Laar PJ, van der Grond J, Hendrikse J (2008) Brain
perfusion territory imaging: methods and clinical appli-
cations of selective arterial spin-labeling MR imaging.
Radiology 246:354–364. doi:10.1148/radio1.
2462061775
43. Bokkers RP, van Osch MJ, Klijn CJ, Kappelle LJ,
Hendrikse J (2011) Cerebrovascular reactivity within
perfusion territories in patients with an internal carotid
artery occlusion. J Neurol Neurosurg Psychiatry
82:1011–1016. doi:10.1136/jnnp.2010.233338
44. Tan JC, Dillon WP, Liu SL, Adler F, Smith WS,
Wintermark M (2007) Systematic comparison of
perfusion-CT and CT-angiography in acute stroke
patients. Ann Neurol 61:533–543. doi:10.1002/
ana.21130
45. Jung SL, Lee YJ, Ahn KJ, Kim YI, Lee KS, Shin YS,
Lee KS, Kim BS (2011) Assessment of collateral flow
with multi-phasic CT: correlation with diffusion
X. Leng and D.S. Liebeskind
weighted MRI in MCA occlusion. J Neuroimaging
21:225–228. doi:10.1111/j.1552-6569.2010.00496.x
46. Leng X, Wong KS, Soo Y, Leung T, Zou X, Wang Y,
Feldmann E, Liu L, Liebeskind DS (2013) Magnetic
resonance angiography signal intensity as a marker of
hemodynamic impairment in intracranial arterial stenosis.
PLoS One 8:e80124. doi:10.1371/journal.pone.0080124
47. Taylor CA, Fonte TA, Min JK (2013) Computational
fluid dynamics applied to cardiac computed tomogra-
phy for noninvasive quantification of fractional flow
reserve: scientific basis. J Am Coll Cardiol
61:2233–2241. doi:10.1016/j.jacc. 2012.11.083
48. Schirmer CM, Malek AM (2007) Prediction of complex
flow patterns in intracranial atherosclerotic disease using
computational fluid dynamics. Neurosurgery
61:842–852. doi:10.1227/01.neu.0000298914.32248.dc
49. Suh DC, Ko YB, Park ST, Yoon K, Lim OK, Oh JS,
Jeong YG, Kim JS (2011) Computational flow dynam-
ics of the severe M1 stenosis before and after stenting.
Neurointervention 6:13–16. doi:10.5469/neuroint.
2011.6.1.13
50. Goldstein LB, Bushnell CD, Adams RJ, Appel LJ,
Braun LT, Chaturvedi S, Creager MA, Culebras A,
Eckel RH, Hart RG, Hinchey JA, Howard VJ, Jauch
EC, Levine SR, Meschia JF, Moore WS, Nixon JV,
Pearson TA (2011) Guidelines for the primary preven-
tion of stroke: a guideline for healthcare professionals
from the American Heart Association/American Stroke
Association. Stroke 42:517–584. doi:10.1161/STR.
0b013e3181fcb238
 Imaging of Spontaneous
Nontraumatic Intracerebral 13
Hemorrhage

Stephen Quinet and Patrick Turski

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234 The focus of this chapter is to provide an over-
view of the most commonly encountered etiol-
Hypertensive Hemorrhage . . . . . . . . . . . . . . . . . . . . . . . . . 234
ogies of spontaneous intracerebral hemorrhage
Evolution of Hematoma on CT and MRI . . . . . . . . . 239 (SICH). Specifically, the goal is to review the
Coagulopathy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241 differential diagnosis of nontraumatic hemor-
rhage based on clinical presentation, imaging
Cerebral Amyloid Angiopathy . . . . . . . . . . . . . . . . . . . . 244
characteristics, and location of hemorrhage. In
T2* Gradient Echo Imaging . . . . . . . . . . . . . . . . . . . . . . . 248 addition to reviewing the various potential
Susceptibility-Weighted Imaging (SWI) . . . . . . . . . . 248 causes of nontraumatic intracranial hemor-
Nontraumatic Subarachnoid Hemorrhage . . . . . . . 251 rhage, this chapter will provide a review of
the variable appearance of hemorrhage on CT
Vascular Malformations . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
and MR. An emphasis will be placed on iden-
Hemorrhagic Transformation of Infarct . . . . . . . . . 270 tification and anticipation of the potential com-
Venous Thrombosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274 plications of acute intracranial hemorrhage.
Hemorrhagic Neoplasm . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
Keywords
Radiation-Induced Vascular Disease Spontaneous intracranial hemorrhage •
and Hemorrhage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
Nontraumatic intracranial hemorrhage •
Vasculitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285 Hypertensive hemorrhage • Coagulopathy •
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290 Cerebral amyloid angiopathy • Cerebral
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290 microbleeds • Cerebral microhemorrhages •
Nontraumatic subarachnoid hemorrhage •
Intracranial aneurysm • Arteriovenous malfor-
mation • Dural arteriovenous fistula • Cavern-
ous malformation • Hemorrhagic
transformation of infarct • Venous thrombosis
• Hemorrhagic neoplasm • Radiation
vasculopathy • Radiation microhemorrhage •
S. Quinet (*) • P. Turski Radiation microangiopathy • Central nervous
Department of Radiology, School of Medicine and Public
system vasculitis
Health, Clinical Science Center, University of Wisconsin,
Madison, WI, USA
e-mail: squine@uw.edu; pturski@uwhealth.org

# Springer Science+Business Media New York 2016 233

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_47
234
Introduction
This chapter provides a review of nine etiologies
of nontraumatic intracranial hemorrhage. Three of
these entities comprise the majority of
nontraumatic intracranial hemorrhage: hyperten-
sion (36 %), aneurysm (36 %), and arteriovenous
malformation (11 %). Other etiologies account for
17 % of nontraumatic hemorrhage, the more com-
mon of which are discussed in detail later in the
chapter (Table 1).
Hemorrhagic strokes represent 10–15 % of all
strokes and are characterized by the sudden onset
of a focal neurologic deficit due to intracranial
hemorrhage, which can be intraparenchymal
(IPH), subarachnoid (SAH), intraventricular
(IVH), or a combination of locations. IPH is
more likely to mimic acute ischemic stroke,
while subarachnoid hemorrhage classically pre-
sents with “thunder clap” headache. IPH will
more commonly manifest as seizures and nau-
sea/vomiting. However, there is considerable
overlap in the presenting symptoms for both IPH
and SAH.
Non-contrast-enhanced head CT is often the
imaging modality selected for the initial evalua-
tion of suspected intracranial pathology. The wide
availability in various clinical settings, the ability
to scan without antecedent safety screenings (as in
MR), and extremely rapid image acquisition and
reconstruction have made CT the imaging modal-
ity generally used in the setting of a neurologic
emergency. If there is high clinical suspicion for
vascular pathology, starting with advanced imag-
ing techniques, most often CTA or MRI/MRA can
be performed [1, 2]. Once intracranial hemorrhage
is detected, follow-up imaging with contrast-
enhanced CT and/or MR imaging, with dedicated
angiographic images (CTA/MRA) without or
Table 1 Etiologies of spontaneous intracerebral hemor-
rhage (SICH)
Location Frequency (%)
Hypertension 36
Aneurysm 36
Arteriovenous malformation (AVM) 11
Others 17
S. Quinet and P. Turski
with perfusion, is often warranted if the etiology
is not immediately obvious. The choice of modal-
ity will depend on patient issues (e.g., rapid clin-
ical deterioration, MR-incompatible medical
devices, uncooperative patient) and the most
highly suspected underlying pathology (e.g., vas-
cular lesion versus a hemorrhagic mass). In some
cases, the patient will be best assessed and poten-
tially treated by proceeding directly to conven-
tional catheter-directed angiography. Given the
appearance and distribution of the hemorrhage
and the clinical presentation, a reasonable short
differential diagnosis can be developed which will
guide selection of the appropriate follow-up imag-
ing modality.
Hypertensive Hemorrhage
Arteriopathy related to long-standing hyperten-
sion is the most common cause of spontaneous
nontraumatic intracranial hemorrhage in adults
[3]. Chronic systemic hypertension produces
degeneration of the deep penetrating arteriolar
branches of the major intracranial arteries on the
basis of intimal hyperplasia and medial degenera-
tion. These affected arterioles are the very small
vessels at termination points of the perforators
arising as end artery branches from the anterior
and posterior circulation. Over time, sustained
hypertension leads to deposition of hyaline, fibrin,
and collagen in the arteriolar walls, leading to
degeneration. Arteriolar sclerosis leads to
narrowing of the vessel lumen and ischemia,
with the ultimate end result being lacunar infarcts.
Additionally, as arteriolar walls degenerate, they
become weakened and begin to dilate partially on
the basis of lipid and collagen replacing arteriolar
smooth muscle cells, leading to formation of
so-called Charcot–Bouchard microaneurysms
(Fig. 1). These are not true aneurysms but rather
are more accurately classified as degenerative
fusiform dilations of chronically diseased vascu-
lature. These weakened, dilated walls are thought
to be more prone to rupture.
Hypertensive hemorrhages occur in predict-
able locations, a characteristic that is useful in
identifying chronic hypertension as the most
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage 235

Fig. 1 Acute hypertensive hemorrhage. (a) Axial CTA
maximum intensity projection image demonstrates
intraparenchymal hemorrhage centered within the right
caudate nucleus, with extension into the ventricular sys-
tem. A punctate enhancing structure in the right basal
ganglia represents an unruptured Charcot–Bouchard
microaneurysm (arrow). (b) Axial source images from
the CTA again show this enhancing punctate structure
(arrow). The coronal MIP reformat CTA image (c) shows
that the microaneurysm is at the termination of a
lenticulostriate artery

likely etiology for intracranial hemorrhage. About Table 2 Location of hypertensive hemorrhage
66 % occur in basal ganglia, particularly in the Location Frequency (%)
putamen and external capsule, 20 % within the Putamen/external capsule 60
thalamus, 10 % within the pons, 5 % within the Thalamus 20
cerebellum, and 5 % within the subcortical white Pons 10
matter (Table 2). It is important to note that while Cerebellum 5
the basal ganglia, thalamus, and brainstem are the Subcortical white matter 5
most common locations for hypertensive hemor-
rhage, this etiology must be considered in the
differential diagnosis of spontaneous lobar hem- locations, as detailed above. However, given the
orrhage, especially in the appropriate clinical set- volume of brain MR imaging done for myriad
ting (Fig. 2). potentially unrelated indications, sequelae of
In the acutely symptomatic patient, acute hem- chronic hypertensive arteriopathy are often inci-
orrhage will be identified in characteristic dentally detected. Specifically, T2* GRE and SWI
236
Fig. 2 CT images in four patients demonstrate the classic
locations of acute hypertensive hemorrhages, including the
(a) putamen, (b) thalamus, (c) pons, and (d) cerebellum.
sequences will reveal loss of signal and “bloom-
ing,” often multifocal, in the distributions of the
deep perforators. Multiple hypertensive
microhemorrhages (Fig. 3) may mimic the presen-
tation of cerebral amyloid angiopathy (CAA). This
topic is discussed in detail later in the chapter.
Some other potential associated findings seen in
the setting of acute hypertensive hemorrhage
S. Quinet and P. Turski
Note that the cerebellar hemorrhage has extended in to the
fourth ventricle. Dilation of the temporal horns of the
lateral ventricles indicates obstructive hydrocephalus
include loss of brain volume, a background of
chronic ischemic microvascular disease in the
white matter, and old small lacunar infarcts. It is
important to consider the possibility of illicit drug
use, especially cocaine and other sympathomimetic
drugs, as potential etiologies of acute intracranial
hemorrhage in typical locations of hypertensive
hemorrhage in otherwise healthy young patients.
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage
Fig. 3 Series of unenhanced head CT images obtained
from a 30-year-old male presenting with right-sided weak-
ness. Multiple hypodense foci were identified in the deep
gray matter and white matter, including within the white
matter of the right frontal lobe (a) and right caudate (b, c).
A large acute hemorrhage was identified centered within
the left putamen. Series of T2* MR images obtained in
The size of a hematoma on initial imaging and
its subsequent expansion portend a worse progno-
sis in this patient population. Expansion of hema-
toma is identified in approximately one-third of
cases of acute hypertensive hemorrhage and is
most likely to occur within several hours of symp-
tom onset. In the setting of active extravasation,
the “swirl sign” may be seen on unenhanced head
CT, characterized by hypodense acute liquid
blood products surrounding a hyperdense orga-
nizing clot. Although not routinely performed in
all institutions if there is a classic imaging
237
follow-up approximately 3 weeks after presentation. (d)
Multiple microhemorrhages are identified within the bilat-
eral basal ganglia and thalami. There has been evolution of
the dominant left putaminal hematoma. (b, c) Additional
microhemorrhages are identified in the central (e) pons and
(f) cerebellum
presentation of acute hypertensive hemorrhage,
dynamic combined CT/CTA imaging can be help-
ful in assessing the severity of the hemorrhage and
the relative urgency of surgical decompression
versus close clinical and imaging surveillance.
Specifically, delayed post-contrast CT imaging
performed shortly after the CTA acquisition is
useful to assess for active blood–brain barrier
breakdown. Contrast pooling within the hema-
toma seen on the delayed post-contrast imaging
is indicative of active arterial extravasation. Addi-
tionally, the so-called spot sign, a focus or foci of
238
Fig. 4 (a) Unenhanced CT shows large acute hemorrhage
centered in the left basal ganglia with associated rightward
mass effect, subfalcine herniation, and intraventricular
extension of hemorrhage. (b) MIP images from
arterial phase enhancement within the center of the
hematoma, is visualized in approximately
one-quarter of cases (Fig. 4) [4]. It has been spec-
ulated that this “spot sign” may actually represent
visualization of the culprit microaneursym within
the area of hemorrhage. The spot sign is present in
20–30 % of acute IPH patients presenting within
3–10 h of symptom onset. The spot sign provides
prognostic information, as up to 75 % of these
patients will go on to have extension of their hema-
tomas. Additionally, the presence of active contrast
extravasation on CTA (Fig. 5) is a predictor of poor
clinical outcome and increased in hospital mortal-
ity (OR >10). It should be noted that the sensitivity
and specificity of the spot sign appear to be depen-
dent upon the timing of CTA and whether delayed
post-contrast imaging is performed.
Factors associated with poor prognosis in
patients with acute intracranial hemorrhage
include age greater than 80 years, posterior fossa
location of hemorrhage, volume of hemorrhage
exceeding 30 mL, and intraventricular extension
of hemorrhage [5]. It is imperative to be aware of,
and actively search for, the potential secondary
effects of acute hypertensive hemorrhage
S. Quinet and P. Turski
subsequently obtained CTA demonstrate the “spot sign”
(arrow) – a focal blush of contrast extravasation centered
within the basal ganglia hemorrhage
including acute hydrocephalus, herniation syn-
dromes, and infarction. Many supratentorial
hematomas can be managed medically; however,
hematomas in the posterior fossa require close
monitoring, often necessitating timely neurosur-
gical decompression, given the potential for rapid
expansion of the hematoma and surrounding
edema within the compact confines of the poste-
rior fossa. There is limited ability of the posterior
fossa structures to enlarge without clinically sig-
nificant, often catastrophic, complications.
Supratentorial herniation, cerebellar tonsillar her-
niation, and obliteration of the fourth ventricle
with resultant obstructive hydrocephalus can
develop in an impressively rapid fashion.
Hypertensive hemorrhages tend to dissect
along the white matter tracts, so that the damage
to the brain tissue and associated neurologic dam-
age incurred by hypertensive hemorrhages are
often much less than the injury related to trau-
matic hemorrhagic contusions. These patients
will have a relatively good prognosis if they
have not experienced any of the above-described
severe secondary effects associated with the
hematoma.
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage 239

Fig. 5 Series of images obtained from a 41-year-old
woman presenting with acute hypertensive hemorrhage
(SBP >220 mmHg). (a) Unenhanced CT shows large
acute hemorrhage centered in the right basal ganglia with
associated mass effect, including rightward midline shift
and subfalcine herniation (Source images from CTA) (b),
and delayed post-contrast CT (c) demonstrate contrast
pooling within the area of hemorrhage (arrows) consistent
with active extravasation. (d) and (e): T2* GRE images
status post-craniectomy demonstrate punctate foci of
susceptibility in the right pons and left caudate and puta-
men consistent with additional microhemorrhages that
were occult on initial unenhanced CT. Blood products
line the site of the evacuated hematoma, in the region of
the external capsule and putamen. (f) Unenhanced head CT
performed 12 months after craniectomy shows expected
evolution of the surgical bed, with volume loss and
hypodensity in the right frontal lobe, subinsular region, as
well as sites of initial hemorrhage in the putamen and
external capsule

platelets) [6]. The attenuation of the hematoma

Evolution of Hematoma on CT and MRI may increase in the first few days after presenta-
tion (up to 80–100 HU) depending on the degree
On unenhanced CT, the increased attenuation on of organization of the clot. After the first few days,
CT (HU 50–60) within an acute hematoma is not an acute hemorrhage evolves in a relatively pre-
due to iron content, but rather it is due to clot dictable fashion on unenhanced CT related to its
retraction and the presence of dense protein con- hydration. As the hematoma evolves, it begins to
tent within the hematoma (e.g., fibrin, cells, draw in water, resulting in the margins of the
240
Fig. 6 Series of unenhanced head CT examinations dem-
onstrating evolution of an intraparenchymal hematoma. (a)
Initial head CT demonstrates large acute hematoma cen-
tered in the right temporoparietal region. Subsequent head
CT examinations performed (b) 11 days, (c) 23 days, and
(d) 29 days later demonstrate expected reduction in
hematoma beginning to decrease in attenuation.
As water is further drawn into the hematoma,
there is continued expected reduced attenuation
within the hematoma itself on the basis of
increased water content. On average, one can
expect the attenuation within the hematoma to
decrease by approximately 1.5 HU per day. An
S. Quinet and P. Turski
attenuation within the hematoma on sequential images
related to increased water content. Pre- and post-contrast
head CT (e) and (f) performed 84 days after the initial
hemorrhage demonstrate a small amount of residual
enhancement at the margins of the hematoma
element of inflammation is present at the margins
of the hematoma as it begins to resorb, leading to
breakdown of the blood–brain barrier. As such, a
peripheral rim of contrast enhancement can be
seen starting approximately 1 week after the initial
hemorrhage, which can persist for up to 8–12
weeks (Fig. 6).
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage
Table 3 Evolution of intracerebral hematoma on MR
Predominant blood
Stage product Time course
Hyperacute Oxyhemoglobin <6 h
Acute Deoxyhemoglobin 6–72 h
Early Intracellular <3 days to 1 week
subacute methemoglobin
Late Extracellular 1–2 weeks to
subacute methemoglobin months
Chronic Hemosiderin 2 weeks to years
MR imaging is even better suited to demon-
strate the evolution of hematoma based on the
signal characteristics of various stages of break-
down blood products and the anticipated time
points at which they appear and transform
(Table 3, Fig. 7) [7, 8]. Hyperacute hematoma
(<6 h) is predominantly composed of oxyhemo-
globin, demonstrating T1 isointense and T2
mildly hyperintense signal. Within an acute hema-
toma in slightly later evolution (7 h–3 days), the
blood products become deoxygenated and there is
conversion of oxyhemoglobin to
deoxyhemoglobin, resulting in hematoma demon-
strating T1 isointense-to-hypointense signal and
T2 markedly hypointense signal. In the early sub-
acute stage, as the blood products continue to
degrade from deoxyhemoglobin to methemoglo-
bin and the red blood cell remains intact, the
hematoma predominantly demonstrates character-
istic signal of intracellular methemoglobin, with
T1 marked hypointensity and T2 marked
hyperintensity. In the late subacute stage, after
the red blood cells have lysed, characteristic sig-
nal of extracellular methemoglobin is seen, with
T1 and T2 marked hyperintensity. In the chronic
state, there is further degradation of blood prod-
ucts and subsequent uptake by macrophages
resulting in deposition of hemosiderin and ferritin
within the hematoma cavity, with marked T1 and
T2 hypointensity.
Hematomas often contain blood products at
different stages of evolution, resulting in a mix-
ture of these imaging features, complicating
image interpretation, and sometimes establishing
chronicity of hemorrhage is difficult. The timing
241
Mass
T1 signal T2 signal effect
Isointense Mildly +++
hyperintense
Iso- to Hypointense +++
hypointense
Hyperintense Hypointense +++/++
Hyperintense Hyperintense +/
Hypointense Hypointense
of evolution of blood products detailed above is
most reliable for intraparenchymal hemorrhage.
Extra-axial hemorrhage tends to progress in a
similar order, but in a more delayed fashion, likely
on the basis of lower oxygen tension. Addition-
ally, subarachnoid and intraventricular hemor-
rhage signal will be affected by mixing with
CSF. Finally, it is possible that on occasion an
acute hematoma may be confused with infarction.
The core of an acute hematoma may be
hyperintense on DWI (representing “T2 shine
through”) with decreased ADC values
(representing “T2 dark through”), mimicking
arterial stroke [9]. However, correlation with
other sequences and/or CT should be helpful to
avoid this potential pitfall.
Coagulopathy
Spontaneous intracranial hemorrhage related to
coagulopathy, whether on the basis of hemato-
logic disease, metabolic disease, or therapeutic
anticoagulation, is an entity that must be
suspected in the appropriate clinical context and
is one that may demonstrate potential clues to its
diagnosis on imaging. Similar to patients with
low hematocrit, blood products within a hema-
toma in the setting of coagulopathy may be
isodense to brain parenchyma on CT imaging.
Identification of a fluid–fluid level within the
hematoma on CT is a characteristic finding of
intracranial hematomas in this patient population
(Figs. 8 and 9). These fluid–fluid levels can per-
sist for several hours after the hemorrhage.
242 S. Quinet and P. Turski

Fig. 7 (continued)
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage
Fig. 7 Evolution of hematoma on MRI. Multiplanar,
multisequence MR imaging demonstrates a typical left
basal ganglia hypertensive hemorrhage. Top row: Demon-
stration of late acute stage hematoma (1–2 days after hem-
orrhage). (a) Sagittal T1, (b) axial T2, (c) coronal T2
FLAIR, (d) axial T2* GRE, (e) and axial DWI sequences.
Several days out, expected signal changes associated with
deoxyhemoglobin are seen, with mildly T1 hypointense
and marked T2 hypointense signal; the T2-WI also shows
Similarly, on MR imaging, heterogeneous signal
within the hematoma cavity with identifiable
fluid–fluid levels can suggest underlying
coagulopathy. Peri-hematoma edema is a
243
expected edema surrounding the hematoma in the region of
the left basal ganglia. Bottom row: Demonstration of late
subacute stage hematoma (2–4 weeks after hemorrhage).
(f) Sagittal T1, (g) axial T2, (h) coronal T2 FLAIR, (i) axial
T2* GRE, and (j) axial DWI sequences. As the hematoma
continues to evolve over the course of several weeks, the
contents of the hematoma are mostly extracellular methe-
moglobin, demonstrating hyperintense signal on both T1-
and T2-WI images
nonspecific finding, but could identify potential
brain tissue that is at risk due to compression,
increased permeability, and inflammation related
to the hematoma (Fig. 10).
244 S. Quinet and P. Turski

Fig. 8 Coagulopathic hemorrhage. Series of unenhanced fluid–fluid levels (arrows), characteristic of intracranial
head CT images from inferior to superior (a–c) demon- hemorrhage in the setting of coagulopathy
strate a large right basal ganglia hemorrhage. Note

Fig. 9 Coagulopathic hemorrhage. Two unenhanced head heterogeneous attenuation and fluid levels (arrows)
CT images demonstrate a large amount of intraventricular involving the right lateral, third, and fourth ventricles
hemorrhage and resulting obstructive hydrocephalus, with

Cerebral Amyloid Angiopathy cerebral amyloid angiopathy (CAA) [10]. Beta

amyloid deposition at these sites increases with
Beta amyloid deposition within the small- and age, accounting for the estimated 20 % of sponta-
small-to-medium-sized arteries of the cortex and neous intracranial hemorrhage in the elderly. This
leptomeninges is the pathologic hallmark of entity is rare before age 60, but is a common cause
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage 245

Fig. 10 Coagulopathic hemorrhage. (a) Unenhanced
head CT demonstrates large right basal ganglia hemor-
rhage with fluid–fluid levels, characteristic of hemorrhage
in the setting of coagulopathy. (b) Axial maximum inten-
sity projection image from the concurrently performed
CTA demonstrates displaced MCA branches. (c–e) Cere-
bral blood flow maps performed part of the CTA exam
demonstrate reduced CBF within and surrounding the area
of hemorrhage. Sagittal T1 (f) and axial T2 (g) images
demonstrate heterogeneous hyperintense signal within the
hemorrhage. (h) T2 FLAIR imaging best demonstrates
degree of vasogenic edema surrounding the hematoma.
DWI (i) and ADC (j) demonstrate a large component of
“T2 shine-through” in the core of the hematoma, but also
true diffusion restriction at the anterior and medial portion
of the hematoma
246
Fig. 11 Lobar hemorrhage and microhemorrhages in
CAA. (a) Sagittal T1 image demonstrates right temporal
lobar hemorrhage with peripheral hyperintensity related to
methemoglobin formation. (b) Coronal T2 FLAIR image
demonstrates peripheral hyperintense rim of the lobar hem-
orrhage and minimal surrounding edema. Periventricular
and subcortical white matter signal abnormalities are also
seen. (c) Axial T2* image demonstrates hypointense signal
in the lobar hemorrhage but also reveals punctate
hypointense foci representing microbleeds in the
of spontaneous intracranial hemorrhage in the
elderly population, second only to hypertension.
CAA can present as lobar macrohemorrhage or as
multiple microhemorrhages, often recurrent
(Fig. 11). The hemorrhage associated with CAA
is often lobar or cortical or combined cortical/
subcortical in location [11]. Pathologically, beta
amyloid deposition is identified within the brain,
perivascular spaces, and outer walls of arterioles
and leptomeningeal vessels leading to fibrinoid
degeneration and necrosis leading to vascular fra-
gility, accounting for the association with hemor-
rhages. Microaneurysms have also been described
as resulting from this degenerative process, likely
also contributing to hemorrhage. Finally, vessel
lumen obliteration may result, leading to micro-
vascular ischemia. Characteristically, these
S. Quinet and P. Turski
subcortical left temporal lobe, left parahippocampal
region, and left occipital lobe. (d) Axial T2* image at a
more superior location demonstrates additional
microbleeds in the left temporal and occipital cortices and
subcortical white matter. (e) Axial T2 image demonstrates
peripheral hyperintense rim of the lobar hemorrhage. (f)
Axial DWI demonstrates hyperintense signal throughout
the hematoma. (g) ADC map demonstrates hypointense
signal throughout the hematoma on the basis of “T2 dark
through,” mimicking an acute infarct
microhemorrhages are most conspicuous as foci
of low signal on T2* GRE and SWI sequences
[12]. Leptomeningeal amyloid angiopathy can
initially present as acute blood layering within
the subarachnoid space. Superficial siderosis,
characterized by diffuse low signal and blooming
on susceptibility-weighted images along the pial
surfaces of the brain and within the basilar cis-
terns, can be seen on follow-up imaging, thought
to be related to recurrent subarachnoid hemor-
rhage (Fig. 12) [13].
The Boston criteria were established in 1990 in
order to standardize the diagnosis of CAA, using
clinical, imaging, and/or pathologic parameters
(Table 4). Definite cerebral amyloid angiopathy
can be diagnosed only postmortem, while proba-
ble CAA is defined in a patient older than
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage
Fig. 12 Superficial siderosis due to recurrent subarach-
noid hemorrhage in a patient with leptomeningeal amyloid
angiopathy (top row, imaging at presentation; bottom row,
follow-up imaging 7 years later). (a) Unenhanced head CT
at presentation demonstrates subarachnoid hemorrhage
layering within the sulci of the left frontal lobe. (b) Axial
T2* GRE image at the level of the midbrain demonstrates
punctate foci of susceptibility along the pial surface of the
right temporal and occipital lobes (arrows) representing
subarachnoid hemorrhage. (c) Axial T2* GRE image
shows punctate foci of susceptibility along the pial surface
55 years, in the setting of appropriate clinical
history, with MRI findings of multiple
microhemorrhages and lobar hemorrhage or con-
vexity subarachnoid hemorrhage without any
other explanation.
CAA coexists with Alzheimer dementia and
other dementias in 20–30 % of patients (Fig. 13).
Symptoms of dementia may be due to leukoaraiosis
resulting from progressive microvascular ischemia.
Unfortunately, as discussed above, advanced
247
of the left frontal and right parietal lobes (arrows)
representing additional subarachnoid hemorrhage. (d)
Axial T2* GRE image obtained 7 months later demon-
strates diffuse hemosiderin staining of the pial surface of
the brain bilaterally, representing superficial siderosis. (e)
Axial SWI image obtained at the same time of
d demonstrates the increased sensitivity of this sequence
for hemorrhage. (f) Axial SWI image more cephalad than
e again shows superficial siderosis to better advantage
compared to routine T2* GRE imaging
hypertensive arteriopathy can similarly present as
multiple small areas of hemorrhage, so there can be
substantial overlap between the imaging appear-
ance of CAA and advanced hypertensive changes
[14]. These two entities account for the majority of
cerebral microhemorrhages detected at imaging.
However, the differential diagnosis of multiple
microhemorrhages includes hemorrhagic metastatic
disease, vasculitis, diffuse axonal injury (DAI),
cerebral embolism (including fat emboli), multiple
248 S. Quinet and P. Turski

Table 4 Boston criteria for diagnosis of cerebral amyloid
angiopathy (CAA)
Diagnostic category Criteria
Definite CAA Postmortem examination:
Lobar, cortical, or cortical/
subcortical hemorrhage
Evidence of severe CAA
vasculopathy
Probable CAA with Supportive clinical history
pathologic evidence Pathologic specimen: lobar,
cortical, or cortical/
subcortical hemorrhage
Probable CAA Supportive clinical history
Age 55
MRI evidence of multiple
lobar, cortical, or cortical/
subcortical hemorrhages
without other explanation
Possible CAA Supportive clinical history
Age 55
MRI evidence of a single
lobar, cortical, or cortical/
subcortical hemorrhage
without other explanation
spin echo T1- and T2-WI sequences.
Susceptibility-weighted T2* images can be of
benefit in further characterizing intracranial hem-
orrhage/hematoma. Gradient echo T2* acquisi-
tions do not have a refocusing radiofrequency
pulse and are thus very susceptible to static vari-
ations in the magnetic field, such as that produced
by the paramagnetic properties of hemosiderin
[15]. This local field inhomogeneity presents as
loss of signal on T2* images. While T2* images
do not have a high contrast resolution and image
clarity of other pulse sequences, they are highly
sensitive to blood products (Fig. 17).
One pitfall in using MR to evaluate hematomas
is related to the signal intensity seen on diffusion-
weighted images. The core of an acute hematoma
can be hyperintense on DWI with decreased ADC
values, simulating an area of infarction.
Susceptibility-Weighted Imaging (SWI)

A more advanced imaging method, which enables

cavernomas, radiation-induced vasculopathy,
cerebral autosomal dominant arteriopathy with
subcortical infarcts and leukoencephalopathy
(CADASIL), and Parry–Romberg syndrome
(Figs. 14 and 15).
Inflammatory CAA is a condition which is
thought to represent an autoimmune response to
the deposited beta amyloid. The imaging manifes-
tation of inflammatory CAA is distinct from those
of multiple microhemorrhages classically associ-
ated with CAA. Inflammatory CAA may manifest
as extensive white matter edema, perivascular
inflammation, and vasculitis (Fig. 16). The imag-
ing appearance of inflammatory CAA may very
closely mimic acute hypertensive encephalopa-
thy/posterior reversible encephalopathy syn-
drome (PRES).
T2* Gradient Echo Imaging
GRE T2* images are significantly more sensitive
to both hyperacute hemorrhage and
microhemorrhages compared to conventional
increased sensitivity to the detection of
microhemorrhages superior to even that of T2*
GRE (Fig. 18), is susceptibility-weighted
imaging (SWI) and susceptibility-weighted angi-
ography (SWAN) [16]. The basis of SWI is a
high-resolution, flow-compensated, three-
dimensional gradient echo sequence
[17–19]. SWI combines magnitude images and
phase shift information related to T2* dephasing
to exploit field inhomogeneity to increase the
ability to detect foci of susceptibility (e.g., cortical
veins, calcification, hemosiderin). Variations in
the local magnetic field result in dephasing of
signal and reduction of T2*. Nonheme-related
iron can be detected and may be useful for the
detection and surveillance of diseases that result in
iron deposition in the brain, such as Parkinson,
Huntington, and Alzheimer dementias. SWI per-
formance and sensitivity are positively related to
field strength.
Using postprocessing techniques, following
application of a high-pass filter to reduce aliasing,
a phase mask is applied to the magnitude images
to increase the conspicuity of foci of
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage 249

Fig. 13 A 77-year-old male with rapidly progressive
dementia, diagnosed with Alzheimer dementia 2 years
prior to this imaging. Unenhanced head CT (a) and T1
(b), T2 (c), and T2 FLAIR (d) MR images demonstrate the
classic appearance of dilated temporal horns of the lateral
ventricles due to hippocampal and medial temporal lobe
atrophy characteristically seen in Alzheimer-type
dementia. Multiple axial T2* GRE images (e–g) demon-
strate innumerable foci of signal loss consistent with
microhemorrhages, most pronounced in the left parietal
lobe, in a cortical–subcortical distribution typical of
CAA. Hippocampal and temporal lobe volume loss, seen
to better advantage on the previous figure, again is com-
patible with Alzheimer dementia
250
Fig. 14 Differential diagnosis of multiple bilateral
microhemorrhages. (a) Axial T2* GRE image demon-
strates punctate microhemorrhages in bilateral basal
ganglia in a patient with hypertensive arteriopathy. (b)
Axial T2* GRE image demonstrates multiple punctate
foci of susceptibility bilaterally at cortical and cortical–-
subcortical junction locations as well as a larger focus in
the right temporal lobe. Findings favored to represent
S. Quinet and P. Turski
radiation-induced telangiectasias and a cavernoma, respec-
tively. (c) Axial T2* GRE image demonstrates
microhemorrhages in the subcortical white matter of the
right frontal and temporal lobes as well as the left thalamus
in a patient with metastatic melanoma. (d) Axial post-
contrast T1-WI (same patient as c) demonstrates enhance-
ment of the right temporal (white arrow) and left thalamic
(black arrow) lesions
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage
Fig. 15 Patient with cerebral autosomal dominant
arteriopathy with subcortical infarcts and leukoence-
phalopathy (CADASIL). (a) Axial T2* GRE image dem-
onstrates microhemorrhages in bilateral thalami in a patient
with CADASIL. (b) Axial T2* GRE image at a slightly
susceptibility. For paramagnetic substances, the
increase in magnetic field leads to a negative
phase (right-handed system) relative to the sur-
rounding parenchyma and CSF. At the worksta-
tion, the separately acquired phase shift
information and the magnitude images combined
into a single image can be displayed using mini-
mum intensity projections (MinIP) for ease of
interpretation. The MinIPs are useful in
confirming the vascular nature of tortuous foci of
susceptibility by demonstrating contiguity with
other vascular structures. SWI has been reported
to provide twofold to fivefold increased sensitiv-
ity for the detection of microbleeds versus routine
MPGR T2* sequences.
An important clinical use of SWI is
distinguishing diamagnetic materials (e.g., cal-
cium) from paramagnetic materials (e.g., hemo-
siderin) [20]. Calcium and hemosiderin may both
present as increased attenuation on unenhanced
CT images and as foci of signal hypointensity and
blooming on conventional gradient echo T2*WI
images. Diamagnetic materials such as calcium
appear as high signal on phase images, allowing
distinction from paramagnetic hemosiderin.
251
higher location again shows the thalamic
microhemorrhages. (c) Axial T2-WI in the same patient
shows multiple hyperintense lesions in the periventricular
white matter bilaterally superimposed on diffuse confluent
deep white matter signal abnormality
Nontraumatic Subarachnoid
Hemorrhage
Nontraumatic subarachnoid hemorrhage (SAH) is
commonly due to three etiologies: arterial aneu-
rysm rupture, extension of intraparenchymal hem-
orrhage into the subarachnoid space, or
perimesencephalic SAH (likely venous). Aneurys-
mal SAH represents the vast majority (>85 %) of
nontraumatic SAH. Intracranial aneurysms occur
in approximately 2 % of the asymptomatic popu-
lation. In order of descending frequency, the most
common sites of intracranial aneurysms are the
anterior communicating artery (AComm) (30 %),
the posterior communicating artery (PComm)
(25 %), the middle cerebral artery (MCA) bifurca-
tion/trifurcation (20 %), the internal carotid artery
(ICA) bifurcation/terminus (7.5 %), basilar tip
(7 %), pericallosal arteries (4 %), and the posterior
inferior cerebellar artery (3 %). Additional miscel-
laneous aneurysm locations account for approxi-
mately 3.5 % of aneurysms (Table 5). While this
chapter aims specifically to review the etiologies of
nontraumatic SAH, both the radiologist and the
252
Fig. 16 Inflammatory cerebral amyloid angiopathy. (a)
Axial T2 FLAIR image demonstrates hyperintense signal
in the subcortical white matter of the left frontal lobe. (b)
Axial T2 FLAIR image a more inferior level in the same
patient demonstrates subcortical white matter signal abnor-
mality in the right temporo-occipital region. (c) Axial post-
contrast T1-WI at the same level of a demonstrates
non-enhancing hypointense signal corresponding to the
FLAIR abnormality in the right frontal lobe. Abnormal
leptomeningeal enhancement is present surrounding the
signal abnormality. (d) Axial post-contrast T1-WI at the
clinical service should be alert to the clinical sce-
nario of ruptured aneurysm and its concomitant
subsequent neurologic deficit leading to the initial
traumatic event. Thus, if the source of trauma is not
clear, further evaluation with dedicated vascular
imaging should be considered, as detailed below.
For reasons described in the introduction,
unenhanced CT is often the modality selected to
S. Quinet and P. Turski
same level of b demonstrates non-enhancing hypointense
signal corresponding to the T2 FLAIR abnormality in the
left temporal and occipital lobes. Abnormal
leptomeningeal enhancement is present surrounding the
signal abnormality. (e) T2* GRE image at a similar level
to a and c demonstrates punctate foci in the cortex and
juxtacortical locations in the left frontal lobe representing
microhemorrhages. (f) T2* GRE image at a similar level to
b and d demonstrates punctate foci in the cortex and
juxtacortical locations in the right temporal and parietal
lobes representing additional microhemorrhages
evaluate for SAH. On non-contrast-enhanced head
CT, the hallmark of SAH is the identification of
hyperdense hemorrhage within the basal cisterns,
interstices of the sulci, and the interhemispheric
fissure. Of note, aneurysmal rupture may also result
in intraparenchymal hemorrhage in up to one-third
of cases. Rare intraparenchymal hemorrhage with-
out associated SAH has also been described in
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage
Fig. 17 Intraparenchymal hemorrhage of unknown etiol-
ogy in a 24-year-old woman. (a) Initial unenhanced head
CT demonstrates large parenchymal hemorrhage centered
in the right parietal lobe. Subarachnoid hemorrhage layers
along the sulci of the right parietal and left frontal lobes.
relation to aneurysm rupture. The aneurysm itself
may occasionally be visible on non-contrast-
enhanced CT as a hyperdense round or lobular
structure, a filling defect surrounded by SAH in a
basilar cistern or fissure, and/or by mural calcifica-
tions, if the aneurysm is long standing. Classifica-
tion systems (Fig. 19, Tables 6 and 7) can be used to
describe both the imaging findings (Fisher grading
scale) and the clinical presentation (Hunt and Hess).
Although CT is very sensitive for the detection
of SAH, if only a small amount of blood is present
in the subarachnoid space, it may be difficult to
253
(b) T2* MR image demonstrates susceptibility-related sig-
nal loss within the hematoma. (c–d) Subtracted three-
dimensional phase-contrast MR angiography demonstrates
excellent subtraction of the hyperintense hematoma pre-
sent on the MR images
detect with CT. Additionally, the sensitivity of
unenhanced CT for SAH decreases as time since
symptom onset increases, related to both evolu-
tion of blood products and dilution of blood
within the CSF. Using MR imaging, fluid attenu-
ation inversion recovery (T2 FLAIR) sequences
are a very sensitive method for detecting SAH
[21]. However, while FLAIR imaging can be uti-
lized to detect very small amounts of blood in the
subarachnoid space in both the acute and subacute
setting, it is not very specific. Potential etiologies
for hyperintense signal in the subarachnoid space
254 S. Quinet and P. Turski

Fig. 18 Comparison of T2* and SWI. (a) Axial T2* GRE image in the same patient shows the increased conspicuity
image demonstrates extensive multiple small hemorrhages of this sequence for microbleeds
related to disseminated aspergillosis. (b) Axial SWAN

Table 5 Frequency of intracranial aneurysm by location Furthermore, if using two-dimensional image

Location Frequency (%) acquisition, inflow enhancement of the CSF
AComm 30 within the subarachnoid space can also have a
PComm 25 similar appearance; this effect is eliminated with
MCA bifurcation/trifurcation 20 use of a 3D acquisition technique. Finally, incom-
ICA bifurcation/terminus 7.5 plete nulling of CSF signal may give false-
Basilar artery tip 7 positive findings in the subarachnoid space, par-
Pericallosal arteries 4 ticularly a problem in the basilar cisterns.
Posterior inferior cerebellar artery 3 The pattern of SAH may suggest the location
Others 3.5 of a culprit aneurysm. For example, SAH concen-
trated in the anterior interhemispheric fissure sug-
gests an aneurysm arising from the AComm, and
on FLAIR imaging other than SAH include SAH asymmetrically involving one of the Sylvian
increased protein content within the subarachnoid fissures implicates the ipsilateral MCA. When
space (e.g., meningitis), retained circulating detected, the patient usually proceeds to CTA
gadolinium-based contrast material from a prior expeditiously to identify a potential culprit aneu-
contrast-enhanced scan in a patient with renal rysm for treatment planning. CTA is currently the
failure, high oxygen tension (e.g., patients on or imaging modality of choice in working up acute
near 100 % supplemental oxygen), and intubated nontraumatic SAH [22, 23]. Regarding detection
or sedated patients receiving propofol. Metal arti- of SAH and aneurysms >3 mm, CT and CTA
fact (e.g., braces, ventriculostomy shunt catheter demonstrate a sensitivity of 98–100 %, specificity
hardware) can interfere with the inversion pulse of 100 %, positive predictive value of 98–100 %,
applied during the FLAIR sequence, resulting in negative predictive value of 96–100 %, and accu-
incomplete nulling of the cerebrospinal fluid. racy of 99 %. Smaller aneurysms <=3 mm may
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage 255

Fig. 19 Series of
unenhanced head CT
examinations with
subarachnoid hemorrhage
of varying degrees of
severity demonstrating
Fisher grades 2–4. (a–b)
Fisher grade 1. Thin
subarachnoid hemorrhage
layers within the suprasellar
cistern, anterior
interhemispheric fissure,
and the Sylvian fissures.
(c–d) In a different patient,
a similar pattern of
subarachnoid hemorrhage
in the basilar cisterns is
seen, but the presence of a
focal clot in the
interhemispheric fissure
upgrades this case to Fisher
grade 2. (e–f) The presence
of intraventricular
hemorrhage (causing
obstructive hydrocephalus,
in this case) results in this
case being classified as
Fisher grade 4. Note the
presence of right posterior
approach ventriculostomy
catheter placed to
decompress the dilated
lateral ventricles
256
Table 6 Fisher grading scale for subarachnoid hemor-
rhage (SAH)
Grade CT findings
1 No subarachnoid hemorrhage
2 Diffuse blood
Clot <3 mm
Thin layer <1 mm
3 Clot >3 mm
Thick layer >1 mm
4 Intraventricular hemorrhage
Table 7 Hunt and Hess grading scale for subarachnoid
hemorrhage (SAH)
Grade Clinical findings
1 Asymptomatic or minimal headache and slight
nuchal rigidity
2 Moderate-to-severe headache, nuchal rigidity,
and no neurologic deficit other than cranial
nerve palsy
3 Drowsiness, confusion, of mild focal deficit
4 Stupor, moderate-to-severe hemiparesis, and,
possibly, early decerebrate rigidity and
vegetative disturbances
5 Deep coma, decerebrate rigidity, moribund
appearance
be difficult to detect using CTA due to limits of
spatial resolution of the modality. A thorough
search for additional asymptomatic aneurysms
once the symptomatic lesion is identified is cru-
cial, as intracranial aneurysms are multiple
approximately 20 % of the time.
Several different display techniques are used
during the interpretation of CTA images for SAH
and/or aneurysm work-up. Maximum intensity
projection (MIP) images are generated from the
axial source images. The source images also can
be reconstructed into three-dimensional
(3D) models, providing the ability to show the
aneurysm in more detail. Specifically, the 3D
reconstructions can be rotated to best demonstrate
the size and morphology of the aneurysm
[24]. Furthermore, these reconstructions offer the
ability to delineate the relevant intracerebral vas-
cular anatomy important for presurgical/
preprocedural planning. On CTA, aneurysms are
identified as contrast-opacified outpouchings
S. Quinet and P. Turski
arising from the parent vessel wall. It is imperative
to evaluate both source images, MIPs, and recon-
structions as important morphologic characteris-
tics of the aneurysm may be best illustrated on
only one these displays (Fig. 20).
Aneurysms can be subdivided by morphology,
including saccular (most commonly responsible
for aneurysmal SAH), mycotic or oncotic,
dissecting, fusiform, and giant (>20 mm). The
morphology of the neck of the aneurysm (e.g.,
broad vs. narrow) determines whether the aneu-
rysm would be more amenable to coiling versus
clipping. Likewise, the fundus-to-neck ratio may
be of benefit for pretreatment planning. Further-
more, the presence of calcifications at the aneurysm
neck and the presence of vessels arising from the
dome of aneurysm also impact treatment. Addi-
tional descriptors include pear-shaped, spherical,
beehive, blister-like, and lobulated. Occasionally,
an aneurysm be partially or completely thrombosed
on initial or follow-up imaging. Giant aneurysms,
defined as aneurysms measuring greater than
20 mm, are more likely to present on the basis of
symptoms due to mass effect, rather than due to
SAH/hemorrhagic rupture.
The radiologist must be aware of potential emer-
gent complications associated with acute SAH.
Acute hydrocephalus can develop very rapidly in
these patients. Particular attention should be paid
to the temporal horns of the lateral ventricles,
which should be only a few millimeters in diameter
in younger patients. The temporal horns can
dilate very quickly, along with the third ventricle
in the setting of acute SAH, and thus should be
scrutinized closely in follow-up examinations, as
these may be the earliest indicators of developing
hydrocephalus.
The annual rate of aneurysmal rupture for a
single aneurysm is approximately 0.7–1 %.
Thus, aneurysm risk stratification is an important
part of initial and follow-up imaging of patients
with intracranial aneurysms. Findings associated
with higher risk of aneurysmal rupture include
growth, PComm or posterior circulation location,
tobacco smoking, initial size >5 mm at presenta-
tion, and the presence of a daughter aneurysm or a
multi-lobulated appearance (Figs. 21 and 22)
[25–29]. Thus, despite a widely used threshold
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage 257

Fig. 20 (continued)
258
Fig. 21 Ruptured bilobed AComm aneurysm. Initial
unenhanced head CT images (a, b) demonstrate extensive
subarachnoid hemorrhage predominantly in the basilar
cisterns, anterior interhemispheric fissure, and Sylvian fis-
sure. A small amount of hemorrhage layers dependently
within the frontal horn of the left lateral ventricle. Axial
source images from subsequently performed CTA (c) dem-
onstrate a bilobed AComm aneurysm. Sagittal maximum
of 7 mm for treatment of intracranial aneurysms,
continued imaging surveillance of patients with
aneurysms <7 mm is warranted. Clinically,
hypertension and tobacco use are associated with
increased risk of aneurysm rupture.
Fig. 20 A 53-year-old woman presenting with the worst
headache of her life. (a–c) Series of unenhanced axial CT
images from inferior to superior demonstrate hyperdense
subarachnoid hemorrhage within the basilar cisterns,
including the suprasellar cistern, interpeduncular fossa,
and perimesencephalic cistern. Blood is also present within
bilateral, left-greater-than-right, Sylvian fissures. The tem-
poral horns of the lateral ventricles are dilated representing
acute communicating obstructive hydrocephalus. (d–e)
S. Quinet and P. Turski
intensity reconstructed CTA image (d) shows that the
aneurysm extends from the cephalad aspect of the
AComm. Rotated three-dimensional reconstructed image
from the CTA source data (e) shows to better advantage the
bilobed morphology of the aneurysm. Lateral projection
digital subtraction angiography via a right ICA injection (f)
again shows the aneurysm
Advanced vascular imaging techniques are
able to determine and quantify flow within aneu-
rysms and arteriovenous malformations (AVMs).
Research is ongoing regarding the stratification of
patients into high- and low-risk categories for
Sequential axial source images from CTA demonstrate a
multi-lobulated right superior hypophyseal artery aneu-
rysm (arrows), extending posteriorly and inferiorly into
the parasellar region. (f) Three-dimensional reformatted
image reconstructed from CTA source image best demon-
strates a bleb/daughter aneurysm arising from the posterior
aspect of the aneurysm (arrow). The bleb/daughter aneu-
rysm area is at high risk for rupture
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage 259

Fig. 22 Risk factors for

aneurysm rupture and
hemorrhage. (a) Time-of-
flight MRA image
demonstrates an inflow jet
within an AComm
aneurysm. (b) Axial source
images from CTA
demonstrate a lobular
AComm aneurysm with a
small bleb at its anterior
margin. (c) MIP images
from the same study
demonstrate the bleb to
better advantage (arrow).
(d) Three-dimensional
reconstruction from CTA
demonstrates the
morphology of the
aneurysm with superb detail

spontaneous aneurysmal rupture. Phase-contrast
MRA (4DFlow) is emerging as a method to char-
acterize the flow features of intracranial aneu-
rysms. It has been proposed that regions of low
wall shear stress (slow flow along the wall) lead to
aneurysm wall degeneration and inflammation
contributing to aneurysm growth. Complex
flow patterns with multiple internal vortices and
discrete inflow jets likely promote aneurysm
rupture.
Another clinical scenario is the patient
presenting with a very small amount of SAH that
cannot be detected by CT, the so-called “sentinel”
headache or hemorrhage (Fig. 23). Occasionally,
initial multimodality work-up of SAH may be
falsely negative on the basis of only a small
amount of aneurysmal hemorrhage, thrombus
within the aneurysm, and/or vasospasm occurring
during CTA and/or conventional angiography.
The diagnosis can only be made after repeat diag-
nostic imaging, CTA and/or DSA, after the acute
episode resolves. Alternatively, the patient may
present with repeat symptomatic SAH, obviously
necessitating earlier repeat imaging work-up.
Therefore, lumbar puncture remains the best
method to exclude CT-negative SAH in patients
in whom this entity is strongly suspected
clinically.
Mycotic aneurysms represent an infectious
process that involves the arterial wall, usually a
260
Fig. 23 Sentinel hemorrhage. (a) Unenhanced CT dem-
onstrates hyperdense subarachnoid blood layering along
the sulci of the right temporal lobe. (b) Unenhanced head
CT slightly more cephalad than a demonstrates blood
within the suprasellar cistern. (c) T2 FLAIR MR image
demonstrates hyperintense subarachnoid hemorrhage
along the sulci of the left temporal lobe and within the
suprasellar cistern, as seen on CT. (d) T2 FLAIR MR
image at a more cephalad location also reveals subarach-
noid blood along the sulci of the right parietal lobe more
posteriorly. Follow-up CTA examination in patient with
S. Quinet and P. Turski
sentinel hemorrhage. (e) Initial pre-contrast imaging dem-
onstrates accumulation of subarachnoid blood predomi-
nantly within the right suprasellar cistern and interposed
between the medial temporal lobe and the brainstem. (f)
Axial source images from CTA demonstrate a saccular
aneurysm from the right PComm. (g) Maximum intensity
projection (MIP) images show the right PComm aneurysm
to better advantage. (h) Conventional catheter digital sub-
traction angiography in the AP projection demonstrates
vasospasm at the time of diagnosis indicating prior sentinel
hemorrhage
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage
Fig. 24 Perimesencephalic subarachnoid hemorrhage.
(a–c) Series of unenhanced head CT images from inferior
to superior demonstrate characteristic location of hemor-
rhage within the interpeduncular fossa and wrapping
distal branch. Mycotic aneurysms may present as
acute intracranial hemorrhages in locations atypi-
cal for aneurysmal hemorrhage and may raise
suspicion for an underlying AVM, other vascular
malformation, or hemorrhagic mass.
Nontraumatic SAH may be non-aneurysmal in
etiology, with perimesencephalic SAH an impor-
tant differential diagnostic consideration, albeit
one which is a diagnosis of exclusion. This entity
has been reportedly responsible for up to
two-thirds of angiographically negative SAH.
Perimesencephalic SAH often presents as a
small amount SAH within the prepontine cistern
and extending along the perimesencephalic cistern
(Fig. 24). Occasionally, SAH is seen in basal por-
tions of the Sylvian fissures. This appearance may
mimic the rupture of a basilar tip aneurysm.
Patients presenting with perimesencephalic SAH
characteristically follow a more benign clinical
course compared to those presenting with SAH
due to other causes. Perimesencephalic SAH rarely
produces vasospasm and is felt likely to be venous
in origin. Anomalies of the basal vein of Rosenthal
have been found to be associated with perimesen-
cephalic SAH. Perimesencephalic SAH is
261
around the perimesencephalic cistern bilaterally. The
patient subsequently underwent two conventional catheter
angiographic examinations which were normal
essentially a diagnosis of exclusion, arrived at
after excluding an arterial source of the bleed. An
initial CTA is usually the modality of choice to
search for culprit lesions. Whether to follow a
negative CTA with conventional angiography
shortly after the initial CTA versus only follow-up
with another CTA is controversial and varies based
on personal, institutional, and/or regional prefer-
ence. However, at a minimum, at least one follow-
up CTA examination is routinely performed sev-
eral weeks out, after the SAH has cleared, to search
for underlying aneurysm or vascular malformation
that was potentially occult or obscured on initial
imaging.
Pseudosubarachnoid hemorrhage is an imag-
ing phenomenon related to marked, severe cere-
bral edema and swelling, seen in the setting of
global anoxic injury. Imaging features of cerebral
edema are seen, with loss of gray–white
differentiation, compression of sulci, and
hypoattenuating cerebral cortex (Fig. 25). As the
brain swells, there is hyperemia and congestion of
the meninges at the base of the brain. The
congested meninges appear to demonstrate
increased attenuation relative to brain tissue.
262
Fig. 25 Pseudosubarachnoid hemorrhage in a patient
with anoxic injury. (a–f) Series of unenhanced CT images
in a patient presenting after an anoxic event. Note the
diffuse hypoattenuation throughout the cortex and subcor-
tical white matter, with loss of gray–white differentiation,
However, the meninges are relatively normal in
attenuation, but appear prominent and engorged
on the basis of congestion. The appearance of
congested basilar meninges juxtaposed to brain
tissue markedly reduced in attenuation accentuates
the attenuation difference between these structures,
falsely giving the appearance that there is diffuse
hyperattenuation in the subarachnoid space.
Vascular Malformations
Arteriovenous malformations (AVMs) and other
vascular malformations are the most common
causes of spontaneous intracranial hemorrhage in
young adults. Subtypes of AVM include pial
S. Quinet and P. Turski
and effacement of sulci and gyri, indicating diffuse cere-
bral edema. The congested meninges of the basal cisterns
appear thickened and hyperdense juxtaposed against the
edematous brain parenchyma
AVMs and dural arteriovenous fistulas (AVFs). Cav-
ernous malformations (cavernomas), developmental
venous anomalies (DVAs), and capillary telangiec-
tasias are other vascular malformations that are also
common causes of nontraumatic intracranial hem-
orrhage in the young adult population and will be
also covered in this section. Though capillary tel-
angiectasias may arise on a congenital or sponta-
neous basis, given the increased incidence of these
lesions in the postradiation treatment setting, the
imaging appearance of capillary telangiectasias is
covered in the section of radiation-induced vascu-
lar malformations.
Pial AVMs are developmental lesions that tend
to present in younger adults, aged 30–60 years.
Fifty percent of patients with pial AVMs present
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage
Fig. 26 Pial arteriovenous malformation in a 32-year-old
man presenting with headache and nausea/vomiting. (a)
Axial unenhanced head CT demonstrates intrapar-
enchymal hemorrhage centered in the right frontal lobe,
with extension into the right lateral ventricle. (b) Axial
source images from subsequent CTA demonstrate
Fig. 27 A 65-year-old male presents with newly diag-
nosed pial AVM presenting as spontaneous intracranial
hemorrhage. Presenting symptoms included severe head-
ache, nausea/vomiting, and lethargy upon awakening. (a)
Axial image from initial unenhanced head CT demon-
strates intraventricular hemorrhage, obstructive hydro-
cephalus, and a partially calcified lesion near the
ependymal surface of the left lateral ventricle. (b) Axial
source image from subsequently obtained CTA at a similar
with intraparenchymal hemorrhage. Pial AVMS
are defined by an anomalous connection between
pial arteries and veins in the absence of an inter-
vening capillary bed [30]. The classic imaging
appearance of AVM includes the presence of mul-
tiple hypertrophied arterial feeding vessels, a
263
expansion of the right frontal lobar hemorrhage. Note is
made of a tangle of abnormal vascular structures at the
superolateral aspect of the hematoma. (c) Lateral projec-
tion from catheter angiography via right internal carotid
injection demonstrates arterial vessels supplying a nidus
along with early opacification of the draining veins
level demonstrates hypertrophied feeding arterial vascular
structures arising from the left MCA distribution, a large
partially calcified nidus, and a dominant draining vein
emptying into the left cerebral vein. (c) Lateral projection
from catheter angiography via left ICA injection demon-
strates en passant angular artery supplying the AVM nidus,
but also continuing on to supply normal brain tissue. A few
Rolandic branches are seen at the cephalad aspect, feeding
the AVM nidus
prominent nidus which can be partially calcified,
and large draining veins (Figs. 26 and 27). On
MR, flow voids can be seen within the larger,
more high-flow vascular components of the
AVM. Arterial feeders to pial AVMs may arise
from either branches of the ICA or the
264
vertebrobasilar system. Aneurysms may be asso-
ciated with the arteries supplying the nidus. The
intraparenchymal nidus is often surrounded by
gliotic brain tissue.
While spontaneous intracranial hemorrhage is
the most typical presentation of patients with pial
AVMs, these lesions may less commonly be
detected in patients presenting with seizures or
undergoing work-up for other clinical indications.
It is important to recognize that CTA, MRA, and
even conventional angiography may be falsely
negative in the setting of acute hematoma related
to a pial AVM due to mass effect and compression
of the hematoma on the supplying arteries, the
nidus, and/or the draining vein. Thus, repeat angio-
graphic imaging should be performed after resolu-
tion of the hematoma if there is strong clinical
suspicion for this entity. The Spetzler–Martin grad-
ing scheme attempts to stratify AVMs based on
surgical risk and takes into account the size of the
AVM, whether the lesion involves eloquent cortex,
and the pattern of venous drainage (e.g., superficial
and/or deep). The risk of hemorrhage of an
unruptured AVM is about 2 % per year, while the
chance of recurrent hemorrhage related to a previ-
ously hemorrhagic AVM has been reported as up to
18 % in the first year after the initial hemorrhage.
Additional risk factors for hemorrhage include
older age at presentation, exclusively deep venous
drainage, and deep location of the AVM.
Dural AVFs account for approximately 10–15 %
of intracranial AVMs (accounting for 6 % of
supratentorial and 35 % of infratentorial vascular
malformations) and are characterized by anomalous
arteriovenous shunting predominantly between
supplying meningeal arteries and draining dural
venous sinuses or meningeal or cortical veins
[31]. In contradistinction to pial AVMs, no intrapar-
enchymal nidus is present in dAVFs. In patients
with dAVF, venous ectasia is the finding most
highly associated with risk of hemorrhage using
the Cognard classification scheme (60 %). Patients
with dAVFs not drained by cortical veins, and there-
fore at lower risk of hemorrhagic complications,
may initially present clinically with symptoms
related to dural venous hypertension. Involvement
of the cavernous sinus and posterior fossa is most
common, though dAVFs may involve any supra- or
S. Quinet and P. Turski
infratentorial location, as well as the spinal canal.
Cavernous–orbital syndrome (e.g., exophthalmos)
may result from dAVF involving the cavernous
sinus, while pulsatile tinnitus and/or a clinical
sign of a bruit may be a clinical manifestation of
dAVFs involving the transverse and/or sigmoid
sinuses.
Dural AVFs can be idiopathic, congenital, or
acquired lesions. In the acquired setting, these
lesions often arise due to prior occlusion of a
dural venous sinus. Venous hypertension, prior
craniotomy, and trauma have also been found to
be associated with the development of dAVF.
While CTA and MRA are helpful in the evaluation
and diagnosis of dAVF, conventional angiography
remains the gold standard diagnostic modality, as
the classification and the risk of subsequent hem-
orrhage are predicated on the pattern of venous
drainage of these lesions (Fig. 28). Additionally,
angiography-directed catheter-based therapy is
often the treatment modality of choice for uncom-
plicated dAVFs. Importantly, the presence of
transosseous vascular channels should alert one
to the potential presence of a dural AVF.
The Borden and Cognard classification schema
have been developed to stratify dAVFs into
“benign” and “aggressive” subtypes, based on
the site of venous drainage (specifically, the pres-
ence or absence of cortical venous drainage), the
direction of flow within the involved or adjacent
dural sinus, and venous outflow architecture
(Tables 8 and 9). These subtypes are intended to
predict the likelihood of hemorrhage as well as
nonhemorrhagic neurologic deficits. In consider-
ing these subtypes, it is important to note that
there is rare progression of “benign” subtypes to
“aggressive” ones (approximately 2 %). Similarly,
spontaneous thrombosis and regression of dAVFs
have also been described. As such, close clinical
and imaging follow-up is necessary to exclude
progression of existing “benign” lesions, espe-
cially in the setting of new symptoms referable
to these lesions, which may herald a change in the
pattern of venous drainage.
Carotid-cavernous fistulas (CCFs) represent
abnormal communication between the carotid arte-
rial system and the cavernous sinus. CCFs are
subdivided into direct, high-flow CCF (type A)
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage 265

Fig. 28 (continued)
266
Fig. 28 Dural arteriovenous fistula. (a) Sagittal
reformatted non-contrast-enhanced head CT demonstrates
hemorrhage within the cerebellum and along the floor of
the posterior fossa extending into the foramen magnum.
(b) Sagittal reformatted maximum intensity projection
image from CTA shows multiple dilated contrast-opacified
vascular structures within the posterior fossa cerebellar
interhemispheric region just inferior to the torcular
Herophili without obvious nidus visualized. A large varix
is identified posteriorly near the midline. AP (c) and lateral
(d) projections from subsequent DSA performed from a
right external carotid injection demonstrate multiple small
arterial feeders (white arrows) arising from the right
and low-flow, dural CCFs (types B–D). Direct
CCFs are high-flow lesions representing direct
communication between the cavernous ICA and
the cavernous sinus, most commonly arising in
the setting of prior trauma or rupture of an ICA
aneurysm [32]. Rarely, an underlying connective
tissue disease such as Ehlers–Danlos or
fibromuscular dysplasia may be the cause. Patients
with direct CCFs typically present acutely,
with rapidly developing symptoms. Dural CCFs
represent abnormal communication between the
cavernous sinus and multiple branches of the ICA
(type B), ECA (type C), or both (type D).
The clinical presentation of dural CCFs is charac-
terized by the insidious onset of symptoms. Dural
CCFs often occur in the setting of chronic hyper-
tension. Dural CCFs may represent sequelae of
cavernous sinus thrombosis and subsequent
revascularization.
S. Quinet and P. Turski
occipital artery with direct communication to the trans-
verse sinus. Note is made of early venous drainage into
cortical veins, some of which are ectatic (black arrows).
AP (e) and sagittal (f) projections from subsequent DSA
performed from a left external carotid injection demon-
strate multiple small arterial feeders (open white arrows)
arising from the left occipital artery with direct communi-
cation to the transverse sinus. Note is made of early venous
drainage into cortical veins, some of which are ectatic
(open black arrows). Additional AP (g) and lateral (h)
DSA images from selective right occipital artery injection
and AP (i) and lateral (j) DSA images from selective left
occipital artery branch injection confirm the above findings
While digital subtraction angiography remains
the gold standard for diagnosis of CCF, a number
of findings on CTA/MRA can suggest this diag-
nosis, including asymmetric enlargement of the
cavernous sinus ipsilateral to the CCF, enlarge-
ment of dural venous sinuses due to increased
venous pressure, and early enhancement of the
cavernous sinus in the arterial phase. Addition-
ally, orbital abnormalities on the basis of venous
engorgement and increased venous outflow from
the cavernous sinus are also characteristic. Typical
findings include enlargement of the superior oph-
thalmic vein, proptosis, enlargement of the
extraocular muscles, and edema within the orbital
fat. Subarachnoid and/or intraparenchymal hem-
orrhage may occur due to rupture of a cortical vein
(Fig. 29).
Cavernous malformations (cavernomas) are
low-flow vascular malformations. Cavernomas
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage 267

Table 8 Cognard classification of dural arteriovenous dark rim. This hypointense rim is especially nota-
fistulas (dAVFs) ble on GRE and SWI sequences, particularly at
Other Risk of higher field strengths, with small lesions occa-
Type Drainage characteristics hemorrhage sionally detectable only on these sequences [34].
I Dural venous Anterograde Low Cavernomas lack both a nidus and enlarged sup-
sinus flow in sinus
plying arteries, distinguishing these lesions from
IIa Dural venous Retrograde Low
sinus flow in sinus pial AVMs.
IIb Dural venous Anterograde Moderate Cavernomas can present as single or multiple
sinus flow in sinus + lesions and can occur sporadically or as an auto-
cortical vein somal dominant condition with variable penetra-
reflux
tion (Fig. 30). Clinically, these patients can
IIa + b Dural venous Retrograde Moderate
sinus flow in sinus +
present with headaches and/or seizures with spon-
cortical vein taneous hemorrhage being a less common presen-
reflux tation. However, cavernomas are often detected
III Cortical vein Cortical vein High incidentally. It is estimated that 10–33 % of
drainage only cavernomas are associated with developmental
IV Cortical vein Cortical vein High
venous anomalies (DVAs) at the time of diagnosis
drainage only +
venous ectasia (Fig. 31). This fact can be exploited using post-
V Cortical vein Associated High contrast imaging; if a DVA is identified adjacent
+ spinal with to an intraparenchymal hematoma, a ruptured
perimedullary progressive cavernoma compressed by the hematoma is
veins myelopathy
strongly suspected, given the extremely low like-
lihood that the DVA itself is responsible for the
Table 9 Borden classification of dural arteriovenous fis-
hemorrhage [35]. It is important to note that a
tulas (dAVFs) partially calcified lesion on unenhanced CT
Other Risk of
carries a differential diagnosis that includes
Type Drainage characteristics hemorrhage AVM and a calcified neoplasm, so further imaging
I Dural No cortical vein Low evaluation, ideally with contrast-enhanced MR
venous reflux and MRA, is recommended. In contradistinction
sinus to other intracranial vascular malformations,
II Dural Cortical vein Moderate cavernomas are most often angiographically
venous reflux
sinus occult, and as such, catheter angiography is not
III Cortical Cortical vein High routinely part of the work-up of these lesions.
vein drainage only DVAs are common and are seen routinely as
incidental lesions on post-contrast CT and MR
imaging. It is important to recognize that DVAs
pathologically are hamartomatous lesions com- represent the sole venous drainage of the affected
posed of single-layered endothelial-lined vessels brain, and thrombosis or resection of these lesions
containing blood and, occasionally, calcification will result in infarction of the drained territory. A
or mineralization, without intervening normal “caput medusa” appearance is common, with mul-
brain tissue [33]. As such, cavernomas may tiple small tributaries converging upon a single
appear as areas of indistinct or amorphous subtle large draining vein, extending to the ependymal
hyperdensity on CT. Cavernomas can also dem- surface of the lateral ventricle or to a cortical vein.
onstrate subtle hypodensity on unenhanced DVAs rarely hemorrhage unless associated with a
CT. On MR, these lesions are classically cavernoma or thrombosis of the prominent anom-
described as heterogeneous in signal intensity, alous draining vein, leading to venous infarct.
with areas of “popcorn-like” bright signal cen- Advanced perfusion imaging techniques, either
trally on both T1- and T2-WI sequences with a CTP or MRP, may demonstrate prolonged mean
268 S. Quinet and P. Turski

transit times and/or increased relative cerebral
blood volume, which may indicate venous con-
gestion within these engorged anomalous venous
structures. Occasionally, non-enhancing
T2/FLAIR signal hyperintensity may be identified
in the subcortical white matter surrounding an
unusually large DVA, which may represent
edema and/or gliosis related to the venous con-
gestion and relative ischemia (Fig. 32). Further-
more, a recent publication has described the

Fig. 29 (continued)
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage 269

ä
Fig. 29 (continued)
270
presence of hypointense foci associated with
DVAs in almost two-thirds of patients when
imaged with susceptibility-weighted sequences
at 3-T [36]. It is thought that this hypointense
signal represents microhemorrhage and/or small
cavernomas, and the increased sensitivity of SWI
suggests that these entities are more commonly
associated with DVAs than previously thought.
Capillary telangiectasias represent dilated cap-
illaries with intervening normal brain tissue. Tel-
angiectasias are most often clinically silent and
the importance of their identification is
distinguishing them from other lesions. Typical
locations include the brainstem (especially the
pons), cerebellum, and spinal cord. On imaging,
they classically present as focal subcentimeter,
non-mass-like areas of hypointense signal on
T2* and SWI sequences with associated
ill-defined, “brushlike” enhancement (Fig. 33).
Telangiectasias are often occult on other conven-
tional MR sequences. They are most often solitary
in the sporadic form, while multiple lesions are
more commonly seen in the setting of prior radi-
ation. Radiation-induced telangiectasias will be
covered later in the chapter.
Hemorrhagic Transformation of Infarct
There are two types of hemorrhage that can occur
following reperfusion of an infarct (Fig. 34,
Table 10) [37]. The first is a petechial type of
hemorrhage, representing a relatively normal evo-
lution of infarct. It is an expected finding after
reperfusion in the area of infarct. Petechial hem-
orrhage is subdivided into type 1 (no mass effect)
Fig. 29 Acquired traumatic carotid–cavernous fistula in a
60-year-old woman with history of recurrent falls resulting
in subdural hematomas. (a) Initial unenhanced head CT
demonstrates an acute intraparenchymal hematoma cen-
tered within the right frontal lobe. (b) Axial CTA source
image demonstrates early contrast enhancement and
enlargement of the right cavernous sinus (arrow) and the
right superior ophthalmic vein. Right proptosis is also
evident. (c) A more cephalad image from the same CTA
again shows the dilated, early draining right superior oph-
thalmic vein. (d) An additional more cephalad image from
the same study shows engorgement of right frontal and
S. Quinet and P. Turski
and type 2 (minimal mass effect). On imaging,
petechial hemorrhage is punctate or confluent in
morphology with little or no associated mass
effect, with a predilection for involvement of
deep gray matter. The other type, which is more
clinically significant, is parenchymal hematoma.
Parenchymal hematoma is divided into type
1 (hematoma involving <= 30 % of the infarct
area) and type 2 (involvement of >30 % of the
infarct area). If the hematoma is relatively small,
these are usually not of major clinical conse-
quence. Type 2 parenchymal hematomas are asso-
ciated with clinical deterioration and poor
outcome. Hemorrhagic transformation can be
seen most commonly a few days after the infarct;
however, earlier transformation can be seen in the
setting of pharmacologic or mechanical
thrombolysis.
The presence of preexisting microhemorrhages
correlates with a worse prognosis after post-
thrombolysis reperfusion, presumably due to
microvasculature structures being diseased and
damaged (Fig. 35). These patients are also more
likely to have parenchymal hemorrhages. Resto-
ration of blood flow to severely ischemic tissue
has been shown to increase the likelihood of hem-
orrhagic transformation, specifically the more
clinically significant entity of parenchymal hema-
toma. Two of the most accepted and long-standing
predictors of hemorrhagic transformation on
unenhanced head CT are the extent of confluent
hypoattenuation in the cortex and subcortical
white matter (e.g., >1/3 MCA territory) and the
presence of hyperattenuating thrombus within the
MCA (“dense MCA sign”). The use of CT perfu-
sion (CTP)-derived blood–brain barrier
parietal cortical veins. The right frontal lobar hemorrhage
is again seen. Frontal (e) and lateral (f) digital subtraction
angiography images performed from a right ICA injection
in the early arterial phase demonstrate early filling of the
cavernous sinus. Lateral arterial phase images from the
same injection in the frontal (g) and lateral (h) projections
show filling of the contralateral cavernous sinus and its
drainage pathways including dilated right superior oph-
thalmic veins. The dilated right cerebral veins seen on
prior CTA are visualized. Post-coiling frontal (i) and lateral
(j) DSA projections show exclusion of the
carotid–cavernous fistula
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage 271

Fig. 30 Series of images in

a patient with familial
multiple cavernomas
(autosomal dominant). (a)
Axial unenhanced head CT
image at the level of the
brachium pontis
demonstrates a well-defined
round heterogeneous
attenuation lesion, with
both hyperdense and
hypodense components. (b)
Axial unenhanced head CT
image at a slightly more
inferior slice further
demonstrates the extent of
the lesion. (c) Sagittal
T1-WI without contrast
demonstrates
heterogeneous signal within
the lesion, with
predominantly T1
hyperintense rim and some
central T1 hyperintensity.
(d) Axial T2-WI without
contrast demonstrates the
classic “popcorn”-type
appearance of the lesion,
with mixed central T2
signal and a hypointense
rim, representing the
methemoglobin and
hemosiderin components of
the lesion, respectively. (e)
Axial T2-WI demonstrates
a smaller but similar
appearing lesion at the
cortical–subcortical
junction in the right
temporal lobe. (f) Axial
T2-WI at more cephalad
location demonstrates
another cavernoma in the
right frontal parasagittal
region

permeability measurements has been shown to be (rCBF), reduced relative cerebral blood volume
helpful in predicting hemorrhagic transformation (rCBV), and delayed transit times [38, 39]. In
in the setting of acute ischemic stroke, including particular, severe hypoperfusion, as demonstrated
findings of reduced relative cerebral blood flow by Tmax >14 s, has been shown to be predictive
272
Fig. 31 Developmental venous anomaly and cavernoma,
sporadic pattern. Unenhanced axial head CT (a) demon-
strates faint hyperattenuation in the right frontal lobe
(arrow). Contrast-enhanced T1 MR image (b) demon-
strates partial visualization of a right frontal developmental
venous anomaly (open arrow). Filtered phase image (c)
of the development of parenchymal hematoma.
CTP-derived blood–brain permeability (BBBP)
and PWI-based assessments of BBB permeability
measurements have also shown promise in
predicting hemorrhagic transformation. Using
MR, severely ischemic tissue is characterized by
S. Quinet and P. Turski
demonstrates increased signal in the right frontal lobe
corresponding to the hyperattenuation on CT. Final multi-
plied and processed SWI image (d) demonstrates
hypointensity at this location. This appearance on phase
imaging and SWI is due to the susceptibility of effects of
diamagnetic calcium within the cavernoma
areas of very low CBV, large DWI lesion volume,
a very low ADC, and Tmax >14 s. Additional
measures of permeability on MR imaging have
also been suggested as predictive of post-
thrombolysis hemorrhage, including late-phase
T2* signal loss in perfusion sequences or
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage
Fig. 32 Series of images in a patient with a right frontal
developmental venous anomaly. (a) Axial post-contrast
T1-WI demonstrates abnormal enhancing tubular struc-
tures in the right frontal lobe converging upon and draining
into an ependymal vein in the right parasagittal region. (b)
Axial SWI image at a similar location demonstrates
hypointense signal within these enhancing structures, on
the basis of deoxyhemoglobin in this slow-flow vascular
malformation. (c) Axial T2* GRE image at a similar
post-contrast FLAIR enhancement. An additional
potential sign thought to be predictive of hemor-
rhagic transformation includes early development
of T2/FLAIR signal hyperintensity within the area
of DWI in an acute infarct, indicating a more
severe ischemic injury. Furthermore, the
hyperintense acute reperfusion marker (HARM)
sign, which has been shown to represent leakage
of gadolinium-based contrast agents across a
damaged blood–brain barrier into the
273
location shows similar signal but decreased conspicuity
of the lesion, relative to its appearance on SWI. (d) Axial
T2/FLAIR image shows confluent signal hyperintensity in
the white matter in the region of the DVA, likely
representing gliosis/leukoaraiosis in the territory drained
by the DVA. MR perfusion imaging demonstrates elevated
cerebral blood volume (e) and prolonged first moment
transit times (f) to the territory of this lesion, probably on
the basis of venous congestion
cerebrospinal fluid on post-contrast T2/FLAIR
images, may be a predictor of hemorrhagic
transformation [40].
The European Cooperative Acute Stroke Study
(ECASS) III randomized control trial demon-
strated the increased risk of symptomatic intracra-
nial hemorrhage in the setting of patients with
acute ischemic stroke treated with intravenous
alteplase thrombolysis compared to placebo
[41]. In the ECASS III protocol, symptomatic
274
Fig. 33 Series of MR images in a patient with presumed
brainstem capillary telangiectasia. Axial unenhanced T1
(a), post-contrast T1 (b), T2 (c), T2* (d), and SW (e)
images demonstrate a presumed capillary telangiectasia
in the right paramedian pons (arrows). The lesion demon-
strates faint, brushlike enhancement and susceptibility-
intracranial hemorrhage was defined as “any
apparently extravascular blood in the brain or
within the cranium that was associated with clin-
ical deterioration, as defined by an increase of four
points or more in the score on the NIHSS, or that
led to death and that was identified as the predom-
inant cause of the neurologic deterioration.”
Importantly, despite the increased incidence of
symptomatic ICH in the thrombolysis group,
there was no difference in mortality. The results
from an earlier arm of the study (ECASS II) are
the basis for the well-established association
between the extent of parenchymal
hypoattenuation on initial unenhanced head CT
and risk for symptomatic ICH. Hypoattenuation
involving more than one-third of the MCA terri-
tory was associated with increased risk of
S. Quinet and P. Turski
related signal loss on T2*, with increased conspicuity on
SWI. The lesion is not apparent on T1 and T2 FSE
sequences. The lesion did not produce symptoms referable
to its location and was stable on serial examinations
performed for other reasons
symptomatic ICH. Though less intensively stud-
ied, CTP-derived permeability measurements
have shown promise in predicting development
of malignant edema following thrombolysis.
Malignant edema was defined in the ECASS III
trial as brain edema with mass effect as the pre-
dominant cause of clinical deterioration (NIHSS
score increase >4 or death).
Venous Thrombosis
Thrombosis of a cortical vein and/or dural sinus
with subsequent hemorrhage is an important con-
sideration in evaluating for potential etiologies of
spontaneous intracranial hemorrhages in unusual
locations (e.g., not within classic arterial
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage
Fig. 34 Examples of hemorrhagic transformation. (a)
Axial unenhanced head CT demonstrates hyperdense pete-
chial hemorrhage centered in the right putamen with sur-
rounding edema. Given the minimal mass effect, this
would be classified as a type 2 petechial hemorrhage. (b)
Axial unenhanced head CT in the same patient demon-
strates and additionally shows hemorrhage within the right
caudate. (c) Axial unenhanced head CT in a different
distributions) or seemingly without an associated
underlying mass or vascular anomaly. Particular
attention should be paid to this entity in the pedi-
atric, young adult female, and hypercoagulable
275
patient demonstrates a hyperdense hemorrhage centered
in the right caudate and the periventricular white matter.
Note hypodensity throughout the cortex and subcortical
white matter in the distribution of the right MCA. The
extent of hematoma is <30 % and can be classified as a
type 2 hematoma. (d) Axial unenhanced head CT image at
a slightly more cephalad location as C shows to better
advantage the intraventricular extension of hemorrhage
patient populations. The clinical presentation of
patients with venous thrombosis is often not
straightforward, with severe, persistent headache
a common reason for referral to imaging.
276
Table 10 Classification of hemorrhagic transformation
(HT) in ischemic strokea
Type I Type II
Petechial No mass effect Minimal mass
hemorrhage effect
Parenchymal <30 % infarct >30 % infarct
hematoma territory territory
a
Based on European Cooperative Acute Stroke Study
(ECASS) I and II definitions
In the pediatric setting, patients may present with
nausea, vomiting, and irritability. If there is exten-
sion of thrombus to involve the cortical veins with
subsequent venous infarction, the patient may
present with seizures. The sequence of events
leading intracranial hemorrhage is thought to be
due to prolonged venous stasis leading to throm-
bosis with subsequent venous hypertension,
ischemia, and infarction.
Given the vague presentation, unenhanced
CT may be the first imaging obtained, depending
on the level of suspicion for this entity.
Hypoattenuation and/or hemorrhage in non-arterial
distributions can be a clue to venous infarction.
Assessment of venous vascular territories in addi-
tion to those of arterial vascular territories is impor-
tant, as certain patterns of signal abnormality
and/or hemorrhage can signal the presence of
venous thrombosis and suggest the site of throm-
bus. Examples include the posterior temporal lobe
(vein of Labbe +/ transverse sinus), parasagittal
frontoparietal lobes (superior sagittal sinus and/or
cortical veins), and bilateral thalami (internal cere-
bral veins) (Figs. 36, 37, 38, and 39).
If infarction has not yet occurred, hyperdense
thrombus within venous sinuses or cortical veins
(“cord sign”) may be the only finding visible early
in the process, though the clot tends to become
less dense after a week and may thus be incon-
spicuous on non-contrast imaging (Fig. 36).
Contrast-enhanced CT, CTA, and optimally CTV
classically demonstrate normally enhancing dura
surrounding non-enhancing thrombus with a dural
sinus (the “empty delta” sign) (Fig. 37).
On MR imaging, a corollary to the cord sign is
presence of T1 hyperintense methemoglobin
within the thrombus. This can be particularly con-
spicuous within the superior sagittal sinus on
S. Quinet and P. Turski
unenhanced sagittal T1-WI (Figs. 37 and 38).
The presence of expansion of the sinus or vein
can be a helpful ancillary finding supporting the
presence of thrombosis, particularly if the throm-
bus is not obviously T1 hyperintense. Likewise,
routine unenhanced spin echo images should be
scrutinized for preserved flow voids in venous
sinuses and cortical veins. However, because
lack of a flow void may be due to either thrombo-
sis or slow flow, additional sequences can be
helpful to distinguish these two entities (e.g.,
post-contrast imaging, TOF, posterior cerebral
artery (PCA), contrast-enhanced time-resolved
4D MRA – 4DCEMRA). TOF or PCA imaging
without and/or with contrast can be performed if
there is high clinical suspicion prospectively
(Figs. 37 and 38). TOF MRV images demonstrat-
ing lack of flow-related enhancement can be sug-
gestive of the diagnosis; however, correlation
with other sequences is mandatory, as inflow
dephasing may mimic thrombosis and T1
hyperintense thrombus may mimic flow-related
enhancement. Unenhanced TOF and PCA tech-
niques are particularly valuable in the patient pop-
ulation unable to receive intravenous contrast
agents. If post-contrast imaging or 4D CE MRA
is used, filling defects within the thrombosed
venous structures can be identified, analogous to
post-contrast CT imaging (Fig. 39) [42]. Post-
contrast imaging can also be helpful in
distinguishing thrombosis versus a hypoplastic
sinus.
Venous infarcts less often demonstrate large
areas of restricted diffusion compared to arterial
infarcts, and the absence of restricted diffusion
does not exclude venous ischemia or infarction
(Fig. 36). As discussed above, DWI and ADC
maps should be interpreted with caution given
the possibility of T2 shine-through and dark-
through effects complicating interpretation.
Hemorrhagic Neoplasm
Hemorrhage into underlying malignant intracra-
nial neoplasms, either primary or metastatic,
accounts for approximately 6–10 % of spontane-
ous intracranial hematomas. Among primary CNS
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage
Fig. 35 Risk factors for hemorrhagic transformation. (a)
Axial T2/FLAIR image demonstrates abnormal
hyperintense signal within the cortex and subcortical
white matter within the infarcted territory. Note is made
of hypointense microhemorrhage in the left frontal lobe
(arrow). (b) Axial T2* GRE image at a similar level
demonstrates the left frontal hemorrhage to better
neoplasms, grade IV astrocytoma/glioblastoma
multiforme is most likely to present as hemor-
rhagic masses (Figs. 40, 41, and 42). Additional
malignant primary CNS neoplasms with a propen-
sity for hemorrhage include oligodendrogliomas
277
advantage (arrow). (c) Perfusion imaging demonstrates
prolonged mean transit times through the majority of the
right MCA distribution. (d) Very low relative cerebral
blood volume is also noted corresponding to the distribu-
tion of the prolonged transit times. The presence of
microhemorrhage increases the risk of hemorrhagic con-
version after thrombolysis
and ependymomas. Metastatic lesions related to
melanoma (Fig. 43), choriocarcinoma, renal cell
carcinoma, and thyroid carcinoma are most likely
to hemorrhage. Of note, despite the increased pro-
pensity of these tumor types to hemorrhage, a
278
Fig. 36 Series of CT and MR images in a 20-year-old
female patient with Crohn disease presenting with bilateral
internal cerebral vein thrombosis. (a) Axial unenhanced
head CT image demonstrates confluent hypodensity within
bilateral thalami, left greater than right. Note hyperdense
thrombus in the expected location of the internal cerebral
veins (arrow). (b) Axial unenhanced head CT image at a
slightly more inferior location shows the same findings. (c)
Axial unenhanced T2-WI at a similar level as the previous
image demonstrates predominantly hyperintense signal in
the thalami corresponding to hypodensity on the CT
images. Scattered foci of hypointensity are also present
within these lesions. (d) Axial T2 FLAIR image at a similar
location confirms the findings seen on T2-WI, with a
S. Quinet and P. Turski
hypointense area in the left thalamus becoming more con-
spicuous on this sequence. (e) Axial T2* GRE image
demonstrates “blooming” of hypointense areas within the
left thalamus with a hyperintense core and petechial hem-
orrhage in the right thalamus. (f) Axial DWI demonstrates
hyperintensity within the left thalamus. (g) Corresponding
axial ADC map image demonstrates hypointensity in the
region of hyperintense signal on DWI, confirming apparent
diffusion restriction. (h) Sagittal TOF MIP image demon-
strates lack of inflow enhancement of the straight sinus and
bilateral internal cerebral veins. (i) Sagittal MIP flow image
from the PC MRA examination demonstrates excellent
subtraction of stationary tissue and absence of flow signal
in the straight sinus and internal cerebral veins
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage 279

Fig. 37 (continued)
280
Fig. 37 Series of CT and MR images in a 51-year-old
patient presenting with seizures and fever in the setting of
illicit drug use. (a) Axial image from unenhanced head CT
demonstrates subarachnoid hemorrhage layering along the
sulci of the right frontal lobe. Hypodensity is also noted
within the cortex and subcortical white matter of the right
frontal lobe with blurring of gray–white distinction. (b)
Enhanced head CT at a slightly more cephalad location
again shows cortical enhancement. (c) Axial image from
subsequent CTA in the same patient demonstrates filling
defect within the superior sagittal sinus (arrow) indicating
thrombosis at this site. (d) Axial CTA image at a higher
location again shows superior sagittal sinus thrombosis
(arrow). (e) Sagittal reformatted CTA again shows the
sagittal sinus thrombosis (arrow). (f) Sagittal unenhanced
T1-WI image demonstrates hyperintense thrombus within
the superior sagittal sinus (arrow). (g) Coronal T2 FLAIR
image demonstrates hyperintense signal within the right
hemorrhagic metastatic lesion discovered on imag-
ing is most likely to be secondary to a lung or breast
primary malignancy, given the much higher inci-
dence and prevalence of these tumors.
S. Quinet and P. Turski
frontal cortex and subcortical white matter. There is corti-
cal enlargement and loss of gray–white distinction. Note
linear hypointense signal at the lateral margin of the right
frontal lobe (arrow) representing a thrombosed cortical
vein. (h) Axial T2* GRE image demonstrates blooming
of hypointense signal within the thrombosed cortical vein
(arrow). (i) Axial T2* GRE image at a slightly higher level
shows the further extent of the thrombosed cortical vein
(arrows). (j) Axial T2/FLAIR image shows abnormal
hyperintense signal within the cortex and subcortical
white matter of the right frontal lobe. Abnormal
hyperintense signal also interdigitates between the sulci
of the right frontal lobe within the subarachnoid space.
(k) Axial post-contrast T1-WI shows abnormal
leptomeningeal enhancement. (l) Axial T2-WI demon-
strates linear hypodense thrombosis within a cortical vein
(arrow). (m) Axial DWI demonstrates hyperintensity
within the right frontal cortex
Potential mechanisms responsible for
intratumoral hemorrhage include neovascularity/
angiogenesis, rapid tumor growth and necrosis,
and vascular invasion. Several factors have been
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage
Fig. 38 Series of MR images in a patient with right frontal
intraparenchymal hemorrhage related to superior sagittal
sinus thrombosis. (a) Sagittal unenhanced T1-WI near the
midline demonstrates hyperintense methemoglobin throm-
bus within the superior sagittal sinus. (b) Sagittal T1-WI
right parasagittal near the midline demonstrates slightly
hyperintense hemorrhage in the right frontal lobe. (c) Sag-
ittal PC MRV demonstrates lack of flow signal within the
majority of the superior sagittal sinus (arrows) with incom-
plete thrombosis of the posterior aspect. (d) Axial T2-WI
suggested which help to favor a malignant etiol-
ogy for intracranial hemorrhage rather than a
benign cause, including locations atypical for
other common benign etiologies of intracranial
hemorrhage (e.g., hypertensive hemorrhage),
multiplicity, and early lesion enhancement. Addi-
tionally, delay and/or distortion of the expected
orderly evolution of hematoma evolution on
MR can be suggestive of underlying malignant
neoplasm, potentially on the basis of intratumoral
hypoxia [43]. Prominent perilesional edema
which persists on follow-up imaging and lack
of development of a complete hemosiderin rim
281
demonstrates the hematoma to be predominantly
hyperintense with internal areas of hypointensity, indicat-
ing different stages of hematoma evolution. Correlating
with the T1 signal, the hematoma is likely in the acute
stage. (e) Axial T2* GRE image demonstrates blooming of
the hypointense components of the hematoma due to sus-
ceptibility effects. (f) Axial DWI image demonstrates
hyperintense signal at the periphery and hypointense signal
in the center of the hematoma
favor hemorrhagic malignant neoplasm over a
benign entity.
Metastatic melanoma can present a diagnostic
dilemma. Melanin can cause intrinsic T1
hyperintensity, as does methemoglobin. Meta-
static melanoma can present as multiple areas
of hemorrhage within the metastasis and/or
as recurrent hemorrhage from metastatic mela-
noma leading to superficial siderosis (hemosid-
erin staining of the subarachnoid space and
pial surface of the brain). As such, this entity
may mimic severe hypertensive arteriopathy
and CAA.
282
Fig. 39 Transverse sinus and vein of Labbe thrombosis. (a)
Axial T2-WI demonstrates hypointense lesion with sur-
rounding hyperintense signal involving the cortex and sub-
cortical white matter of the posterior left temporal lobe. (b)
Axial post-contrast T1-WI demonstrates heterogeneous but
predominantly hypointense signal within the affected left
temporal lobe associated with abnormal leptomeningeal
enhancement overlying the involved cortex. (c) Axial T2*
GRE image shows “blooming” of hypointense blood prod-
ucts within the hematoma. Note the linear hypointense
signal and blooming at the lateral margin of the left temporal
lobe representing thrombosis within the vein of Labbe
S. Quinet and P. Turski
(arrow). (d) Axial T2* GRE image at a slightly more
cephalad location shows similar findings as C. Vein of
Labbe thrombosis is again seen (arrow). (e) Axial post-
contrast T1-WI demonstrates a filling defect within the left
transverse sinus (arrow). (f) Axial T2* GRE image demon-
strates blooming of thrombus within the left transverse sinus
(arrow). (g) Coronal MRA image shows lack of enhance-
ment of the left transverse sinus. (h) Coronal MRA image a
slightly more posterior location demonstrates enhancement
of the dura of the left transverse sinus surrounding a
non-enhancing filling defect representing the thrombus
(arrow), the so-called empty delta sign
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage
Fig. 40 Grade IVastrocytoma. Axial post-contrast T1 (a),
T2* (b), and SW (c) images demonstrate a
microhemorrhage in the subcortical right frontal lobe
(white arrow) and macrohemorrhage (black arrow) in the
Radiation-Induced Vascular Disease
and Hemorrhage
Telangiectasias and cavernomas have been
described with increased frequency in patients
with history of prior CNS irradiation, particularly
in childhood, relative to the general population.
Radiation-induced telangiectasias have been
shown to occur in at least 20 % of children receiv-
ing childhood radiation, with a suggestion of
apparent increased frequency of telangiectasias
occurring in patients receiving higher radiation
doses [44]. The increased frequency in the setting
of childhood CNS radiation is thought to reflect
the established increased radiation sensitivity of
the pediatric brain, with possibly a contribution
from increased expression of angiogenic factors
(e.g., vascular endothelial growth factor) in the
pediatric population. Pathologically, telangiecta-
sias represent dilated capillaries with thin walls
and intervening normal brain parenchyma.
Attempted collateralization in the setting of
small venous thrombosis has been suggested as a
potential etiology of radiation-induced telangiec-
tasia development. Both telangiectasias and
cavernomas can be categorized as a subtype of
283
right temporal lobe. Extensive enhancing leptomeningeal
metastatic disease is present within the Sylvian fissure and
studding the pial surfaces of the right frontal and temporal
lobes
delayed radiation injury and are thought to repre-
sent a proliferative vasculopathy involving small
vessels, arising on the basis of radiation injury to
the microvasculature [45]. These lesions are
thought to arise along a different pathophysio-
logic pathway than large-vessel radiation-induced
vasculopathy, which is characterized by acceler-
ated atherosclerosis on the basis of hyalinization
and fibrinoid necrosis of the vessel wall, intimal
proliferation, and thrombosis. Telangiectasias
have a much shorter latency period, on average
arising under a year after radiation. Conversely,
cavernomas may arise as late as 26 years follow-
ing radiation. It has been speculated that given this
discrepancy in latency periods, cavernomas may
progress from telangiectasias. A shorter latency
period for development of cavernomas has been
described with childhood radiation doses in
excess of 30 Gy.
On MR imaging, radiation-induced telangiec-
tasias most often appear as homogeneous small
(often <5–10 mm) areas of T2 hypointensity with
“blooming” on T2* GRE and SWI sequences
(Fig. 44). Occasionally, telangiectasias can be het-
erogeneous in signal with at T2 hypointense rim.
These lesions may demonstrate stippled or punc-
tate enhancement. The appearance of multiple
284
Fig. 41 Grade IV astrocytoma. Initial unenhanced head
CT (a) demonstrates a large acute hemorrhage involving
the right cerebral hemisphere. Post-contrast T1 image (b)
shows heterogeneous enhancement of an underlying mass.
(c) T2 image demonstrates blood products in varying states
of degradation within the hematoma associated with the
S. Quinet and P. Turski
soft tissue mass. (d) SWI MinIP image demonstrates
susceptibility-related signal loss throughout the hemor-
rhagic mass. Vasogenic edema surrounds the mass. Diffu-
sion tensor imaging/tractography (e) and functional MRI
(f) were subsequently performed for pretreatment planning
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage
Fig. 42 Grade IV astrocytoma. Unenhanced head CT (a)
demonstrates a large intraparenchymal hemorrhage cen-
tered within the right frontal lobe and surrounding
vasogenic edema causing marked mass effect. Follow-up
MR imaging with T2 (b), T2* (c), perfusion (d), and
coronal reformatted post-contrast T1 sequences show the
underlying mass in better detail. The mass is shown to be
radiation-induced telangiectasias can mimic
CAA. The lesions are most often asymptomatic,
unless associated with hemorrhage, which usually
occur in the setting of a cavernoma.
The appearance of radiation-induced
cavernomas is no different than those arising on
a congenital or genetic basis and is described
earlier in this chapter [46]. The appearance of a
cavernoma can often be distinguished from neo-
plastic entities based on their characteristic
285
complex, with hemorrhagic and necrotic components with
enhancement of the solid components. Contralateral exten-
sion via the corpus callosum is best appreciated on the
coronal image. Increased cerebral blood volume within
the solid components of the mass is identified on perfusion
imaging
imaging appearance as well as a typical relative
lack of surrounding edema.
Vasculitis
Central nervous system vasculitis is a rare
cause of spontaneous intracranial hemorrhage,
which can be intraparenchymal or subarachnoid
in location [47]. In the absence of history or
286
Fig. 43 Metastatic melanoma. Axial T1 images (a–b),
post-contrast T1 (c–d), and axial T2* (e–f) MR images
demonstrate multifocal subcortical metastatic lesions in a
patient with melanoma. Several of the lesions demonstrate
T1 hyperintense signal which may represent melanin
strong clinical suspicion for the diagnosis,
vasculitis can be an extremely difficult diagnosis
to make prospectively in a patient presenting
with intracranial hemorrhage. Vasculitis is
defined as inflammation of the vessel wall.
Inflammatory changes can lead to irregularity
and compromise of the vessel lumen, which
may ultimately manifest as ischemia, infarct,
and/or hemorrhage.
CNS vasculitis is considered primary if there is
no underlying cause identified (i.e., primary
angiitis of the CNS). Secondary CNS vasculitis
is more common and due to systemic vasculitides
S. Quinet and P. Turski
and/or methemoglobin. T2* images demonstrate blooming
of hypointense signal confirming the presence of methe-
moglobin. The presence of enhancement confirms the met-
astatic nature of these microhemorrhages
(e.g., granulomatosis with polyangiitis), connec-
tive tissue diseases (e.g., systemic lupus
erythematosus), angioinvasive infection, and illicit
drugs (most commonly cocaine, amphetamines,
and other sympathomimetics). On imaging, CTA
or MRA may show nonspecific irregular beading
of the arterial structures (Figs. 45 and 46). Impor-
tantly, small vessel disease is below current imag-
ing resolution of routine MRI or MRA imaging,
and a negative examination does not exclude the
diagnosis. DSA remains the most sensitive imag-
ing modality to suggest the diagnosis, but is simi-
larly nonspecific. DSA classically demonstrates
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage
Fig. 44 Radiation-induced microangiopathy and telangi-
ectasias and/or cavernomas. Axial unenhanced head CT at
the time of presentation (a) shows a large partially calcified
mass centered in the left hemisphere with associated marked
mass effect, resected shortly thereafter and shown to be
anaplastic ependymoma. (b–f) Series of MRI images
obtained approximately 2 years after resection and focused
radiation in the setting of new right-hand weakness and
dysarthria. Axial T2/FLAIR image (b) demonstrates
hyperintense signal in the subcortical white matter of the
left frontal lobe and periventricular region, compatible with
postradiation change. Axial T2 (c) and SWI (d) images
demonstrate multiple foci of hypointensity and
287
susceptibility-related signal loss representing multiple pre-
sumed capillary telangiectasias and/or cavernomas within
the radiation treatment bed. Axial high B-value DWI (e) and
ADC map (f) images demonstrate small foci of restricted
diffusion within the cortex of the left frontal lobe.
Reconstructed maximum intensity projection image from
three-dimensional phase-contrast MRA (g) demonstrates
attenuation of flow signal throughout the M1 segment and
distal branches of the left MCA. Digital subtraction angiog-
raphy images obtained shortly after this MR in coronal
projection of right (h) and left (i) common carotid injections
confirm the findings on the PCA images. The right intracra-
nial vasculature is normal and shown for comparison
288
Fig. 45 Series of MR images in a patient with CNS
vasculitis. (a) Coronal T2/FLAIR image demonstrates
hyperintense signal along the sulci of the right frontal
lobe (white solid arrow) consistent with subarachnoid
hemorrhage. (b) DWI demonstrates an area of diffusion
restriction within the right parietal lobe (white open
arrow), representing infarction. (c) Perfusion-weighted
multifocal areas of small- and middle-sized arterial
stenosis and dilations. Severe, diffuse atheroscle-
rotic vascular disease can mimic vasculitis on CTA,
MRA, and DSA imaging. Leptomeningeal biopsy
may ultimately be required to arrive at the diagno-
sis in the absence of an established diagnosis of a
systemic disorder or suggestive clinical history.
S. Quinet and P. Turski
imaging demonstrates prolonged transit times to the area
of infarct (black solid arrow). Series of MRA images in a
patient with CNS vasculitis. (d–f) Coronal MIP images
from TOF MRA demonstrate diffuse thinning and irregu-
larity of the proximal arterial vasculature, most notably
within the M1 segments of the MCS
Other less common vascular disorders that
uncommonly present with intracranial hemor-
rhage include moyamoya disease and syndrome,
acute hypertensive encephalopathy/posterior
reversible encephalopathy syndrome (PRES),
and reversible cerebral vasoconstriction syn-
drome (RCVS).
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage
Fig. 46 Series of CT and MR images in a patient with
vasculitis secondary to systemic lupus erythematosus
(SLE). (a) Axial unenhanced CT image demonstrates
right frontal subarachnoid hemorrhage (arrow). (b) Axial
unenhanced CT image at a more inferior location demon-
strates right thalamic hematoma and intraventricular exten-
sion of hemorrhage (open arrow) and a left parietal infarct
(arrow). (c) Axial T2* GRE image demonstrates blooming
of hypointense right thalamic hemorrhage. (d) T2-WI at a
similar location demonstrates mixed intensity within the
289
right thalamic hematoma on the basis of blood products of
differing ages. (e) Axial post-contrast T1-WI demonstrates
peripheral enhancement of the right thalamic hematoma.
Care must be taken to distinguish this lesion from a mass
lesion or abscess. (f–h) Series of CTA images from the
same patient demonstrates diffuse irregularity of the intra-
cerebral arterial vasculature. As in the previous case, find-
ings are particularly evident in the M1 segments of the
bilateral MCA
290
Summary
Unenhanced CT and dynamic CTA and post-
contrast CT are the primary tools for evaluating
intracranial hemorrhage. It is important to evaluate
the hemorrhage based on location (e.g., cortical,
basal ganglia, SAH, IVH). Additionally, evidence
of active bleeding (e.g., spot sign), an underlying
lesion (e.g., vascular malformation, neoplasm), the
size of the hematoma, and associated mass effect
(e.g., subfalcine or transtentorial herniation, acute
hydrocephalus, infarct) require close attention. MR
imaging is well suited to age the hemorrhage,
though evaluation can be complicated when asso-
ciated with an underlying lesion and/or recurrent
hemorrhage occurs. Advanced vascular imaging
techniques, with both CT and MRI, are helpful in
prognostication and risk stratification of subse-
quent hemorrhage in a subset of entities causing
nontraumatic intracerebral hemorrhage.
References
1. Khosravani H, Mayer SA, Demchuk A, Jahromi BS,
Gladstone DJ, Flaherty M, Aviv RI (2013) Emergency
noninvasive angiography for acute intracerebral hem-
orrhage. Am J Neuroradiol 34(8):1481–1487.
doi:10.3174/ajnr.A3296
2. Broderick JP, Adams HP, Barsan W, Feinberg W,
Feldmann E, Grotta J, Zuccarello M (1999) Guidelines
for the management of spontaneous intracerebral hem-
orrhage: a statement for healthcare professionals from a
special writing group of the Stroke Council, American
Heart Association. Stroke 30(4):905–915.
doi:10.1161/01.STR.30.4.905
3. Fischbein NJ, Wijman CAC (2010) Nontraumatic
intracranial hemorrhage. Neuroimaging Clin N Am
20(4):469–492. doi:10.1016/j.nic.2010.07.003
4. Wada R, Aviv RI, Fox AJ, Sahlas DJ, Gladstone DJ,
Tomlinson G, Symons SP (2007) CT angiography
“spot sign” predicts hematoma expansion in acute
intracerebral hemorrhage. Stroke 38(4):1257–1262.
doi:10.1161/01.STR.0000259633.59404.f3
5. Rønning P, Sorteberg W, Nakstad P, Russell D, Helseth
E (2008) Aspects of intracerebral hematomas – an
update. Acta Neurol Scand 118(6):347–361.
doi:10.1111/j.1600-0404.2008.01023.x
6. Huisman TAGM (2005) Intracranial hemorrhage:
ultrasound, CT and MRI findings. Eur Radiol 15
(3):434–440. doi:10.1007/s00330-004-2615-7
S. Quinet and P. Turski
7. Gomori JM, Grossman RI (1988) Mechanisms respon-
sible for the MR appearance and evolution of intracra-
nial hemorrhage. Radiographics 8(3):427–440.
doi:10.1148/radiographics.8.3.3380989
8. Barkovich AJ, Atlas SW (1988) Magnetic resonance
imaging of intracranial hemorrhage. Radiol Clin North
Am 26(4):801–820
9. Silvera S, Oppenheim C, Touzé E, Ducreux D, Page P,
Domigo V, Meder J-F (2005) Spontaneous intracere-
bral hematoma on diffusion-weighted images: influ-
ence of T2-shine-through and T2-blackout effects.
Am J Neuroradiol 26(2):236–241
10. Viswanathan A, Greenberg SM (2011) Cerebral amy-
loid angiopathy in the elderly. Ann Neurol 70
(6):871–880. doi:10.1002/ana.22516
11. Chao CP, Kotsenas AL, Broderick DF (2006) Cerebral
amyloid angiopathy: CT and MR imaging findings.
Radiographics 26(5):1517–1531. doi:10.1148/
rg.265055090
12. Haacke EM, DelProposto ZS, Chaturvedi S, Sehgal V,
Tenzer M, Neelavalli J, Kido D (2007) Imaging cere-
bral amyloid angiopathy with susceptibility-weighted
imaging. Am J Neuroradiol 28(2):316–317
13. Wang J, Gong X (2011) Superficial siderosis of
the central nervous system: MR findings with
susceptibility-weighted imaging. Clin Imaging 35
(3):217–221. doi:10.1016/j.clinimag.2010.06.003
14. Schrag M, McAuley G, Pomakian J, Jiffry A, Tung S,
Mueller C, Kirsch WM (2010) Correlation of
hypointensities in susceptibility-weighted images to
tissue histology in dementia patients with cerebral
amyloid angiopathy: a postmortem MRI study. Acta
Neuropathol 119(3):291–302. doi:10.1007/s00401-
009-0615-z
15. Chavhan GB, Babyn PS, Thomas B, Shroff MM,
Haacke EM (2009) Principles, techniques, and appli-
cations of T2*-based MR imaging and its special appli-
cations. Radiographics 29(5):1433–1449. doi:10.1148/
rg.295095034
16. Goos JDC, van der Flier WM, Knol DL, Pouwels PJW,
Scheltens P, Barkhof F, Wattjes MP (2011) Clinical
relevance of improved microbleed detection by
susceptibility-weighted magnetic resonance imaging.
Stroke 42(7):1894–1900. doi:10.1161/
STROKEAHA.110.599837
17. Haacke EM, Mittal S, Wu Z, Neelavalli J, Cheng Y-CN
(2008) Susceptibility-weighted imaging: technical
aspects and clinical applications, part 1. Am J
Neuroradiol 30(1):19–30. doi:10.3174/ajnr.A1400
18. Mittal S, Wu Z, Neelavalli J, Haacke EM (2008)
Susceptibility-weighted imaging: technical aspects
and clinical applications, part 2. Am J Neuroradiol 30
(2):232–252. doi:10.3174/ajnr.A1461
19. Barnes SRS, Haacke EM (2009) Susceptibility-
weighted imaging: clinical angiographic applications.
Magn Reson Imaging Clin N Am 17(1):47–61.
doi:10.1016/j.mric.2008.12.002
13 Imaging of Spontaneous Nontraumatic Intracerebral Hemorrhage
20. Schweser F, Deistung A, Lehr BW, Reichenbach JR
(2010) Differentiation between diamagnetic and para-
magnetic cerebral lesions based on magnetic suscepti-
bility mapping. Med Phys 37(10):5165. doi:10.1118/
1.3481505
21. Noguchi K, Seto H, Kamisaki Y, Tomizawa G,
Toyoshima S, Watanabe N (2000) Comparison of
fluid-attenuated inversion-recovery MR imaging with
CT in a simulated model of acute subarachnoid hem-
orrhage. Am J Neuroradiol 21(5):923–927
22. Papke K, Kuhl CK, Fruth M, Haupt C, Schlunz-
Hendann M, Sauner D, Brassel F (2007) Intracranial
aneurysms: role of multidetector CT angiography in
diagnosis and endovascular therapy planning.
Radiology 244(2):532–540. doi:10.1148/radiol.
2442060394
23. McKinney AM, Palmer CS, Truwit CL, Karagulle A,
Teksam M (2008) Detection of aneurysms by
64-section multidetector CT angiography in patients
acutely suspected of having an intracranial aneurysm
and comparison with digital subtraction and 3D rota-
tional angiography. Am J Neuroradiol 29(3):594–602.
doi:10.3174/ajnr.A0848
24. Yoon DY, Lim KJ, Choi CS, Cho BM, Oh SM, Chang
SK (2007) Detection and characterization of intracra-
nial aneurysms with 16-channel multidetector row CT
angiography: a prospective comparison of volume-
rendered images and digital subtraction angiography.
Am J Neuroradiol 28(1):60–67
25. Villablanca JP, Duckwiler GR, Jahan R, Tateshima S,
Martin NA, Frazee J, Vinuela FV (2013) Natural history
of asymptomatic unruptured cerebral aneurysms evalu-
ated at CT angiography: growth and rupture incidence
and correlation with epidemiologic risk factors. Radiol-
ogy 269(1):258–265. doi:10.1148/radiol.13121188
26. Wermer MJH, van der Schaaf IC, Algra A, Rinkel GJE
(2007) Risk of rupture of unruptured intracranial
aneurysms in relation to patient and aneurysm charac-
teristics: an updated meta-analysis. Stroke 38
(4):1404–1410. doi:10.1161/01.STR.0000260955.
51401.cd
27. Wiebers DO, Whisnant JP, Huston J 3rd, Meissner I,
Brown RD Jr, Piepgras DG, International Study of
Unruptured Intracranial Aneurysms Investigators
(2003) Unruptured intracranial aneurysms:
natural history, clinical outcome, and risks of surgical
and endovascular treatment. Lancet 362
(9378):103–110
28. Wiebers DO, Piepgras DG, Meyer FB, Kallmes DF,
Meissner I, Atkinson JL, Brown RD Jr (2004) Patho-
genesis, natural history, and treatment of unruptured
intracranial aneurysms. In: Mayo clinic proceedings,
vol 79. Elsevier, pp 1572–1583. Retrieved from http://
www.sciencedirect.com/science/article/pii/S00256196
11618685
29. Etminan N, Beseoglu K, Barrow DL, Bederson J,
Brown RD, Connolly ES, Macdonald RL (2014)
291
Multidisciplinary consensus on assessment of
unruptured intracranial aneurysms: proposal of an inter-
national research group. Stroke 45(5):1523–1530.
doi:10.1161/STROKEAHA.114.004519
30. Mullan S, Mojtahedi S, Johnson DL, Macdonald RL
(1996) Embryological basis of some aspects of cerebral
vascular fistulas and malformations. J Neurosurg
85(1):1–8
31. Gandhi D, Chen J, Pearl M, Huang J, Gemmete JJ,
Kathuria S (2012) Intracranial dural arteriovenous fis-
tulas: classification, imaging findings, and treatment.
Am J Neuroradiol 33(6):1007–1013. doi:10.3174/ajnr.
A2798
32. Sliker CW (2008) Blunt cerebrovascular injuries:
imaging with multidetector CT angiography.
Radiographics 28(6):1689–1708. doi:10.1148/
rg.286085521
33. Wilson CB (1992) Cryptic vascular malformations.
Clin Neurosurg 38:49–84
34. De Souza JM, Domingues RC, Cruz LCH, Domingues
FS, Iasbeck T, Gasparetto EL (2008) Susceptibility-
weighted imaging for the evaluation of patients with
familial cerebral cavernous malformations: a compari-
son with T2-weighted fast spin-echo and gradient-echo
sequences. Am J Neuroradiol 29(1):154–158.
doi:10.3174/ajnr.A0748
35. Perrini P, Lanzino G (2006) The association of
venous developmental anomalies and cavernous
malformations: pathophysiological, diagnostic,
and surgical considerations. Neurosurg Focus 21(1):e5
36. Takasugi M, Fujii S, Shinohara Y, Kaminou T,
Watanabe T, Ogawa T (2013) Parenchymal
hypointense foci associated with developmental
venous anomalies: evaluation by phase-sensitive MR
imaging at 3T. Am J Neuroradiol 34(10):1940–1944.
doi:10.3174/ajnr.A3495
37. Smith SD, Eskey CJ (2011) Hemorrhagic stroke.
Radiol Clin North Am 49(1):27–45. doi:10.1016/j.
rcl.2010.07.011
38. Hom J, Dankbaar JW, Soares BP, Schneider T, Cheng
S-C, Bredno J, Wintermark M (2010) Blood–brain
barrier permeability assessed by perfusion CT predicts
symptomatic hemorrhagic transformation and malig-
nant edema in acute ischemic stroke. Am J
Neuroradiol. doi:10.3174/ajnr.A2244
39. Jain AR, Jain M, Kanthala AR, Damania D, Stead LG,
Wang HZ, Jahromi BS (2013) Association of CT per-
fusion parameters with hemorrhagic transformation in
acute ischemic stroke. Am J Neuroradiol 34
(10):1895–1900. doi:10.3174/ajnr.A3502
40. Warach S (2004) Evidence of reperfusion injury,
exacerbated by thrombolytic therapy, in human focal
brain ischemia using a novel imaging marker of early
blood–brain barrier disruption. Stroke 35(11_Suppl_1):
2659–2661. doi:10.1161/01.STR.0000144051.32131.09
41. Hacke W, Kaste M, Bluhmki E, Brozman M,
Dávalos A, Guidetti D, Machnig T (2008)
292
Thrombolysis with alteplase 3 to 4.5 hours after acute
ischemic stroke. N Engl J Med 359(13):1317–1329
42. Ozsvath RR, Casey SO, Lustrin ES, Alberico RA,
Hassankhani A, Patel M (1997) Cerebral venography:
comparison of CT and MR projection venography. AJR
Am J Roentgenol 169(6):1699–1707
43. Atlas SW, Grossman RI, Gomori JM, Hackney DB,
Goldberg HI, Zimmerman RA, Bilaniuk LT (1987)
Hemorrhagic intracranial malignant neoplasms: spin-
echo MR imaging. Radiology 164(1):71–77.
doi:10.1148/radiology.164.1.3588929
44. Koike S, Aida N, Hata M, Fujita K, Ozawa Y, Inoue T
(2004) Asymptomatic radiation-induced telangiectasia
in children after cranial irradiation: frequency, latency,
S. Quinet and P. Turski
and dose relation. Radiology 230(1):93–99.
doi:10.1148/radiol.2301021143
45. Rabin BM, Meyer JR, Berlin JW, Marymount MH,
Palka PS, Russell EJ (1996) Radiation-induced
changes in the central nervous system and head and
neck. Radiographics 16(5):1055–1072. doi:10.1148/
radiographics.16.5.8888390
46. Jain R, Robertson PL, Gandhi D, Gujar SK, Muraszko
KM, Gebarski S (2005) Radiation-induced cavernomas
of the brain. Am J Neuroradiol 26(5):1158–1162
47. Blitstein MK, Tung GA (2007) MRI of cerebral
microhemorrhages. Am J Roentgenol 189
(3):720–725. doi:10.2214/AJR.07.2249
 ICA–ECA Collaterals
14
Maxim Mokin and Adnan H. Siddiqui

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294 Knowledge and identification of multiple anas-
tomotic connections between the external
Ophthalmic Artery Anastomoses . . . . . . . . . . . . . . . . . . 294
carotid artery and the internal carotid artery
Internal Carotid Artery Anastomoses . . . . . . . . . . . . 295 branches are paramount for the diagnosis and
Posterior Circulation Anastomoses . . . . . . . . . . . . . . . 298 treatment of vascular diseases of the brain,
head, neck, and spine. In this chapter, common
Noninvasive Imaging Modalities . . . . . . . . . . . . . . . . . . 299
extracranial–intracranial anastomoses of the
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301 ophthalmic artery, internal carotid artery, and
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301 of the posterior circulation that are critical for
the proceduralist to recognize in order to avoid
complications during embolization procedures
or open surgery are reviewed. Finally, the role
M. Mokin of noninvasive imaging in the evaluation of the
Department of Neurosurgery, School of Medicine and
collateral circulation and cerebrovascular
Biomedical Sciences, University at Buffalo, State
University of New York, Buffalo, NY, USA reserve in clinical conditions associated with
arterial occlusive disease is discussed.
Department of Neurosurgery, Gates Vascular Institute/
Kaleida Health, Buffalo, NY, USA
e-mail: maximmokin@gmail.com Keywords
A.H. Siddiqui (*) External carotid artery • Internal carotid artery
Department of Neurosurgery, School of Medicine and • Anastomosis • Stroke • Collaterals •
Biomedical Sciences, University at Buffalo, State Angiography
University of New York, Buffalo, NY, USA
Department of Neurosurgery, Gates Vascular Institute/ List of Abbreviations
Kaleida Health, Buffalo, NY, USA
AP Anteroposterior
Department of Radiology, School of Medicine and CRA Central retinal artery
Biomedical Sciences, University at Buffalo, State
ECA External carotid artery
University of New York, Buffalo, NY, USA
ICA Internal carotid artery
Toshiba Stroke and Vascular Research Center, University
MMA Middle meningeal artery
at Buffalo, State University of New York, Buffalo, NY,
USA OA Ophthalmic artery
Jacobs Institute, Buffalo, NY, USA
e-mail: asiddiqui@ubns.com; adnan.h.siddiqui@gmail.
com

# Springer Science+Business Media New York 2016 293

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_1
294
Introduction
Knowledge and identification of multiple anasto-
motic connections between the external carotid
artery (ECA) and the internal carotid artery
(ICA) branches are paramount for the diagnosis
and treatment of vascular diseases of the brain,
head, neck, and spine. The expanding role of
minimally invasive approaches for the treatment
of cerebrovascular diseases, including arteriove-
nous malformations, fistulas, moyamoya disease,
and other conditions, such as epistaxis and head
and neck tumors, requires careful exploration of
the arterial supply to these structures to ensure
success from the intervention. In this chapter,
common extracranial–intracranial anastomoses
that are critical for the proceduralist to recognize
in order to avoid complications during emboliza-
tion procedures or open surgery are reviewed. The
role of noninvasive imaging in the evaluation of
the collateral circulation and cerebrovascular
reserve in clinical conditions associated with arte-
rial occlusive disease is discussed.
The development of the cranial vascular sup-
ply during embryonic stages and a comprehensive
review of all potential extracranial–intracranial
anastomoses are beyond the scope of this chapter.
Readers are referred to the classic work by
Lasjaunias et al. [1] for a more detailed descrip-
tion of vascular anatomy on this subject.
Ophthalmic Artery Anastomoses
The ophthalmic artery (OA) most commonly
arises from the paraclinoid segment of the ICA
in adults. Variations in the OA origin can be seen
on angiography, such as from the cavernous por-
tion of the ICA (the so-called “dorsal OA”), from
the middle meningeal artery (MMA), or from the
middle cerebral artery (MCA) [2–4]. The order
and appearance of the OA branches depend on
the course of this artery in relationship to the
optic nerve [5]. The OA branches are commonly
divided into three groups: ocular, orbital, and
extraorbital. The retina and choroid are supplied
by the central retinal artery (CRA) and the ciliary
arteries. Their visualization on angiography is
M. Mokin and A.H. Siddiqui
critical to avoid occlusion of the CRA, which
will result in monocular blindness. The CRA is
the first branch of the OA, when the OA crosses
above the optic nerve. When the OA crosses
below the optic nerve, the lateral posterior ciliary
artery originates from the OA first and the CRA
second. In cases in which the CRA cannot be
reliably identified on diagnostic angiography
and embolization is planned, provocative testing
with superselective injection of sodium amobar-
bital and lidocaine should be considered to iden-
tify whether potential anastomoses to the eye
exist at or distal to the site of microcatheter
placement.
Visualization of a choroidal blush (best appre-
ciated during angiography on lateral views as a
thin crescent of contrast material) on ECA injec-
tions should raise suspicion for potential anasto-
moses with the OA [6]. However, it should be
noted that absence of the choroidal blush does
not exclude the existence of such anastomoses.
For example, when embolization is executed in
the territory known to have potential connections
with the OA, repeat diagnostic angiographic runs
should be performed, as dangerous anastomoses
might become visible only following occlusion of
ECA branches with embolic agents. Identification
of OA–MMA anastomoses can also help avoid
visual complications in cranial base surgery, such
as during resection of meningiomas [7].
An isolated OA origin of the MMA is a rare
anatomic variant, especially when present bilater-
ally [8, 9]. In such cases, the entire MMA or some
of its branches are not visualized on a selective
ECA injection. An ICA injection will opacify the
MMA in a slightly delayed fashion, via the prox-
imal OA through the superior orbital fissure
(Fig. 1). Embolization of a lesion supplied by
such anomalous meningeal vessels is challenging
because of the high risk for emboli into OA
branches, causing retinal ischemia [10].
Multiple anastomoses between branches of the
ECA and OA exist, and only the most clinically
relevant variants will be described here. For a
more detailed review, the articles by Hayreh [11]
and Perrini et al. [5] are recommended. The recur-
rent meningeal artery is a critical anastomosis
when attempting embolization of convexity
14 ICA–ECA Collaterals
Fig. 1 Ophthalmic origin of the middle meningeal artery.
Cerebral angiogram, lateral projection, early (a) and late
arterial phase (b) of selective left internal carotid artery
injection demonstrate reconstitution of the middle menin-
geal artery (white arrows) via the ophthalmic artery (black
meningiomas. It typically projects posteriorly to
the OA through the superior orbital fissure.
Another important group of anastomoses is
between the anterior and posterior ethmoidal
branches of the OA and the distal branches of
the internal maxillary artery, namely, the
sphenopalatine and greater palatine arteries
(Fig. 2). The presence of such anastomoses should
be explored on cerebral angiography when plan-
ning embolization for epistaxis [12].
Additional examples of potential anastomoses
involving the OA include the superficial temporal
artery branches such as the zygomatico-orbital
artery, terminal nasal branches of the facial artery,
and the infraorbital branch of the internal
maxillary artery (Fig. 3). Extracranial–intracranial
295
arrow) through the recurrent meningeal artery (black
arrowhead). (c) Notice that the middle meningeal artery
does not opacify on the selective external carotid artery
injection. sta superficial temporal artery, ima internal max-
illary artery
anastomoses have been described between the
MMA, superficial temporal artery, occipital artery,
and the cortical leptomeningeal arteries [13]
(Fig. 4).
Internal Carotid Artery Anastomoses
Important anastomoses within the petrocavernous
segment of the ICA are described below. They can
often be identified on cerebral angiograms in
cases of cervical ICA occlusion and are a critical
component of dural arteriovenous malformations
in this region.
The inferolateral trunk is a remnant of the
dorsal OA and consists of several branches that
296
Fig. 2 Ophthalmic artery–internal maxillary artery anas-
tomoses. Cerebral angiogram in a patient with chronic
right internal carotid artery occlusion (asterisk on
anteroposterior [AP] view), AP (a) and lateral (b) views,
shows reconstitution of the ophthalmic artery (white
arrows) through the anterior and posterior ethmoidal
Fig. 3 Superficial temporal artery anastomoses to the
ophthalmic artery. Cerebral angiogram, AP (a) and lateral
(b) views, shows opacification of the ophthalmic artery
(black arrows) and the supraclinoid internal carotid artery
M. Mokin and A.H. Siddiqui
branches (black arrows point to the nasal blush at the site
of anastomoses within the nasal cavity). White arrowhead
– infraorbital artery of the internal maxillary artery. Black
arrowhead – cavernous internal carotid artery; ima internal
maxillary artery
(black arrowheads) via the zygomatico-orbital branch
(white arrow) of the superficial temporal artery. Another
anastomosis involves the infraorbital branch (white arrow-
heads) of the internal maxillary artery (ima)
14 ICA–ECA Collaterals 297

Fig. 4 Cortical
leptomeningeal
anastomoses. Cerebral
angiogram, early arterial
intracranial AP (a) and
lateral (b) views and late
arterial intracranial AP (c)
and lateral (d) views, shows
cortical leptomeningeal
anastomoses between the
middle meningeal artery
(black arrows) and the
middle cerebral artery
(white arrows)

provide supply to the cranial nerves and the
gasserian ganglion. Anteroinferiorly, a character-
istic serpentine-like anastomosis with the maxil-
lary artery through the artery of the foramen
rotundum can be observed (Fig. 5). Other anasto-
moses include the recurrent artery of foramen
lacerum (which connects with the ascending pha-
ryngeal artery), branches of the MMA (seen lat-
erally through the foramen spinosum), branches
of the accessory meningeal artery (seen posteri-
orly), and branches of the ascending pharyngeal
artery (carotid branch of the superior pharyngeal
branch through the foramen lacerum) (Fig. 6).
A rare finding on cerebral angiography is
hypertrophy of the vasa vasorum after complete
occlusion of the cervical ICA [14] (Fig. 7). Exten-
sive proliferation of the vasa vasorum can lead to
anterograde recanalization of the ICA distally, such
as at the level of the petrocavernous segment, and
can be confused with anastomoses originating from
the ascending pharyngeal artery. Visualization of
tortuous vascular channels that project over the
expected course of the cervical ICA rather than
the ascending pharyngeal artery helps in differen-
tiating the two conditions. The ascending pharyn-
geal artery typically originates from the posterior
wall of the ECA trunk, although rarely its origin
can be seen from the proximal cervical ICA.
The meningohypophyseal trunk of the ICA
supplies the pituitary gland, the clival region,
298 M. Mokin and A.H. Siddiqui

and the meninges of the tentorium. Through its

tentorial branches, which are seen more anteriorly,
it can form anastomoses with petrous branches of
the MMA. Another common anastomosis of the
meningohypophyseal trunk involves its lateral
clival artery and neuromeningeal branches of the
ascending pharyngeal artery.

Posterior Circulation Anastomoses

Early during embryologic development of the

Fig. 5 Artery of foramen rotundum. Cerebral angiogram,
AP view, demonstrates a characteristic serpentine-like artery
of foramen rotundum (white arrows) directed at the base of
the sella, which is an anastomosis between the internal
maxillary artery (ima) and the cavernous segment of the
internal carotid artery (black arrow). oa ophthalmic artery
central nervous system, the posterior circulation
forms primitive anastomoses with the ICAs (the
trigeminal, otic, hypoglossal, and proatlantal con-
nections) that eventually disappear, except for the
posterior communicating arteries, which remain
in adults. The most common anastomosis
involves the occipital artery and the vertebral
artery connecting at the C1 or C2 segmental levels
(Fig. 8). Other examples of anastomotic connec-
tions with the vertebral artery include the ascend-
ing pharyngeal artery and the ascending cervical
and deep cervical arteries.

Fig. 6 Ascending
pharyngeal artery
anastomosis. Cerebral
angiogram, AP (a) and
lateral (b) views, shows a
typical appearance of the
ascending pharyngeal
artery. The superior
pharyngeal branch of the
pharyngeal trunk (white
arrows) gives off the carotid
branch, which creates an
anastomosis with the
inferolateral trunk (not
shown). cca common
carotid artery, fa facial
artery, ima internal
maxillary artery, oa
occipital artery
14 ICA–ECA Collaterals 299

Fig. 7 Vasa vasorum of the

internal carotid artery.
Cerebral angiogram,
cervical AP (a) and lateral
(b) views and cranial AP (c)
and lateral (d) views in a
case of chronic cervical left
internal carotid artery
occlusion. The petrous
segment of the internal
carotid artery (black
arrows) reconstitutes via a
prominent network of vasa
vasorum, which appear as
tortuous channels following
the expected course of the
cervical internal carotid
artery (white arrows)

to estimate the ability of such connections to pro-

Noninvasive Imaging Modalities
Catheter angiography remains the gold standard
for the identification of specific anastomotic con-
nections between the ECA and ICA branches
because of its excellent spatial resolution and the
proceduralist’s ability to perform superselective
catheterization of individual vessels. Noninvasive
imaging can be used in clinical practice as a way
vide adequate cerebral blood flow in conditions
such as occlusive disease of the ICA and the
vertebrobasilar system and moyamoya disease.
Measurement of cerebrovascular reserve can
help predict stroke risk in patients with carotid
stenosis or occlusion, based on a meta-analysis
of studies evaluating cerebrovascular reserve
impairment [15].
300 M. Mokin and A.H. Siddiqui

The techniques used most commonly to eval-

Fig. 8 Occipital artery–vertebral artery anastomosis.
Cerebral angiogram, lateral view, selective occipital artery
(oa) injection, demonstrates opacification of the
vertebrobasilar system (white arrows) via the C1 segmen-
tal artery (black arrow)
uate cerebrovascular reserve include computed
tomographic and magnetic resonance perfusion
imaging, single-photon emission computed
tomography, and positron emission tomography
[16–20] (Fig. 9). Systemic administration of cere-
bral vasodilator agents that can cross the
blood–brain barrier, such as acetazolamide,
allows the calculation of changes in cerebral
blood flow and visualization of reactivity in spe-
cific vascular territories in which vasoocclusive
disease is present [21] (Fig. 10).
Arterial spin-labeling magnetic resonance
imaging is another technique that can assess tem-
poral dynamics of arterial blood inflow and iden-
tify brain regions with impaired hemodynamics
[18, 22, 23]. It can be used to visualize the perfu-
sion territory of the ECA in order to depict the
status of the collateral circulation and quantita-
tively compare flow pre- and postoperatively [23].

Fig. 9 Watershed
perfusion deficit. In a
patient with chronic
occlusive disease of the left
internal carotid artery, there
is a perfusion deficit in the
anterior cerebral artery
(ACA)–middle cerebral
artery watershed territory
(indicated by dotted lines)
14 ICA–ECA Collaterals
Fig. 10 Acetazolamide challenge. (a) In a patient with
chronic right internal carotid artery occlusion, there is
decreased cerebral blood flow in the right internal carotid
artery territory, in comparison to the contralateral side.
Summary
To ensure success and safety when treating cere-
brovascular diseases of the head and neck,
proceduralists should be aware of dangerous
extracranial–intracranial carotid arterial anasto-
moses, which are highly variable in their appear-
ance and location. Catheter angiography remains
the gold standard for the identification of such
anastomotic routes and should be carefully
reviewed before performing embolization of arte-
riovenous malformations, tumors, or epistaxis.
Noninvasive imaging can estimate the ability of
such connections to provide adequate cerebral
blood flow in conditions such as occlusive disease
of the ICA and the vertebrobasilar system.
References
1. Lasjaunias P, Berenstein A, ter Brugge K (2001) Sur-
gical neuroangiography: 1 clinical vascular anatomy
and variations. Springer, Berlin
2. Sade B, Tampieri D, Mohr G (2004) Ophthalmic artery
originating from basilar artery: a rare variant. AJNR
Am J Neuroradiol 25:1730–1731
301
(b) A robust increase in cerebral blood flow within the
right hemisphere is seen after administration of acetazol-
amide, indicating good cerebrovascular reserve
3. Watanabe A, Hirano K, Ishii R (1996) Dural carotico-
cavernous fistula with both ophthalmic arteries arising
from middle meningeal arteries. Neuroradiology
38:806–808
4. Uchino A, Saito N, Takahashi M, Kozawa E,
Mizukoshi W, Nakajima R, Okano N (2013) Persistent
dorsal ophthalmic artery and ophthalmic artery arising
from the middle meningeal artery diagnosed by MR
angiography at 3 T. Surg Radiol Anat 35:775–782
5. Perrini P, Cardia A, Fraser K, Lanzino G (2007) A
microsurgical study of the anatomy and course of the
ophthalmic artery and its possibly dangerous anasto-
moses. J Neurosurg 106:142–150
6. Schurr PH (1951) Angiography of the normal ophthal-
mic artery and choroidal plexus of the eye. Br J
Ophthalmol 35:473–478
7. Hayashi N, Kubo M, Tsuboi Y, Nishimura S,
Nishijima M, Ahmed Abdel-Aal M, Endo S (2007)
Impact of anomalous origin of the ophthalmic artery
from the middle meningeal artery on selection of sur-
gical approach to skull base meningioma. Surg Neurol
68:568–572
8. Plas B, Bonneville F, Dupuy M, Sol JC, Chaynes P
(2013) Bilateral ophthalmic origin of the middle men-
ingeal artery. Neurochirurgie 59:183–186
9. Royle G, Motson R (1973) An anomalous origin of the
middle meningeal artery. J Neurol Neurosurg Psychia-
try 36:874–876
10. Trivelatto F, Nakiri GS, Manisor M, Riva R,
Al-Khawaldeh M, Kessler I, Mounayer C (2011) Pre-
operative onyx embolization of meningiomas fed by
302
the ophthalmic artery: a case series. AJNR Am J
Neuroradiol 32:1762–1766
11. Hayreh SS (2006) Orbital vascular anatomy. Eye
(Lond) 20:1130–1144
12. Willems PW, Farb RI, Agid R (2009) Endovascular
treatment of epistaxis. AJNR Am J Neuroradiol
30:1637–1645
13. Handa J, Waga S, Handa H (1971) Dural-cortical arte-
rial anastomosis as a collateral channel in carotid occlu-
sive disease. Clin Radiol 22:302–307
14. Meguro T, Muraoka K, Terada K, Hirotsune N,
Nishino S (2011) Recanalisation of the internal carotid
artery via the vasa vasorum after coil occlusion. Br J
Radiol 84:e23–e26
15. Gupta A, Chazen JL, Hartman M, Delgado D,
Anumula N, Shao H, Mazumdar M, Segal AZ,
Kamel H, Leifer D, Sanelli PC (2012) Cerebrovascular
reserve and stroke risk in patients with carotid stenosis
or occlusion: a systematic review and meta-analysis.
Stroke 43:2884–2891
16. Konstas AA, Goldmakher GV, Lee TY, Lev MH (2009)
Theoretic basis and technical implementations of CT
perfusion in acute ischemic stroke, part 1: theoretic
basis. AJNR Am J Neuroradiol 30:662–668
17. van Laar PJ, van der Grond J, Mali WP, Hendrikse J
(2006) Magnetic resonance evaluation of the cerebral
circulation in obstructive arterial disease. Cerebrovasc
Dis 21:297–306
M. Mokin and A.H. Siddiqui
18. Donahue J, Sumer S, Wintermark M (2014) Assess-
ment of collateral flow in patients with cerebrovascular
disorders. J Neuroradiol 41:234–242
19. van Laar PJ, van der Grond J, Bremmer JP, Klijn CJ,
Hendrikse J (2008) Assessment of the contribution of
the external carotid artery to brain perfusion in patients
with internal carotid artery occlusion. Stroke
39:3003–3008
20. Ozgur HT, Kent Walsh T, Masaryk A, Seeger JF,
Williams W, Krupinski E, Melgar M, Labadie E
(2001) Correlation of cerebrovascular reserve as mea-
sured by acetazolamide-challenged SPECT with angio-
graphic flow patterns and intra- or extracranial arterial
stenosis. AJNR Am J Neuroradiol 22:928–936
21. Vagal AS, Leach JL, Fernandez-Ulloa M, Zuccarello
M (2009) The acetazolamide challenge: techniques and
applications in the evaluation of chronic cerebral ische-
mia. AJNR Am J Neuroradiol 30:876–884
22. Bokkers RP, van Laar PJ, van de Ven KC, Kapelle LJ,
Klijn CJ, Hendrikse J (2008) Arterial spin-labeling MR
imaging measurements of timing parameters in patients
with a carotid artery occlusion. AJNR Am J
Neuroradiol 29:1698–1703
23. Dang Y, Wu B, Sun Y, Mo D, Wang X, Zhang J, Fang J
(2012) Quantitative assessment of external carotid
artery territory supply with modified vessel-encoded
arterial spin-labeling. AJNR Am J Neuroradiol
33:1380–1386
 Imaging Biomarkers of Angiogenesis
and the Microvascular Environment 15
in Cerebral Tumors

Alan Jackson, Ibrahim Djoukhadar, and David J. Coope

Contents Clinical Applications of Microvascular

Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304 Diagnosing Glioma Mimics . . . . . . . . . . . . . . . . . . . . . . . . . 321
Abnormalities of Tumor Blood Vessels . . . . . . . . . . . 304 Grading Glioma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
Identifying Transformation in Low-Grade
Antiangiogenic Therapies . . . . . . . . . . . . . . . . . . . . . . . . . . 307 Glioma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
Distinguishing Recurrent Tumor from Treatment
Imaging the Tumor Vascular
Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
Microenvironment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
Predicting Treatment Response . . . . . . . . . . . . . . . . . . . . . . 323
Imaging Biomarkers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
Biomarkers of Perfusion and the Vascular
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
Microenvironment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
Dynamic Susceptibility Contrast-Enhanced
MRI (DSCE-MRI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
Dynamic Relaxivity Contrast-Enhanced
MRI (DRCE-MRI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
Dynamic Contrast-Enhanced Computed
Tomography (DCE-CT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
Imaging Microvessel Morphology and Hypoxia
with MRI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
Vessel Size Imaging (VSI) and Vessel
Architectural Imaging (VAI) . . . . . . . . . . . . . . . . . . . . . . . . . 312
Capillary Heterogeneity Imaging . . . . . . . . . . . . . . . . . . . . 314
Arteriovenous Overlap Imaging (AVOL) . . . . . . . . . . . 315
Imaging Hypoxia Using MRI . . . . . . . . . . . . . . . . . . . . . . . 316
Molecular Imaging of the Tumor Vascular
Microenvironment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
Imaging Hypoxia Using PET . . . . . . . . . . . . . . . . . . . . . . . . 319
Imaging Integrin Expression . . . . . . . . . . . . . . . . . . . . . . . . . 319

A. Jackson (*) • I. Djoukhadar • D.J. Coope

Wolfson Molecular Imaging Centre, University of
Manchester, Withington, Manchester, UK
e-mail: Alan.Jackson@manchester.ac.uk; Ibrahim.
Djoukhadar@Manchester.ac.uk; david.
coope@manchester.ac.uk

# Springer Science+Business Media New York 2016 303

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_18
304
Abstract
The structure and organization of blood vessels
within tumor tissue is very different from that
seen in normal tissues. Tumor blood vessels
show abnormalities in microstructure and hier-
archical organization, which result from multi-
ple factors including local tumor
characteristics, angiogenic drive, and the abil-
ity of the angiogenic process to keep pace with
tumor growth. Tumor microvasculature is inef-
ficient compared to that seen in normal tissues,
and, particularly in rapidly growing tumors,
blood flow is often inadequate to meet the
demands for oxygen and nutrient delivery and
clearance of waste material. Understanding the
microvascular environment and its variation
between and within tumors is critical for an
understanding of tumor behavior and therapeu-
tic response. A wide range of quantitative
imaging techniques have been developed in
an attempt to provide noninvasive, repeatable
assays of microvascular characteristics which
can then be studied in terms of their spatial
variability and change over time. This chapter
reviews the currently available imaging bio-
markers and their current clinical application.
Keywords
Angiogenesis • Blood vessels • Permeability •
Flow • Dynamic contrast-enhanced imaging •
Positron emission tomography • Hypoxia •
Integrins
Introduction
The development of new blood vessels (angiogen-
esis) is an essential component of tumor growth
and is one of the classic hallmarks of the cancer
cell [1]. The development of new blood vessels,
their maturation, and organization give rise to a
distinctive vascular microenvironment that has
been shown to vary significantly between tumor
types and even within individual tumors. The
characteristics of the vascular microenvironment
will place constraints on the delivery of nutrients,
oxygen, and drugs as well as the distribution of
A. Jackson et al.
macromolecules, which can in turn directly affect
the behavior of tumor cells. The development of
therapeutic approaches, which target the angio-
genic process or the neovasculature, has led to
enormous research activity in this area over the
past 20 years. Imaging techniques have been
widely applied to study the process of angiogen-
esis and the characteristics of the tumor vascular
microenvironment (TVM). Imaging has unique
strengths allowing repeated, noninvasive assess-
ment of the whole tumor or multiple tumors in the
same patient. In addition, imaging provides spa-
tial data allowing the study of regional variations
in the microenvironment within a single tumor.
Imaging also allows the development of quantita-
tive metrics, which reflect the physiological, phar-
macological, or pathological processes occurring
within the tumor. Such quantitative metrics are
referred to as imaging biomarkers (IB), and con-
siderable research activity has been focused on the
discovery, technical, and biological validation of
IB that allow characterization of the TVM. In this
chapter, the characteristics of the TVM and the IB
available for its characterization and their current
applications are reviewed.
Abnormalities of Tumor Blood Vessels
In normal tissues a single layer of endothelial cells
(EC) lines blood vessels forming a barrier limiting
the passage of macromolecules. EC are
surrounded by tightly connected pericytes and
both of these enveloped in a basement membrane.
In normal tissues pericytes are heavily involved in
the suppression of EC proliferation and EC are
quiescent, showing no evidence of division or
sprouting of new vessels [2]. Pericytes produce a
number of signaling molecules, which include
vascular endothelial growth factor (VEGF),
angiopoietin-1 (Ang-1), fibroblast growth factors
(FGFs), NOTCH, and chemokines, which support
normal endothelial cell repair. Endothelial cells
can also respond directly to dissolved oxygen
concentration and have receptors for hypoxia-
inducible factors such as prolyl-hydroxylase
domain 2 (PHD2) and hypoxia-inducible factor-
2α (HIF-2α) which support mechanisms to enable
15 Imaging Biomarkers of Angiogenesis and the Microvascular Environment in Cerebral Tumors 305

remodeling of vessels to optimize oxygen delivery. capillary network, which drains into venules and
Within normal tissue blood vessels are arranged in larger veins. These vascular networks vary from
a clear hierarchy, with inflow of blood managed by tissue to tissue but are, in general, highly structured
arterial vessels and branching arterioles feeding a and internally consistent.

Loosening junctions
a Selection of tip cell (VE-Cadherin)
Matrix remodelling (MMPs)
Tip-cell formation
(VEGFR-2, DLL4, JAGGED1
NRP1, integrins, HIF-1α
MT1-MMP, PGC-1α)
Angiogenic factors
(VEGF, VEGF-C, FGFs,
ANG-2, chemokines)
Pericyte detachment
(ANG-2)
Quiescent vessel Flow
Permeability, vasodilation
and extravasation (VEGF)
b Stalk elongation and tip guidance
Lumen formation
(VE-cadherin, CD34, sialomucins, VEGF)

Pericyte recruitment (PDGF-B, ANG-1,

NOTCH, ephrin-B2, FGF)
Tip-cell guidance and adhesion
(semaphorins, ephrins, integrins)

Liberation of angiogenic
factors from ECM
(VEGF, FGFs)

ECM
Stalk elongation Myeloid cell
Flow recruitment (ANG-2, Adjacent vessel
(VEGFR-1, NOTCH, WNT,
SDF-1α, PIGF) sprout
NRARP, PIGF, FGFs, EGFL7)

c Quiescent phalanx resolution

Phalanx cell
Transendothelial lipid transport (VEGF-B)
(PHD2, HIF-2α
Vascular maintenance VE-cadherin, TIE-2)
(VEGF, ANG-1, FGFs, NOTCH)
Barrier formation
(VE-cadherin, ANG-1)

Pericyte maturation
Basement membrane (PDGF-B, PDGFR-β
Flow deposition (TIMPs, PAI-1) ephrin-B2, ANG-1,
NOTCH, TGF-β1)

Fig. 1 (continued)
306
Tumor vasculature demonstrates a wide range
of abnormalities dependent on the tumor type. As
tumors develop, they depend on the development
of new blood vessels to meet increasing demand
for nutrient delivery and waste removal. This pro-
cess of neovascularization is referred to as angio-
genesis [2]. Angiogenesis is controlled by a
number of signaling mechanisms mediated by
small molecules known as cytokines which may
be released from tumor cells, EC, hypoxic normal
cells, or inflammatory tissue [3]. In normal tissue
the angiogenic process forms a normal part of
development [2]. In tumors, the presence of hyp-
oxia and hypoglycemia promotes the release of
pro-angiogenic signaling molecules including
VEGF, Ang-2, FGFs, and chemokines into the
local microenvironment. In normal vessels
exposed to these pro-angiogenic stimuli, pericytes
detach from the basement membrane under the
action of matrix metalloproteinases (MMP) and
the tight junctions between endothelial cells
loosen, increasing the permeability of the endo-
thelial membrane (Fig. 1).
Plasma proteins extravasate into the extravas-
cular space and form an early extracellular matrix,
while integrin signaling stimulates migration of
EC onto the matrix. New vessels grow in an
ordered process that involves migration of a single
EC, known as the tip cell along the new matrix.
Nearby EC, known as stalk cells, divide to form
an elongated vascular core that then develops a
Fig. 1 The consecutive steps of blood vessel branching are
shown, with the key molecular players involved denoted in
parentheses. (a) After stimulation with angiogenic factors,
the quiescent vessel dilates and an endothelial tip cell is
selected (DLL4 and JAGGED1) to ensure branch formation.
Tip-cell formation requires degradation of the basement
membrane, pericyte detachment, and loosening of endothe-
lial cell junctions. Increased permeability permits extrava-
sation of plasma proteins (such as fibrinogen and
fibronectin) to deposit a provisional matrix layer, and pro-
teases remodel preexisting interstitial matrix, all enabling
cell migration. For simplicity, only the basement membrane
between endothelial cells and pericytes is depicted, but in
reality, both pericytes and endothelial cells are embedded in
this basement membrane. (b) Tip cells navigate in response
to guidance signals (such as semaphorins and ephrins) and
A. Jackson et al.
patent lumen in response to a range of signaling
molecules. These new vessels fuse with existing
ones to initiate blood flow and, in time, develop a
pericyte coverage. As they mature they may
reform EC tight junctions and basement
membranes.
In general terms, rapidly growing aggressive
tumors provide greater angiogenic drive and will
result in increasingly disordered microvascular
environments. In some, slowly growing, tumors
such as low-grade meningiomas, the microvas-
cular architecture can be remarkably structured
with well-developed hexagonal arteriolar and
capillary territories clearly seen. In more aggres-
sive tumors such as glioma, tumor vessels are
disorganized, irregular, and tortuous with bizarre
branching patterns. They do not maintain the
normal arterial/capillaries/venous hierarchy and
all components of the tumor vessel wall are
abnormal. EC tight junctions are absent, render-
ing tumor vessels leaky to macromolecules and
often with a propensity to microhemorrhage.
These neovascular structures are dynamic, show-
ing constant changes in blood vessel develop-
ment and variations in hemodynamics so that
blood flow in individual regions may change in
volume and direction even over short time
periods. Primary malignant brain neoplasms
exhibit the greatest degree of vascular prolifera-
tion of any tumor type, and angiogenesis is a key
step in their progression [4].
adhere to the extracellular matrix (mediated by integrins) to
migrate. Stalk cells behind the tip cell proliferate, elongate,
and form a lumen, and sprouts fuse to establish a perfused
neovessel. Proliferating stalk cells attract pericytes and
deposit basement membranes to become stabilized.
Recruited myeloid cells such as tumor-associated macro-
phages (TAMs) and TIE-2-expressing monocytes (TEMs)
can produce pro-angiogenic factors or proteolytically liber-
ate angiogenic growth factors from the ECM. (c) After
fusion of neighboring branches, lumen formation allows
perfusion of the neovessel, which resumes quiescence by
promoting a phalanx phenotype, reestablishment of junc-
tions, deposition of basement membrane, maturation of
pericytes, and production of vascular maintenance signals.
Other factors promote transendothelial lipid transport (From
Ref. [3] with permission)
15 Imaging Biomarkers of Angiogenesis and the Microvascular Environment in Cerebral Tumors
Most tumors have far less blood flow than
would be expected from their vascularity,
reflecting the inefficiency of the tumor vascula-
ture. In addition, leakage of large molecules and
the lack of normal lymphatic drainage can result
in high interstitial pressures, which further impair
regional perfusion. Consequently, regional hyp-
oxemia and necrosis are common, particularly in
aggressive tumors.
Antiangiogenic Therapies
The ubiquitous nature of the angiogenic process
led to great excitement in the 1990s with the hope
that therapeutic approaches, which targeted path-
ological angiogenesis, would provide a “magic
bullet” that could be used to deprive a large
range of tumors of the blood vessels necessary
for development and expansion. The development
of antiangiogenic drugs and of antivascular
agents, which target abnormal, developing ves-
sels, led to one of the most extensive areas of
anticancer research in recent decades. Despite
this, antiangiogenics, while showing efficiency
as single or combination agents in a number of
tumor types, have failed to provide the prolonged
tumor suppression of tumor activity commonly
observed in animal models. Many mechanisms
have been identified which imbue tumor cells,
EC, or other components of the tumor stroma
with resistance to angiogenic inhibition, and
some agents have proven toxic when combined
with chemotherapy [5]. Compensatory increase in
other pro-angiogenic signaling pathways can sup-
port tumor angiogenesis independently from
VEGF, and some evidence is beginning to suggest
that antiangiogenic therapy may cause cancer
cells to become increasingly malignant in behav-
ior. Secondary hypoxia can induce mutations
resulting in hypoxic-resistant tumor cells and
alternate pro-angiogenic “escape” mechanisms.
In glioma there is also evidence that EC can be
formed from primary tumor stem cells via a pro-
cess that is relatively insensitive to blockade of
VEGF. While the lack of curative single-agent
efficiency is disappointing, it is clear that
307
antiangiogenic agents offer significant survival
benefit in combination therapies and research
into this mechanism of drug action continues,
albeit with new expectations. The main questions
which now face imaging researchers in this field
are more demanding and include: Can IB predict
therapeutic response to specific agents or combi-
nation therapies? Can IB guide selection of opti-
mal therapeutic combinations? Can IB be used to
optimize the timings of combination therapies?
Can IB identify and/or predict specific mecha-
nisms of resistance and appropriate subsequent
interventions?
Imaging the Tumor Vascular
Microenvironment
It is clear that the TVM represents one of the most
profoundly important factors both in tumor behav-
ior and therapeutic response. It controls the ability
of individual regions of proliferating tumor cells to
obtain sufficient nutrients and oxygen. Failure of
vascular development, adequate to the require-
ments of the developing tumor, results in regional
hypoxia with the consequent effects on local tumor
and blood vessel growth and a drive toward geno-
typic instability and tumor cell mutation. It can also
result in regional necrosis and tumor cell death.
Features of the TVM also control the passage of
macromolecules and inflammatory cells into the
tissue and the generation of hematogenous meta-
static deposits. The ability to study and measure
these biological processes is essential, and many
can be characterized and quantified in exquisite
detail on the basis of tissue biopsy or even by
examination of circulating tumor cells and soluble
biomarkers. Despite this, imaging techniques
remain the only available approach to studying
the TVM in situ without the need for biopsy or
surgical resection. Furthermore, imaging tech-
niques support studies of the entire tumor, can
provide information on the heterogeneity of indi-
vidual biological features, and can, in many cases,
be optimized to provide biological data of direct
relevance to tumor biology. These features,
together with the minimally invasive nature of
308
most imaging techniques and the ability to repeat
studies whenever necessary, give imaging signifi-
cant advantages over other methodological
approaches.
In recent years, the concept of the imaging
biomarker (IB) has become the basis for the devel-
opment of novel imaging methods. A biomarker is
defined as “a quantitative parameter which directly
reflects the physiology, anatomy or pathological
process under investigation.” The use of soluble
and tissue biomarkers, like prostate-specific anti-
gen or CA 125, has led to the development of
“biomarker roadmaps” designed to standardize
the discovery, technical, and biological validation
of individual biomarkers to ensure that they are fit
for purpose. The concept of an IB is identical and
requires identification of a potentially valuable,
measurable imaging feature, proof that it is techni-
cally measurable in single or multi-institutional
settings, characterization of its repeatability and
reproducibility, and evidence to support its biolog-
ical validity. For imaging techniques the technical
validation steps are often more extensive than those
required for soluble and tissue biomarkers and can
be technically challenging, particularly where
imaging techniques must be standardized across
different manufacturers, imaging platforms, and
institutions. Despite this, the development of
novel IB now forms one of the main areas of
imaging research both in cancer and other diseases.
In the remainder of this chapter, the IB avail-
able for quantified imaging of the TVM is
reviewed, the evidence for their technical and
biological validity is discussed, and their applica-
tions to the present time are reviewed.
Imaging Biomarkers
The development of IB to characterize the tumor
microenvironment and the angiogenic process has
been and continues to be a major area of imaging
research. All major modalities have been utilized,
including CT, MR, single-photon emission com-
puted tomography (SPECT), positron emission
tomography (PET), ultrasound, and optical imag-
ing. The last two are largely experimental, being
A. Jackson et al.
used primarily in the intraoperative setting or in
animal models of disease, and will not be
discussed further.
Most early imaging studies focused on the char-
acterization of aspects of the TVM involved in the
angiogenic process, specifically the amount of
blood vessels and the increases in capillary endo-
thelial permeability induced by pro-angiogenic
cytokines. Both of these features can be estimated
using dynamic contrast-enhanced imaging tech-
niques, which are now widely employed both in
clinical practice and in trials of antiangiogenic and
vascular targeting agents. However, more recently
there has been significant interest in developing
biomarkers that target other aspects of the TVM,
leading to the description of an increasingly wide
range of IB and, in particular, of PET-based molec-
ular imaging agents.
Biomarkers of Perfusion
and the Vascular Microenvironment
The early development of IB for tumor microvas-
cular structure focused on attempts to measure the
known histological features of angiogenic tissue.
Classically, histochemical staining studies had
identified the importance of the density of new
vessels (microvascular density; MVD) measured
on stained tissue. MVD has been shown to relate
to malignancy and outcome in a number of tumors
and is routinely used in histological studies [6]. It
was quickly realized that a theoretical surrogate of
MVD could be extracted from dynamic contrast-
enhanced MRI.
Dynamic Susceptibility Contrast-
Enhanced MRI (DSCE-MRI)
Early studies concentrated on the use of dynamic
T2*WI acquisitions, often referred to as Dynamic
Susceptibility Contrast-Enhanced MRI (DSCE-
MRI), in neoplastic and cerebral vascular disease.
The initial focus on DSCE-MRI arose from the
fact that system architecture at that time made
rapid dynamic acquisitions with T2* star
15 Imaging Biomarkers of Angiogenesis and the Microvascular Environment in Cerebral Tumors
weighting more straightforward than dynamic
T1-WI imaging. More importantly, the signal
change in T2*WI images arises from the effects
of the locally altered magnetic field which extends
some distance from the contrast molecule. In iso-
lated capillary beds this means that signal is gener-
ated from protons within the extravascular
extracellular space (EES), even if contrast mole-
cules remain entirely intravascular. This suscepti-
bility effect gives rise to slightly higher
proportional signal changes in the normal gray
and white matter which is a significant advantage
in normal cerebral capillary beds. The analysis of
the dynamic signal intensity time course curves
(SI-TCC) from DSCE-MRI is also relatively
straightforward. Dynamic time course signal
changes can be transformed into estimates of con-
trast concentration time course curves (CAA-TCC)
from which estimates of proportional cerebral
blood volume can be derived as the area under
the contrast agent concentration curve during its
first passage. This area correlates with the propor-
tion of the volume of the voxel filled by blood
vessels often referred to as the relative cerebral
blood volume (rCBV). By scaling to normal
white matter or pure vascular tissues (large veins),
rCBV can be normalized to produce absolute
values (CBV), which have shown close agreement
with histological assessment of MVD in a number
of cancer types and show similar correlations with
tumor grade, aggressiveness, and survival data [7].
The basic analysis approaches for the DSCE-
MRI assumed that contrast agent remains intra-
vascular in the presence of an intact blood–brain
barrier. In tumor tissue this causes very significant
errors since contrast agent leakage into the EES
produces contradictory signal changes, due to T1
effects. These give rise to artificially low estimates
of CBV (so-called T1 shine through). These leak-
age errors can be reduced by preloading the tissue
with a dose of contrast agent prior to the dynamic
contrast study, by the use of low flip angle
sequences with low T1 sensitivity, or by signal
post-processing [8]. Increasingly, contrast agent
preloading, signal processing, or a combination of
the two is routinely used. The use of low flip angle
acquisition sequences has been relatively
309
unpopular since, although these do abrogate the
effect of the so-called T1 shine through signal,
they also result in significant decreases in signal-
to-noise ratio with consequent reductions in the
accuracy of subsequent curve fitting analyses.
The growing application of DSCE-MRI in
neuro-oncology led to a number of additional
analytical approaches designed to provide addi-
tional data about the TVM. It was rapidly realized
that in tumor vasculature, the normal shape of the
CC-TCC was distorted by retention of contrast
agent following the first passage of the contrast
bolus. This elevation of contrast concentration
was first described as the relative recirculation
(rR) parameter and has also been widely referred
to as the percentage recovery (describing the per-
centage of signal recovery to the baseline). This
has been variously attributed to contrast linkage
and to retention of contrast agent within the abnor-
mal microvasculature due to vessel tortuosity and
regional reductions in perfusion pressure. These
metrics allow identification of areas of tumor
where capillary flow is present, but delayed,
corresponding to regions with reduced perfusion
pressure and reduced flow and has been used to
differentiate GBM from metastasis [9].
The development of more complex image anal-
ysis techniques led to the development of pharma-
cokinetic models designed with the intention of
measuring regional blood flow and/or capillary
endothelial permeability. These approaches require
correction of the tissue residue function from the
individual tumor voxels to correct for variations in
the width of the injected arterial bolus, known as the
arterial input function (AIF). Correcting for the
variation between patients resulting from individual
variations in AIF allows estimations of regional
capillary blood flow. Unfortunately, the accuracy
of these estimates is greatly affected by delays in
the arrival of the contrast bolus within the tissue and
by variations in dispersion of the contrast bolus,
which result from a combination of flow delays
and irregular vessel flow patterns. The result of
these effects is that regional capillary flow estimates
can only be acquired reliably where there can be an
assumption of normal arterial/capillary/venous hier-
archy, which is not the case in tumor tissue [10].
310
Most early antiangiogenic agents targeted
VEGF or VEGF signaling. Since VEGF produces
rapid and significant increases in the permeability
of the endothelial membrane, considerable work
was directed toward the quantification of perme-
ability by measurement of MR contrast agent
leakage. Analysis of DSCE-MRI using a
two-compartment leakage model was described
in an attempt to estimate permeability surface
area product of the local tumor capillary bed.
These pharmacokinetic analysis techniques
allow estimation of the contrast transfer coeffi-
cient (Ktrans) between the plasma and extravascu-
lar extracellular space (EES). However, modeling
of leakage effects using T2* dynamic images is
challenging, and attention rapidly shifted to the
use of T1-WI dynamic acquisitions.
Dynamic Relaxivity Contrast-Enhanced
MRI (DRCE-MRI)
On T1-WI images, signal change observed in
response to intravascular and extravascular contrast
agent is additive, simplifying attempts to estimate
the magnitude of contrast agent leakage in compar-
ison to DSCE-MRI. Furthermore, the majority of
clinical MR scanning systems are capable of acquir-
ing DRCE-MRI data. DRCE-MRI data can be ana-
lyzed using simple semiquantitative metrics,
although in research applications these have com-
monly been avoided since intensity changes can
vary between systems due to differences in acqui-
sition sequence design, receiver gain settings, and
other technical variables. Early studies attempted to
minimize these sources of variation by the use of
simple, normalized, metrics such as the maximal
rate of change of signal intensity or the ratio of pre-
and postcontrast signal intensity. Subsequent
improvements in hardware and particularly in gra-
dient coil design have minimized variations in
observed signal change, and these together with
the use of normalized metrics have led to routine
clinical use of semiquantitative measurements in
prostate, breast, pancreatic, cervical, colonic, and
rectal cancers, bone sarcoma, as well as brain
tumors. The clinical preference for semiquantitative
metrics reflects the complexity of pharmacokinetic
A. Jackson et al.
approaches and the lack of available standardized
software for clinical application (Fig. 2) [11].
Despite this a desire for better standardization
across imaging systems and for greater physiolog-
ical specificity led to the development and wide-
spread application of pharmacokinetic analytical
approaches [12].
Pharmacokinetic analysis of DRCE-MRI data
requires transformation of signal change measure-
ments into estimates of contrast agent concentra-
tion to allow the measurement of CC-TCC (Fig. 3).
This can be performed assuming that the base-
line T1 values of the tissue are measured prior to
contrast agent administration. While accurate mea-
surement of T1 values can be time-consuming, the
use of multiple flip angle gradient echo sequences
produces adequate accuracy in acceptable acquisi-
tion times and has become routine. Analysis of the
CC-TCC can then be undertaken using a variety of
pharmacokinetic distribution models. All of these
models require provision of an explicit input func-
tion representing the arterial input to the tissue
(AIF). The accuracy of measurement of AIF has a
considerable effect on the accuracy of derived
values of parameters from the analysis, and many
authors have studied the optimal sampling and
modeling of the AIF. Where an AIF cannot be
measured with sufficient accuracy, then a
population-based AIF may be applied.
The most common pharmacokinetic analysis
approach is a modified version of the original
Toft’s pharmacokinetic model that produces esti-
mates of vascular fraction (vp), extravascular
extracellular space fraction (ve), and the transfer
contrast coefficient (Ktrans) [14]. The Ktrans param-
eter is widely used in DRCE-MRI studies, often as
a potential surrogate of capillary endothelial per-
meability. However, although Ktrans will be
affected by the permeability, it is a bulk transfer
coefficient, so that it is also affected by the surface
area of the capillary endothelial bed and by rate of
delivery. This means that in very low flow states,
particularly where leakage is high, the transfer coef-
ficient will depend almost entirely on flow, whereas
in high flow states with low permeability, it will
depend almost entirely on permeability. Nonethe-
less it rapidly became evident that the VEGF inhi-
bition was indeed associated with rapid and
15 Imaging Biomarkers of Angiogenesis and the Microvascular Environment in Cerebral Tumors
Day0
Day0
Day2
Day90
R10 Ktrans vp ve
Fig. 2 Axial view of central slices of 3D longitudinally
co-registered kinetic and semiquantitative parametric maps
obtained in a 26-year-old woman, who has Type II neurofi-
bromatosis with a progressive VS (arrow heads) and an
occipital located meningioma (arrow) undergoing treatment
with bevacizumab. Images show comparisons of pharmaco-
kinetic leak derived parameters Ktrans, vp, and ve with
semiquantitative parameters: (1) absolute signal enhance-
ment (SE1min), (2) relative signal enhancement (SErel) which
profound reductions in Ktrans [13]. Although the
modified Toft’s model has been widely used, a
variety of alternate pharmacokinetic models exist,
ranging from simple two-compartment models,
such as the original Toft’s model, to more complex
models, such as the adiabatic tissue homogeneity
model, where more specific physiological informa-
tion is sought by definition of additional biological
variables (Table 1).
The choice of appropriate pharmacokinetic
models depends on a wide range of factors. In
some instances the demands of an individual
study may require absolute estimation of physio-
logical variables such as regional capillary blood
flow (F) which can be provided only by more
complex models. In contrast, the use of increas-
ingly complex models introduces uncertainty into
parameter estimates due to the need to optimize to
an increasing number of variables. This in turn
places significant demands on the image acquisi-
tion strategy requiring high temporal resolution
311
High
low
SE1min SErel ∑SE ∑SErel Rse1/se2
uses a baseline value for normalization in order to reduce
the dependence on biological and imaging system
variables, (3) the sum of SE over a fixed post-injection
duration (SE), (4) the sum of SErel over a fixed post-injection
duration (SErel), and (5) signal enhancement ratio,
commonly defined as the ratio of early
to late contrast

enhancement ratio, e.g., Rse1=se2 ¼ SI1min post  SIpre =


SI5min post  SIpre (From Ref. [11] with permission)
data with high signal-to-noise characteristics. The
need to “square the circle” by producing very high
spatial resolution data with high temporal resolu-
tion measurement of the AIF and high signal-to-
noise characteristics has led to extensive research
into MR image acquisition leading to a number of
novel fast image acquisition approaches for
DCE-MRI and the development of dual injection
methods where separate dynamic acquisitions are
acquired to provide matched high spatial and high
temporal resolution data.
Dynamic Contrast-Enhanced
Computed Tomography (DCE-CT)
With the development of rapid multi-slice CT
acquisitions, it was natural that the analytical
approaches taken with DCE-MRI would be
applied to CT data. Dynamic CT (DCT) has a
number of potential advantages over MRI. The
312
Fig. 3 DCE-MRI data acquisition and analysis. (a) Multi-
ple images (typically 25–100) are acquired as a bolus of
contrast agent passes through tissue capillaries; (b) the
region of interest for a tumor and the feeding vessel arterial
input function are defined; and (c) signal intensity values for
each voxel are converted into contrast agent concentration
using a map of T1 values. These steps allow calculation of
(d) WTV and T1 values. Next, voxels are classified as
enhancing or not after which parameters based on the
amount and proportion of enhancement can be defined,
along with the IAUC60. Finally, a pharmacokinetic model
may be applied to derive parameters such as Ktrans.
main one of these is that the concentration of
contrast agent is directly linearly related to the
measured attenuation value; this makes measure-
ments of semiquantitative parameters and the
application pharmacokinetic models more simple
and alleviates many of the problems associated
with multicentered studies in MRI [16]. The main
problem with DCT is of course the radiation dose,
which is significant and limits the application of
the technique into clinical trials requiring repeated
imaging measurements.
A. Jackson et al.
R
Abbreviations: [CA]dt area under the contrast agent–time
curve, AIF arterial input function, Cp contrast agent concen-
tration in plasma, CA contrast agent, DCE-MRI dynamic
contrast-enhanced MRI, EF enhancing fraction, ETV
enhancing tumor volume, IAUC60 initial area under con-
centration agent–time curve at 60 s, Ktrans volume transfer
constant between plasma and the extracellular extravascular
leakage space, ROI region of interest, S signal, T1 longitu-
dinal relaxation time, ve volume of extracellular extravas-
cular leakage space, vp blood plasma volume, WTV whole
tumor volume (From Ref. [13] with permission)
Imaging Microvessel Morphology
and Hypoxia with MRI
Vessel Size Imaging (VSI) and Vessel
Architectural Imaging (VAI)
As described above, a great deal of the develop-
ment of dynamic contrast-enhanced techniques
arose from a desire to study changes in regional
blood volume or capillary endothelial permeabil-
ity. An important feature of antiangiogenic
15 Imaging Biomarkers of Angiogenesis and the Microvascular Environment in Cerebral Tumors
Table 1 Parameters derived in DCE-MRI
Parameter Definition
Primary end pointsa
Ktrans Volume transfer constant between
plasma and the EES
IAUC60 Initial area under concentration agent-
time curve at 60 s
Alternative functional biomarkersb
kep Rate constant between EES and plasma
Ve Volume of EES per unit volume of tissue
Vp Blood plasma volume
F Blood flow
PS Permeability surface area product per
unit mass of tissue
Alternative biophysical measurementsc
WTV Whole tumor volume
ETV Enhancing tumor volume
EF Enhancing fraction (ETV:WTV ratio)
T1 Longitudinal relaxation time
DCE-MRI dynamic contrast-enhanced MRI, EES extracellular extravascular leakage space, NA not applicable
a
As recommended by Leach et al. [15]
b
Require pharmacokinetic modeling
c
Do not require pharmacokinetic modeling
therapy includes reduction in the abnormal
tumoral vasculature to produce the so-called “nor-
malization” of vascular structures. This normali-
zation process is characterized by a reduced
number of vessels of increasingly normal caliber
and reestablishment of the arterial/capillaries/
venous hierarchy [17]. As discussed above the
signal changes resulting from T1 and T2* relaxa-
tion measurements differ significantly. One prin-
cipal difference is that the relaxivity effect, which
produces the reduction in T1, is limited to the
immediate spatial environment of the proton,
while the variation in local field intensity which
results in T2* contrast extends a significant dis-
tance beyond the proton, affecting not only the
vessels containing contrast but also the surround-
ing tissue. As a consequence of this difference in
the contrast mechanisms, it is possible to model
the effect of vessel size and vessel distribution
and, using paired T1 and T2*WI acquisitions, to
derive a calculated parameter which reflects aver-
age vessel size (Rv) [18]. This technique is
referred to as vessel size imaging (VSI). VSI is
increasingly employed in research settings and
313
Unit Notes
min1 Composite measure of permeability, capillary
surface are and flow
mmol min Similar measure to Ktrans, but also influences
by Ve
min1 NA
NA NA
NA Relatively poor reproducibility
ml/ Temporal resolution achieved in most studies s
g min1 too slow
ml/ Temporal resolution achieved in most studies is
g min1 too slow
mm3 Easy to measure
mm3 Easy to measure
none Easy to measure
ms Easy to measure
has been used, in drug trials, to demonstrate vessel
normalization in response to antiangiogenic ther-
apies (Fig. 4) [20].
More recently this approach has been extended
following the realization that the difference in
contrast mechanisms between T1 and T2*
dynamic acquisitions produces not only the dif-
ference in signal intensity but also an apparent
temporal variation between the maximal effect of
T1- and T2*-based signal changes [21]. Modeling
of the temporal discrepancies of these signal
changes allows estimation of the homogeneity
and direction of flow. Furthermore the signals
are differentially affected by oxygen tension so
that an estimate of regional oxygen extraction can
also be calculated. This technique, known as ves-
sel architectural imaging (VAI), has also been
applied in studies of antiangiogenic therapies
showing different patterns of change in the early
phases of treatment in patients who respond to
therapy. VAI has also been used to study patients
with glioblastoma treated with cediranib and
chemoradiation, demonstrating that increases in
perfusion and oxygenation, seen in approximately
314 A. Jackson et al.

Fig. 4 Reversibility of Drug Holiday

normalization: (a) Vascular
and volume changes as a Post-contrast T1-weighted MRI
function of time in a patient
who did not take drug from
days 43 to 56 and was
imaged on day 55 (shown as
drug holiday). T1-WI
anatomic images after
intravenous administration
of gadolinium-DTPA,
similar to Fig. 2. Note that at
day 55, there is a rebound in
tumor enhancement, which
decreases again after
restarting the drug as seen
on follow-up imaging on
day 110. In this patient,
maps of relative vessel size
(similar to Fig. 1c) also Relative Vessel Size
show fluctuation with the
drug holiday and −1 1 28 55 111
resumption of AZD2171 Study Days
treatment. (b)
Measurements of imaging Drug Holiday
parameters confirm the 120 Day 43 to 55
Value relative to pretreatment (%)

reversibility of vascular
normalization by drug
interruption followed by VS
renormalization after
80
AZD2171 is resumed
(From Ref. [19] with
P
permission)

40
FLAIR

CE-T1
0
−1 1 28 55 111
Study Days

50 % of patients, are associated with improved
survival. In comparison an increase was seen in
only one of 14 patients with GBM treated with
chemotherapy and radiation alone [21].
Capillary Heterogeneity Imaging
Extensions of current pharmacokinetic modeling
approaches have also been developed for DRCE-
MRI. St Lawrence et al. extended their original
adiabatic tissue heterogeneity model to provide
spatial maps of tumor blood flow (F), extraction
fraction (E), permeability–surface area product
(PS), transfer constant (Ktrans), washout rate
(kep), interstitial volume (Ve), blood volume
(Vb), capillary transit time (tc), and capillary het-
erogeneity (a1) [22]. This has been termed the
gamma capillary transit time (GCTT) model, and
the introduction of the capillary transit time and
heterogeneity terms represents novel biomarkers
that could provide new evidence about microvas-
cular structure. Jensen et al. [23] applied the
model in a group of patients with glioblastoma
and found close correlation between the capillary
transit time, HIF-1, and VEGF expression in
active tumor. Furthermore, patient survival corre-
lated with only two DCE parameters: capillary
15 Imaging Biomarkers of Angiogenesis and the Microvascular Environment in Cerebral Tumors
Fig. 5 Example of Avol in a representative case of
untreated GBM. (Top) Row 1 shows a series of T1-WI
postcontrast images, and row 2 shows the same series with
the arterial (red) and venous (blue) ICA components
heterogeneity in active tumor and interstitial vol-
ume in areas of peri-tumoral edema.
Arteriovenous Overlap Imaging (AVOL)
La Violette et al. [24] studied a group of 43 patients
with high-grade glioma using independent compo-
nent analysis to separate the vasculature into arte-
rial and venous components. They observed that
the abnormal vascular flow patterns in tumors gave
rise to a number of voxels that were classified both
as arterial and venous (Fig. 5).
315
overlaid. The overlap of the two is indicated in yellow.
(Bottom) Mean MR signal within voxels defined by ICA
components and the overlap (From Ref. [24] with
permission)
They used this arteriovenous overlap (AVOL)
as an index of abnormal tumor microvasculature,
showing that median survival was greater in
patients where the Avol decreased in response to
bevacizumab treatment. They propose the Avol as
an independent biomarker of abnormal microvas-
cular hemodynamics. In a subsequent study the
same authors demonstrated significant sensitivity
of the Avol measurement to contrast leakage and
modify their recommendations to suggest that the
measurement should be performed on a second
dose of contrast following a preloading dose to
saturate the T1 effects of contrast leakage [25].
316
Imaging Hypoxia Using MRI
Hypoxia plays a central role in tumor development,
angiogenesis, tumor growth, and resistance to treat-
ment. Hypoxia is intimately related to the structure
and function of the vascular microenvironment.
Inadequate supply of oxygen and other nutrients
to tumor cells and to tumor support and inflamma-
tory cells results in local hypoxia and which forms
one of the main pro-angiogenic stimuli. Likewise,
failure of angiogenesis to keep pace with the rate of
tumor proliferation or metabolism will give rise to
areas of hypoxic tissue within the tumor. The pres-
ence of hypoxia has a profound effect on tumor
biology and has been shown to correlate with the
probability of metastatic spread, tumor recurrence,
resistance to chemotherapy and radiotherapy, inva-
sive phenotype, and decreased patient survival. It is
clear that prolonged or severe hypoxia induces
phenotypic changes such as loss of genotype sta-
bility, loss of apoptotic potential, and oncogene
activation. Imaging methods for the identification
of hypoxic tissue have classically relied on the use
of positron emission tomography (vide infra).
However, there is increasing interest in the use of
MR-based techniques to identify areas of hypoxic
tissue. As described in the previous section, VAI
has been shown to allow measurement of regional
oxygen consumption although the biomarker has
not yet been fully biologically qualified. Similarly,
capillary heterogeneity imaging demonstrates
close correlation with the expression of HIF-1.
Over the past several years, susceptibility-
weighted imaging (SWI) has become part of rou-
tine neuroimaging. SWI is used to detect magnetic
susceptibility of different tissues and is now clin-
ically used to assess neoplastic microvascular,
microbleeds and necrotic areas. One potential
cause of increased signal on SWI is the presence
of increasing concentrations of deoxyhemoglobin
providing another potential IB for hypoxic tissue.
In a recent study, 64 patients with brain tumors
underwent SWI demonstrating that R2* values
were significantly different between low-grade
and high-grade glioma. The authors suggest that
the R2* value may reflect alterations in deoxygen-
ated hemoglobin which are higher in high-grade
than low-grade glioma [26] (Fig. 6).
A. Jackson et al.
However, although these signal changes may
have some diagnostic value, they have not been
biologically qualified as an IB of tissue hypoxia,
and it is clear that the signal would also be affected
by other causes of susceptibility contrast such as
calcification or microhemorrhage.
Attempts to qualify the presence of hypoxia
using T1-WI oxygen-enhanced magnetic reso-
nance imaging (OE-MRI) measurements have
recently been described [27]. These and previous
works have demonstrated that increasing the con-
centration of inspired oxygen produces a signal
detectable in blood and tissues on T1-WI imaging.
Animal models of the area under the curve
resulting from oxygen enhancement showed pos-
itive correlation with histologically demonstrated
hypoxic tissue and a negative correlation with
vessel density (Fig. 7).
These relationships were not seen with param-
eters derived from DCE-MRI measurements.
Clinical data confirm that the same changes were
seen in patients with glioblastoma apparently
identifying central areas of hypoxic tissue. The
use of T1-WI dynamic imaging together with
oxygen enhancement appears highly promising
although further biological and technical valida-
tion is still required .
Molecular Imaging of the Tumor
Vascular Microenvironment
Understanding of the tumor vascular microenvi-
ronment and of the angiogenic process continues
to increase at an almost exponential rate. Enor-
mous investment leads to the ongoing identifica-
tion of numerous novel therapeutic targets and
candidate compounds which will pass into pre-
clinical and, eventually, for the more successful
agents, into clinical trial. Although dynamic
contrast-enhanced imaging methods have pro-
vided, and continue to provide, the mainstay of
imaging biomarker support for such studies, they
describe only selected elements of the tumor
microvascular environment. There is a growing
demand for biomarkers that can quantify specific
molecular processes to provide a more compre-
hensive toolkit for the investigation of the TVM
15 Imaging Biomarkers of Angiogenesis and the Microvascular Environment in Cerebral Tumors 317

a
a b c

b 40.00 a

62

30.00
R2* values

20.00

10.00

0.00

low-grade high-grade meningiomas brain

gliomas gliomas metastasis
brain tumors

6.00 b

5.00

4.00
23
CBF

3.00
38
2.00
13
1.00

0.00
Low grade High grade Meningiomas Metastasis
gliomas gliomas
Brain tumors

Fig. 6 (a) A 46-year-old man with a histologically proven
glioblastoma multiforme (grade IV) located in the right
frontal lobe. (a) Axial T1-WI postcontrast image demon-
strates nodular enhancement in the tumor. (b) The R2* map
shows an area of high R2* values (31.22) in the tumor.
There is hemorrhage at the bottom of the tumor. (c) The
ASL CBF map demonstrates nodular hyperperfusion. b R2*
and rCBF values of four types of brain tumors. (a) The
decreasing order of R2* values is high-grade gliomas,
brain metastasis, meningiomas, and low-grade gliomas. (b)
The sequence rCBF values are meningiomas, high-grade
gliomas, brain metastasis, and low-grade gliomas (From
Ref. [26] with permission)
318
Fig. 7 (a) Tumors from five patients with GBM. Top row:
area under the DR1 curve for OE-MRI, AUCOE displayed
using a hot color scale for positive values and negative
values using a cool color scale. All images windowed to
maximize visible dynamic range symmetrically around
zero AUCOE change. Bottom row: area under the DR1
curve for DCE-MRI, IAUCGd displayed using a hot color
and of angiogenesis, antiangiogenic, and
antivascular therapies. This has led to the devel-
opment of a number of novel radiotracers, princi-
pally designed for use with positron emission
tomography (PET), which target molecules
expressed by vascular and vascular support cells.
A. Jackson et al.
scale with all images windowed to maximize visible
dynamic range. Numerals indicate patient tumor index.
(b) Correlation between negative AUCOE areas following
oxygen inhalation with vessel density (circles) and per-
centage hypoxic area (crosses) in all U87MG tumors
(From Ref. [27] with permission)
Many candidate tracers are currently in develop-
ment and have been used for preclinical studies.
For a review of the current state-of-the-art in this
field, the reader is directed toward a number of
major review articles. In many cases the focus of
these agents is on systemic tumors, and there has
15 Imaging Biomarkers of Angiogenesis and the Microvascular Environment in Cerebral Tumors
been little application in brain tumors in humans.
The main area of interest with regard to human
glioma has been around imaging of hypoxia and
integrin expression.
Imaging Hypoxia Using PET
PET imaging is an ideal modality for evaluating
hypoxia. The PET imaging agent fluoromiso-
nidazole (18F-FMISO) is an 18F fluorinated
derivative of misonidazole, an azomycin hypoxic
cell sensitizer. It binds covalently to intracellular
molecules at a rate inversely proportional to intra-
cellular O2 concentration. 18F-FMISO diffuses
freely across the blood–brain barrier and is
reduced at any hypoxic site where it is not
excreted or highly metabolized and its metabolites
have a rapid plasma clearance, so normalization
for delivery is not required. Quantitatively, the
degree of hypoxia can be defined by hypoxic
fraction or volume or can be expressed by its
severity, defined as the region with the lowest
oxygen concentration, and its relative level. 18F-
FMISO imaging studies cannot provide values of
regional oxygen concentration, but the maximum
tumor/blood ratio is a convenient surrogate for the
worst level of hypoxia in the image. 18F-FMISO
is currently the most widely used positron emis-
sion tomography (PET) imaging agent for hyp-
oxia. The imaging procedure is a well-tolerated
procedure by patients and imaging takes around
20–30 min, starting from 75 to 150 min after
injection.
18F-FMISO has been validated in several ani-
mal models and human disease conditions, and its
signal is independent of other factors associated
with hypoxia, such as regional glucose concentra-
tion, glutathione levels, other transporters,
and pH.
In 3 studies of 34 patients with glioma, only
glioblastoma demonstrated constant uptake of
18F-FMISO, whereas neither low-grade glioma
nor anaplastic astrocytoma showed this effect. A
recent study in GBM patients, imaged preopera-
tively using 18F-FMISO, showed positive corre-
lations between relative hypoxia (defined as the
319
ratio of the hypoxic volume to the MRI
T2-defined tumor volume) and the net rate of
cell proliferation, as well as between the biologi-
cal aggressiveness ratio (defined as the ratio
between the net rate of cell proliferation and the
net rate of invasion) and relative hypoxia, scaled
to the blood activity of the tracer.
18F-FMISO and 15O-H2O PET have also
been used in brain tumors to provide matched
measures of tumor hypoxia and perfusion.
Increased 18F-FMISO brain tumor retention was
observed in GBM (seven out of seven) and was
associated with an increased 18F-FMISO tumor
distribution volume, which was also used as a
quantitative criterion for hypoxia. 18F-FMISO
accumulated in both hypo- and hyperperfused
tumor regions, suggesting that hypoxia in glio-
blastoma may develop irrespective of the magni-
tude of perfusion [28]. Other studies have
demonstrated correlations between 18F-FMISO
uptake and expression of vascular endothelial
growth factor (VEGF)-R1 and Ki67 expression
in glioblastomas and a relationship between the
uptake volume and intensity and both tumor pro-
gression and overall survival (Fig. 8).
Relatively poor tumor-to-background ratios
combined with slow clearance from normal tissue
have led to the ongoing search for improved hyp-
oxia imaging agents. For a fuller review of the
potential advantages and disadvantages of these
alternate imaging agents, the reader is directed to a
number of comprehensive review articles.
Imaging Integrin Expression
Integrins are membrane receptors comprised of an
α and a β subunit that mediate interactions
between cells. To date, 18 different α and 8 differ-
ent β subunits have been identified, forming
24 different integrin receptors [29]. A common
property of many integrins is their interaction with
the arginine-glycine-aspartic acid (RGD)
sequence found in extracellular matrix proteins
like vitronectin, fibrinogen, thrombospondin, and
fibronectin. The most extensively studied of these
in the angiogenic process is the integrin αvβ3
320
Fig. 8 (a–c) 15O-H2O PET perfusion images in three
patients with glioblastoma multiforme.
Corresponding late 18F-FMISO PET images show tumor
hypoxia in low perfusion, in intermediate perfusion with an
which is highly expressed on the surface of acti-
vated EC. However, integrins are commonly
expressed on the surface of tumor cells, and the
interpretation of the signal obtained from integrin
imaging can therefore be complex [30]. The first
integrin-specific PET tracer used in humans was
[18F]Galacto-RGD. The initial study of mixed
tumor types demonstrated wide variability of
tumor uptake with high background in the kid-
neys, liver, spleen, and bowel. Immunohis-
tological examination of resected tissue
demonstrated that the number of αvβ3-positive
vessels per field of view correlated highly with
tumor uptake of [18F]Galacto-RGD [31]. A sim-
ilar correlation between uptake and immunohisto-
chemical staining was also demonstrated in a later,
larger study. In glial cell tumors there was moder-
ate but variable uptake, most marked in areas of
high proliferation or tumor infiltration. Immuno-
histochemical staining was positive both in EC
and in tumor cells, and tracer uptake correlated
A. Jackson et al.
inverse pattern compared with hypoxia, and in high perfu-
(d–f) sion. PET images are normalized to their own maximum
(From Ref. [28] with permission)
with immunohistochemical assays of αvβ3
integrin expression [32] (Fig. 9).
A number of modifications designed to
improve the pharmacokinetic properties of RGD
peptides have been described, including the
development of multimeric tracers designed to
more effectively identify tumors in areas of high
physiologic integrin expression. Direct compari-
son of monomeric, dimeric, and tetrameric cyclic
RGD tracers conjugated with DOTA and
radiolabeled with 111In in mice with SK-RC-52
renal cell carcinoma xenografts demonstrated
tumor uptake of the tetramer which exceeded
that of both the dimer and the monomer [34],
and a 64Cu-DOTA-RGD octamer demonstrated
higher uptake than tetramer versions in human
glioblastoma tumor grafts [35].
One significant alternate integrin-specific
tracer is [18F]-fluciclatide, which has been devel-
oped by GE Healthcare and is becoming widely
available. This commercially available tracer
15 Imaging Biomarkers of Angiogenesis and the Microvascular Environment in Cerebral Tumors 321

Fig. 9 Examples of
patients with glioblastoma
multiforme of the left
frontal lobe (a, b) and right
parieto-occipital lobe (c, d).
Note the intense peripheral
enhancement in the
gadolinium-DTPA-
enhanced MRI scans in both
tumors (a, c). However, the
tumor in (b) shows only
very faint tracer uptake in
the [18F]Galacto-RGD PET
(maximum standardized
uptake value [SUVmax],
1.2), whereas the tumor in
(d) demonstrates
substantially more intense
[18F]Galacto-RGD uptake
(SUVmax, 2.8) (From Ref.
[33] with permission)

shows different binding affinities with the highest

affinity for αvβ5 (IC50 0.1 nM) followed by
integrin αvβ3 (IC50 11.1 nM). A study in human
glioma heterografts treated with sunitinib demon-
strated a therapy-induced reduction in tumor
uptake over a 2-week period. This correlated
with a reduction in tumor MVD suggesting that
this agent is capable of demonstrating vascular
responses to transgenic inhibition prior to the
occurrence of significant volumetric changes [36].
The potential role of integrin imaging in the
clinical environment remains uncertain. Observa-
tions of associations between integrin expression
and tumor progression in melanoma and tumor
grade in sarcoma have suggested potential clinical
utility. Similarly, preclinical studies have demon-
strated that integrin αvβ3 has an important role in
promoting aggression and metastatic potential in
breast cancer and glioma. These observations sug-
gest specific potential clinical utility for integrin
imaging, particularly in the selection of patients
for antiangiogenic therapies.
Clinical Applications of Microvascular
Imaging
The use of IB microvascular structure and func-
tion has been the subject of extensive clinical
research, and some better-characterized IB are
now passing into routine clinical use. Since
much of the development of DCE-MRI tech-
niques occurred in the brain, a number of these
have now been shown to have actual or potential
neuro-oncological application.
Diagnosing Glioma Mimics
In a large series of 5,000 stereotactic biopsies,
only 64.4 % were glioma, and it remains true,
despite improvements in routine clinical imaging,
that these lesions cannot always be confidently
diagnosed. DCE-MRI techniques have been
shown to allow improved differentiation between
322
glial cell tumors and common diagnostic mimics.
Distinguishing solitary metastases from high-
grade glioma can be difficult; however, several
studies have shown that rCBV values are higher
in GBM particularly in the region surrounding the
malignant tumor [9]. This has been attributed to
the presence of mixed vasogenic and cytotoxic
edema due to tumor cell invasion around the mar-
gins of GBM. The diagnosis of cerebral lym-
phoma is also greatly strengthened by the low
rCBV values that these tumors exhibit, clearly
distinguishing them from GBM which shows
high rCBV despite very similar enhancement pat-
terns on contrast-enhanced static images
[37]. Another major differential diagnosis of
GBM is cerebral abscess. Abscess may have
very similar classical imaging appearances but is
of course an important urgent medical diagnosis.
Once again rCBV in the rim of an enhancing
access is found to be significantly lower than
that seen in HGG [38]. Although there is signifi-
cant overlap between these groups, diagnostic
sensitivity and specificity become very high
when combined with measurements of apparent
diffusion coefficient from diffusion-weighted
imaging, which demonstrates restricted diffusion
in the glial tumors [38].
Grading Glioma
Many studies have described variations in DCE-
MRI-derived biomarkers between different grades
of glioma [39]. To a large extent this is an inter-
esting biological validation of the imaging bio-
marker but represents little or no clinical benefit in
the management of the patient. In all cases tissue
will be obtained and tissue diagnosis will be
made, which remains, at the present time, the
gold standard diagnostic criteria [40]. Typically
studies show increases in angiogenic activity, pro-
portionate to increasing aggressiveness and grade
of the tumor. Multiple authors have reported
increases in rCBV related to grade, capable of
differentiating low-grade from high-grade glioma
[41]. Some authors have also demonstrated simi-
lar correlations between grade and vascular frac-
tion from DRCE-MRI although it is noteworthy
A. Jackson et al.
that differentiation between grade III and grade IV
glioma is unreliable with either technique.
Identification of areas of dedifferentiation in
low-grade tumors and of aggressive areas in
HGG, particularly grade III, avoids erroneous
undergrading of the tumor due to failure to sample
the most aggressive component. In many clinical
practices this has led to routine use of rCBV maps
to identify optimal sites for biopsy.
Identifying Transformation in
Low-Grade Glioma
Low-grade glioma has a propensity to undergo
malignant transformation, but the time scale is
highly variable between individuals. Monitoring
of patients with low-grade disease is aimed at
identifying transformation at an early stage and
has classically relied on the development of con-
trast enhancement within the tumor. Contrast
enhancement can result from contrast leakage
due to increased endothelial permeability or
from increased intravascular contrast due to
increased proportional blood volume. Measure-
ment of increased rCBV using the DSCE-MRI
has been shown to allow early detection of dedif-
ferentiation by 6–12 months compared with con-
ventional imaging and is now in common clinical
use [42].
Distinguishing Recurrent Tumor from
Treatment Effects
Following chemoradiotherapy there is a signifi-
cant incidence of apparent progression of the
tumor on conventional radiological criteria that
later resolves. This is referred to as pseudo-
progression and can cause significant problems
with management and treatment planning. The
pseudo-progression syndrome appears to result
from leakiness of endothelial membranes due to
vascular damage following chemoradiotherapy.
The incidence of pseudo-progression has been
reported as 9 % with radiation alone and 21–31 %
with concomitant temozolomide. Routine MRI
sequences are unable to distinguish between
15 Imaging Biomarkers of Angiogenesis and the Microvascular Environment in Cerebral Tumors
true progression and early or late related treatment
effects. Pseudo-progression is characterized on
DSCE-MRI by lower rCBV, lower relative peak
height of the enhancement curve, and higher
percentage signal recovery compared to recurrent
tumor [43]. However, although some investigators
have showed extremely high discriminative
power of rCBValone, others have failed to achieve
this despite the use of multimodality IB including
MR spectroscopy and ADC measurements [43].
In the late stages following chemoradio-
therapy, radiation necrosis may occur between
4 and 12 months after radiation. It is related to
the areas of highest radiation dose and also causes
significant problems with treatment planning. In
one study, 51 % of patients with radiological
evidence of glioma progression following
temozolomide were found to have radiation necro-
sis following surgical resection with no evidence of
recurrent tumor. Microvascular IB, particularly
CBV, show differences between radiation necrosis
and recurrent tumor, being higher in tumor, but
these are not sufficiently great to allow confident
diagnosis [44]. In one study sensitivity and speci-
ficity were 83.3 % and 100 %, respectively. More
recently Larsen et al. reported sensitivities and
specificities of 100 % in a small study of 14 patients
[45]; however, these levels of accuracy have not
been repeated in larger studies. Although DSCE-
MRI may help make the differential diagnosis, it is
clear that the microvascular imaging biomarkers
are not yet sufficiently discriminative to be used
alone.
Predicting Treatment Response
Many groups have demonstrated a relationship
between progression-free survival and/or overall
survival and IB describing the TVM. As early as
2006 it was shown that glioma patients with
higher proportions of high rCBV had reduced
prognosis. Furthermore these workers showed
that changes in tumor CBV during the early treat-
ment course also predicted for survival. Better
survival was predicted by a decrease in the frac-
tional low-CBV tumor volume at week 1 of RT
vs. before RT and by a decrease in the fractional
323
high-CBV tumor volume at week 3 vs. week
1 [46]. These findings have been confirmed by a
number of other similar studies. Furthermore,
rCBV following resection has also been shown
to be predictive of overall survival in patients with
cerebral metastases.
Summary
Despite an explosion of research concerning the
biology of the TVM, knowledge of the subject
continues to increase exponentially as the drive
for new anticancer therapeutics continues. The
availability of minimally invasive, accurate,
reproducible IB forms a major component of
that research drive. The past 20 years has seen
the development and evolution of sophisticated
image acquisition and analysis techniques aimed
at developing IB which address the most rele-
vant components of disease physiology and
which are truly able to provide important infor-
mation of direct relevance. A vast amount of
development continues to occur around dynamic
contrast-enhanced methodologies, which have
now become ubiquitous in drug discovery and
drug trials for a large range of antiangiogenic
and vascular targeting agents. Increasingly
we are seeing these biomarkers being used
in clinic although there remains a very signifi-
cant delay between the initial biological valida-
tions of novel IB in their eventual clinical
qualification.
Despite major strides there are enormous chal-
lenges ahead. Although the development of IB has
been almost exponential in nature, there are sig-
nificant and growing anxieties about the complex-
ity of utilizing IB across multiple centers and in
large-scale studies. A number of international ini-
tiatives are currently under way to try and stan-
dardize acquisition and analysis techniques and to
improve both technical and biological validation
and qualification strategies. This process of con-
solidation will continue in parallel with innova-
tion because, although IB are expensive and
challenging to utilize, they offer many unique
properties and unparalleled advantages for the
study of tumor biology.
324
References
1. Hanahan D, Weinberg RA (2000) The hallmarks of
cancer. Cell 100(1):57–70
2. Patan S (2004) Vasculogenesis and angiogenesis.
Cancer Treat Res 117:3–32
3. Carmeliet P, Jain RK (2011) Molecular mechanisms
and clinical applications of angiogenesis. Nature
473(7347):298–307
4. Cheng SY, Huang HJ, Nagane M, Ji XD, Wang D, Shih
CC et al (1996) Suppression of glioblastoma
angiogenicity and tumorigenicity by inhibition of
endogenous expression of vascular endothelial growth
factor. Proc Natl Acad Sci U S A 93(16):8502–8507
5. Chen HX, Cleck JN (2009) Adverse effects of antican-
cer agents that target the VEGF pathway. Nat Rev Clin
Oncol 6(8):465–477
6. Nico B, Benagiano V, Mangieri D, Maruotti N,
Vacca A, Ribatti D (2008) Evaluation of microvascular
density in tumors: pro and contra. Histol Histopathol
23(5):601–607
7. Yao WW, Zhang H, Ding B, Fu T, Jia H, Pang L
et al (2011) Rectal cancer: 3D dynamic contrast-
enhanced MRI; correlation with microvascular density
and clinicopathological features. Radiol Med
116(3):366–374
8. Jackson A, Kassner A, Annesley-Williams D, Reid H,
Zhu XP, Li KL (2002) Abnormalities in the
recirculation phase of contrast agent bolus passage in
cerebral gliomas: comparison with relative blood vol-
ume and tumor grade. AJNR Am J Neuroradiol
23(1):7–14
9. Cha S, Lupo JM, Chen MH, Lamborn KR, McDermott
MW, Berger MS et al (2007) Differentiation of glio-
blastoma multiforme and single brain metastasis by
peak height and percentage of signal intensity recovery
derived from dynamic susceptibility-weighted con-
trast-enhanced perfusion MR imaging. AJNR Am J
Neuroradiol 28(6):1078–1084
10. Thacker NA, Scott ML, Jackson A (2003) Can
dynamic susceptibility contrast magnetic resonance
imaging perfusion data be analyzed using a model
based on directional flow? J Magn Reson Imaging
17(2):241–255
11. Jackson A, Li KL, Zhu X (2014) Semi-quantitative
parameter analysis of DCE-MRI revisited: monte-
carlo simulation, clinical comparisons, and clinical val-
idation of measurement errors in patients with type
2 neurofibromatosis. PLoS One 9(3):e90300
12. Jackson A (2004) Analysis of dynamic contrast
enhanced MRI. Br J Radiol 77(Spec No 2):S154–S166
13. O’Connor JP, Jackson A, Parker GJ, Roberts C, Jayson
GC (2012) Dynamic contrast-enhanced MRI in clinical
trials of antivascular therapies. Nat Rev Clin Oncol
9(3):167–177
14. Tofts PS, Kermode AG (1991) Measurement of the
blood-brain barrier permeability and leakage space
using dynamic MR imaging. 1. Fundamental concepts.
Magn Reson Med 17(2):357–367
A. Jackson et al.
15. Leach MO, Brindle KM, Evelhoch JL, Griffiths JR,
Horsman MR, Jackson A et al (2005) The assessment
of antiangiogenic and antivascular therapies in early-
stage clinical trials using magnetic resonance imaging:
issues and recommendations. Br J Cancer
92(9):1599–1610
16. Naish JH, McGrath DM, Bains LJ, Passera K,
Roberts C, Watson Y et al (2011) Comparison of
dynamic contrast-enhanced MRI and dynamic
contrast-enhanced CT biomarkers in bladder cancer.
Magn Reson Med 66(1):219–226
17. Jain RK (2001) Normalizing tumor vasculature with
anti-angiogenic therapy: a new paradigm for combina-
tion therapy. Nat Med 7(9):987–989
18. Schmainda KM, Rand SD, Joseph AM, Lund R, Ward
BD, Pathak AP et al (2004) Characterization of a first-
pass gradient-echo spin-echo method to predict brain
tumor grade and angiogenesis. AJNR Am J
Neuroradiol 25(9):1524–1532
19. Batchelor TT, Sorensen AG, di Tomaso E, Zhang WT,
Duda DG, Cohen KS et al (2007) AZD2171, a
pan-VEGF receptor tyrosine kinase inhibitor, normal-
izes tumor vasculature and alleviates edema in glio-
blastoma patients. Cancer Cell 11(1):83–95
20. Sorensen AG, Batchelor TT, Zhang WT, Chen PJ,
Yeo P, Wang M et al (2009) A “vascular normalization
index” as potential mechanistic biomarker to predict
survival after a single dose of cediranib in recurrent
glioblastoma patients. Cancer Res 69(13):5296–5300
21. Emblem KE, Mouridsen K, Bjornerud A, Farrar CT,
Jennings D, Borra RJ et al (2013) Vessel architectural
imaging identifies cancer patient responders to anti-
angiogenic therapy. Nat Med 19(9):1178–1183
22. St Lawrence K, Verdecchia K, Elliott J, Tichauer K,
Diop M, Hoffman L et al (2013) Kinetic model opti-
mization for characterizing tumour physiology by
dynamic contrast-enhanced near-infrared spectros-
copy. Phys Med Biol 58(5):1591–1604
23. Jensen RL, Mumert ML, Gillespie DL, Kinney AY,
Schabel MC, Salzman KL (2014) Preoperative dynamic
contrast-enhanced MRI correlates with molecular
markers of hypoxia and vascularity in specific areas of
intratumoral microenvironment and is predictive of
patient outcome. Neuro Oncol 16(2):280–291
24. LaViolette PS, Cohen AD, Prah MA, Rand SD,
Connelly J, Malkin MG et al (2013) Vascular change
measured with independent component analysis of
dynamic susceptibility contrast MRI predicts
bevacizumab response in high-grade glioma. Neuro
Oncol 15(4):442–450
25. LaViolette PS, Daun MK, Paulson ES, Schmainda KM
(2014) Effect of contrast leakage on the detection of
abnormal brain tumor vasculature in high-grade gli-
oma. J Neurooncol 116(3):543–549
26. Liu Z, Liao H, Yin J, Li Y (2014) Using R2* values to
evaluate brain tumours on magnetic resonance imag-
ing: preliminary results. Eur Radiol 24(3):693–702
27. Linnik IV, Scott ML, Holliday KF, Woodhouse N,
Waterton JC, O’Connor JP et al (2013) Noninvasive
15 Imaging Biomarkers of Angiogenesis and the Microvascular Environment in Cerebral Tumors
tumor hypoxia measurement using magnetic resonance
imaging in murine U87 glioma xenografts and in
patients with glioblastoma. Magn Reson Med
28. Bruehlmeier M, Roelcke U, Schubiger PA, Ametamey
SM (2004) Assessment of hypoxia and perfusion in
human brain tumors using PET with
18F-fluoromisonidazole and 15O-H2O. J Nucl Med
45(11):1851–1859
29. Takada Y, Ye X, Simon S (2007) The integrins.
Genome Biol 8(5):215
30. Desgrosellier JS, Cheresh DA (2010) Integrins in can-
cer: biological implications and therapeutic opportuni-
ties. Nat Rev Cancer 10(1):9–22
31. Haubner R, Kuhnast B, Mang C, Weber WA, Kessler H,
Wester HJ et al (2004) [18F]Galacto-RGD: synthesis,
radiolabeling, metabolic stability, and radiation dose
estimates. Bioconjug Chem 15(1):61–69
32. Schittenhelm J, Schwab EI, Sperveslage J, Tatagiba M,
Meyermann R, Fend F et al (2013) Longitudinal
expression analysis of alpha v integrins in human gli-
omas reveals upregulation of integrin alpha v beta 3 as
a negative prognostic factor. J Neuropathol Exp Neurol
72(3):194–210
33. Schnell O, Krebs B, Carlsen J, Miederer I, Goetz C,
Goldbrunner RH et al (2009) Imaging of integrin alpha
(v)beta(3) expression in patients with malignant glioma
by [18F] Galacto-RGD positron emission tomography.
Neuro Oncol 11(6):861–870
34. Liu S, Hsieh WY, Jiang Y, Kim YS, Sreerama SG,
Chen X et al (2007) Evaluation of a (99m)Tc-labeled
cyclic RGD tetramer for noninvasive imaging integrin
alpha(v)beta3-positive breast cancer. Bioconjug Chem
18(2):438–446
35. Li ZB, Cai W, Cao Q, Chen K, Wu Z, He L et al (2007)
(64)Cu-labeled tetrameric and octameric RGD pep-
tides for small-animal PET of tumor alpha(v)beta
(3) integrin expression. J Nucl Med 48(7):1162–1171
36. Battle MR, Goggi JL, Allen L, Barnett J, Morrison MS
(2011) Monitoring tumor response to antiangiogenic
sunitinib therapy with 18F-fluciclatide, an 18F-labeled
alphaVbeta3-integrin and alphaV beta5-integrin imag-
ing agent. J Nucl Med 52(3):424–430
37. Calli C, Kitis O, Yunten N, Yurtseven T, Islekel S,
Akalin T (2006) Perfusion and diffusion MR imaging
in enhancing malignant cerebral tumors. Eur J Radiol
58(3):394–403
325
38. Erdogan C, Hakyemez B, Yildirim N, Parlak M (2005)
Brain abscess and cystic brain tumor: discrimination
with dynamic susceptibility contrast perfusion-
weighted MRI. J Comput Assist Tomogr
29(5):663–667
39. Patankar TF, Haroon HA, Mills SJ, Baleriaux D, Buck-
ley DL, Parker GJ et al (2005) Is volume transfer
coefficient (K(trans)) related to histologic grade in
human gliomas? AJNR Am J Neuroradiol
26(10):2455–2465
40. Thompson G, Mills SJ, Stivaros SM, Jackson A (2010)
Imaging of brain tumors: perfusion/permeability. Neu-
roimaging Clin N Am 20(3):337–353
41. Yoon JH, Kim JH, Kang WJ, Sohn CH, Choi SH,
Yun TJ et al (2014) Grading of cerebral glioma
with multiparametric MR imaging and 18F-FDG-
PET: concordance and accuracy. Eur Radiol
24(2):380–389
42. Law M, Young RJ, Babb JS, Peccerelli N, Chheang S,
Gruber ML et al (2008) Gliomas: predicting time to
progression or survival with cerebral blood volume
measurements at dynamic susceptibility-weighted con-
trast-enhanced perfusion MR imaging. Radiology
247(2):490–498
43. Di Costanzo A, Scarabino T, Trojsi F, Popolizio T,
Bonavita S, de Cristofaro M (2014) Recurrent glioblas-
toma multiforme versus radiation injury: a
multiparametric 3-T MR approach. Radiol Med
119:616–624
44. Sugahara T, Korogi Y, Tomiguchi S, Shigematsu Y,
Ikushima I, Kira T et al (2000) Posttherapeutic
intraaxial brain tumor: the value of perfusion-sensitive
contrast-enhanced MR imaging for differentiating
tumor recurrence from nonneoplastic contrast-
enhancing tissue. AJNR Am J Neuroradiol
21(5):901–909
45. Larsen VA, Simonsen HJ, Law I, Larsson HB, Hansen
AE (2013) Evaluation of dynamic contrast-enhanced
T1-weighted perfusion MRI in the differentiation of
tumor recurrence from radiation necrosis. Neuroradi-
ology 55(3):361–369
46. Cao Y, Tsien CI, Nagesh V, Junck L, Ten Haken R,
Ross BD et al (2006) Survival prediction in high-grade
gliomas by MRI perfusion before and during early
stage of RT [corrected]. Int J Radiat Oncol Biol Phys
64(3):876–885
 Part IV
Stroke Imaging
 Patterns of Ischemic Stroke: From
Lacunar to Territorial to Multiple 16
Embolic to Watershed Hypotensive

Jessica Rotman and Robert Zimmerman

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329 Neuroimaging is integral to stroke diagnosis
and management. Ischemic stroke prognosis,
Etiologies of Ischemic Strokes . . . . . . . . . . . . . . . . . . . . . 330
risk of recurrence, and management decisions
Imaging Findings Associated with Ischemic are influenced by stroke subtype, which can be
Strokes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332 categorized by anatomic distribution and caus-
CT Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332
MRI Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337 ative mechanism. Computed tomography
(CT) and multimodal magnetic resonance
Vascular Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
imaging (MRI) facilitate the identification of
MR Perfusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341 the extent, etiology, and time of onset of ische-
Localization of Ischemic Strokes . . . . . . . . . . . . . . . . . . 344 mic strokes. Understanding the underlying
Anterior Territory Infarcts . . . . . . . . . . . . . . . . . . . . . . . . . . . 344 vascular anatomy and patterns of ischemic
Posterior Territory Infarcts . . . . . . . . . . . . . . . . . . . . . . . . 347 stroke enables physicians to localize diseased
Branch Occlusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347 blood vessels and better personalize patient
Basilar Occlusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351 management.
Watershed Infarcts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
Lacunar Infarcts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
Keywords
Ischemia • Lacunar infarct • Watershed infarct •
Anoxia and Hypoxia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
Anoxia • Hypoxia • Emboli • CT • MRI
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
Introduction

Stroke is a leading cause of mortality and morbid-

ity in the developed world. The goals of imaging
J. Rotman (*)
PGY-3 Diagnostic Radiology Resident, New York in the setting of acute stroke are to establish a
Presbyterian Hospital, Weill Cornell Medical Center, diagnosis as early as possible and to deduce accu-
New York, NY, USA rate information regarding the intracranial vascu-
e-mail: jessicarotman@gmail.com; jar9120@nyp.org
lature and brain perfusion to guide in the selection
R. Zimmerman of appropriate therapy.
Diagnostic Radiology Department, New York Presbyterian
The ability to determine the age of an
Hospital, Weill Cornell Medical Center, New York, NY,
USA ischemic stroke provides useful clinical
e-mail: rdzimmer@med.cornell.edu

# Springer Science+Business Media New York 2016 329

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_27
330
information for the patient, his or her family, and
the medical team. The time window to administer
intravenous tissue plasminogen activator is
currently 4.5–6 h from the time when the patient
was last seen to be normal [1, 2]. Different
imaging characteristics associated with different
stages of infarction significantly affect the deci-
sion of whether or not to administer thrombolytic
therapy.
The various MRI patterns of acute ischemic
stroke, including topography, size, and multiplic-
ity, are essential factors that can reflect the most
likely mechanisms of origin. Leptomeningeal
arteries comprise the terminal branches of the
cerebral and cerebellar arteries, which penetrate
the cortex and subjacent white matter. Infarcts
involving the leptomeningeal artery system are
often described as territorial infarcts. Ischemic
infarcts secondary to hypoperfusion that occur at
the border zones between major arterial territories
are known as watershed infarcts, while deep cere-
bral infarcts typically caused by hypertension are
referred to as lacunar infarcts. In 1991, the
Oxfordshire Community Stroke Project (OCSP)
proposed four subgroups of cerebral infarctions
based on presenting signs and symptoms
(Table 1). Patterns of brain lesions on MRI
according to a modified Oxfordshire method are
classified based on the anatomic distribution of
infarcts: (1) total anterior circulation infarcts,
(2) partial anterior circulation infarcts, (3) poste-
rior circulation infarcts, (4) watershed infarcts,
(5) centrum ovale infarcts, and (6) lacunar
infarcts [3].
In 1993, a multicenter clinical trial known as
TOAST, or Trial of ORG 10172 in Acute Stroke
Treatment, classified ischemic strokes based on
their respective causative mechanisms: (1) large-
vessel atherosclerotic disease, (2) small-vessel
atherosclerotic disease, (3) cardioembolism,
(4) other determined etiologies, and
(5) undetermined of multiple possible etiologies
(Table 2) [4, 5].
In this chapter, the etiologies of ischemic
stroke, characteristic imaging patterns associated
with the evolution of ischemic strokes, and stroke
categories based on causative mechanisms will be
reviewed.
J. Rotman and R. Zimmerman
Table 1 Topographic clinical pattern of brain infarction
according to modified Oxfordshire method [3]
Type of infarct Clinical description
Lacunar Acute stroke that includes one of the
major recognized lacunar
syndromes: (1) pure motor, (2) pure
sensory, (3) sensorimotor, (4) ataxic
hemiparesis, (5) dysarthria (clumsy
hand syndrome)
Total anterior Ischemia in both deep and superficial
circulation territories of the middle cerebral
artery characterized by higher
cerebral dysfunctions such as
dysphasia, dyscalculia, visuospatial
disorder, homonymous visual field
defects, ipsilateral motor, and/or
sensory deficit of at least two areas of
the face, arm, and leg
Partial anterior Clinical syndrome includes only two
circulation of the three aforementioned
components, with higher cerebral
dysfunction alone or with more
restricted sensorimotor deficits than
those classified as lacunar infarcts
Posterior Ipsilateral cranial nerve palsy with
circulation contralateral motor and/or sensory
deficit, bilateral motor and/or
sensory deficit, disorder of conjugate
eye movement, cerebellar
dysfunction without ipsilateral long-
tract deficit, or isolated
homonymous visual defect
Adapted with permission from Lippincott Williams and
Wilkins/Wolters Kluwer Health: Stroke, Adams HP Jr,
Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon
DL, Marsh EE 3rd, Classification of subtype of acute
ischemic stroke. Definitions for use in a multicenter clinical
trial. TOAST. Trial of Org 10172 in Acute Stroke Treat-
ment [4], 1993
Etiologies of Ischemic Strokes
Large-vessel disease is the culprit in approxi-
mately 13 % of all patients who present with
first-time strokes [6]. Caused by atherosclerotic
debris and ulceration, large-vessel disease is
defined as occlusion or stenosis of greater than
50 % of a major brain artery or branch cortical
artery. Large-vessel atherosclerosis leads to
infarction when there is hemodynamically signif-
icant narrowing, which is defined as a decrease in
vessel lumen diameter by 50–70 % [7]. In such
cases, however, it is also possible for blood flow to
16 Patterns of Ischemic Stroke: From Lacunar to Territorial to Multiple Embolic to Watershed. . .
Table 2 Stroke categories according to TOAST method
[4, 5]
1 Large-vessel disease
2 Small-vessel disease
3 Cardioembolism
4 Other etiology
5 Undetermined or multiple possible etiologies
TOAST trial of ORG 10172 in acute stroke treatment
be preserved in the setting of collateral circulation
involving the circle of Willis and leptomeningeal
vessels. Conversely, complete internal carotid
artery (ICA) occlusion in the neck may still
cause cerebral infarctions in the setting of emboli,
which may be multiple and simultaneous or
focally broken up to produce multiple infarctions.
While 35 % of patients over the age of 50 have
atherosclerotic changes in their cervical arteries,
only one third will have symptomatic vascular
disease [7]. Dissection of the carotid or vertebral
artery should always be considered as a cause of
stroke in young otherwise healthy patients. Ische-
mic infarction typically occurs as a result of distal
embolization from a cervical dissection rather
than occlusion and typically manifests on imaging
several hours to days after the onset of the dissec-
tion, although may also be delayed for several
weeks.
Cardioembolism is the principal cause of
ischemic cerebrovascular disease. Cardiogenic
embolism is responsible for approximately
15–27 % of first-time strokes [6]. High and
medium risk sources of cardioembolism are sum-
marized in Table 3. In those with large infarcts
without ICA occlusion, the frequency of
cardioembolic disease is higher than in large
infarcts with ICA occlusion or limited infarcts
without ICA occlusion [8]. The incidence of
cardioembolism is greater in patients below the
age of 45, presumably due to a lower rate of
atherosclerotic disease in this younger age
group. Cardioembolic ischemic stroke is the
most common subtype associated with recurrent
strokes and is associated with the highest 1-month
mortality and in-hospital mortality in comparison
to other subtypes [9, 10].
Approximately 16 % of ischemic strokes are
associated with atrial fibrillation and 10 % are
331
Table 3 Sources of cardioembolism stratified by risk
[7, 9–12, 67]
High risk Mechanical prosthetic valve
Mitral stenosis with atrial fibrillation
Atrial fibrillation
Left atrial/atrial appendage thrombus
Sick sinus syndrome
Recent myocardial infarction (<4 weeks)
Left ventricular thrombus
Dilated cardiomyopathy
Akinetic left ventricular segment
Atrial myxoma
Infective endocarditis
Medium Mitral valve prolapse
risk Mitral annulus calcification
Mitral stenosis without atrial fibrillation
Left atrial turbulence
Atrial septal aneurysm
Patent foramen ovale
Atrial flutter
Lone atrial fibrillation
Bioprosthetic cardiac valve
Nonbacterial thrombotic endocarditis
Congestive heart failure
Hypokinetic left ventricular segment
Myocardial infarction (>4 weeks, <6
months)
Rovira [7] with kind permission from Springer Science and
Business Media
probably secondary to atrial appendage embolism
[11, 12]. On brain imaging, large cortical territo-
rial infarction is suggestive of a cardioembolic
mechanism, especially if the patient does not
have ICA occlusion. Although simultaneous
acute infarctions are generally attributed to
cardioembolism, the majority of cases are gener-
ally not caused by a proximal embolus from the
heart or aortic arch, but rather by artery-to-artery
embolism or fragmentation of a proximal artery
embolus [7]. The median volume of infarcts sec-
ondary to cardiogenic embolism is greater than
twice that of infarcts caused by artery-to-artery
embolism [13].
Although uncommon, acute stroke caused by
fat emboli is potentially life threatening. Acute
ischemia in the presence of long bone or pelvic
fractures, cardiac surgery, or joint replacement
therapy should raise concern for fat embolism.
332
Fat emboli from epicardial or bone marrow fat can
pass through pulmonary capillaries without
shunting lesions and result in systemic emboliza-
tion, commonly to the brain. Patients typically
present with the clinical triad of hypoxia, altered
mental status, and petechiae. Other signs that may
be present include tachycardia, fever, anemia, and
thrombocytopenia.
Small-vessel occlusion is characterized by one
of the traditional lacunar syndromes without evi-
dence of cerebral or cortical dysfunction and is the
cause of 25 % of all first-ever strokes [6]. Lacunes
are thought to occur most often in patients with
hypertension or diabetes [14]. In addition to athero-
sclerosis and lipohyalinosis, other potential causes
of small-vessel occlusion include vasculopathy,
artery-to-artery or cardiogenic embolism, arterial
dissection, and hypercoagulability-associated
thrombosis [15]. In patients with their first lacunar
stroke, vascular imaging of the cervical and intra-
cranial arteries for a cardiac source of embolism is
indicated.
Only 2 % of strokes are classified in the “other
etiology” category [6]. Strokes of other deter-
mined causes include non-atherosclerotic
vasculopathies, hypercoagulable states, and
hematologic disorders, whereas strokes with neg-
ative or incomplete evaluations, or as a result of
two or more causes, are labeled strokes of
undetermined cause.
Anoxic and hypoxic brain injuries occur when
there is decreased oxygen content in the blood.
Anoxia results from near-complete absence of
oxygen in the blood for more than 5 min, while
hypoxia occurs when there is partial but more
prolonged hypoxemia [7]. Anoxia can occur
after cardiac arrest, prolonged seizures, strangu-
lation/hanging, near drowning, and smoke/car-
bon monoxide inhalation. Carbon monoxide
poisoning produces anoxic injury by preventing
oxygen from binding to hemoglobin. General
sites of the brain susceptible to ischemia are
those with increased metabolic rate and a paucity
of redundant blood supply, including the hippo-
campus, globus pallidus (in particular in carbon
monoxide poisoning), amygdala, cerebellum,
and occipital lobes, in decreasing order of
frequency.
J. Rotman and R. Zimmerman
Imaging Findings Associated
with Ischemic Strokes
Brain imaging can accurately localize an ischemic
stroke lesion to a specific vascular distribution
[16, 17]. Identification of occlusion or stenosis
of extracranial and/or intracranial arteries can
delineate the ischemic nature of a lesion and
help determine its etiology – for example, whether
it occurs as a consequence of slow flow, proximal
embolic occlusion, or small-vessel disease.
CT and MRI findings change rapidly through-
out the initial week after an infarct due to the
underlying pathophysiologic changes taking
place. Although CT is extremely useful for
detecting large ischemic infarcts of more than
6–8 h in duration, it does not reliably
reveal infarcts of less than 4 h in duration or
optimally characterize the extent of such
infarctions. The goals of CT in the acute setting
include exclusion of intracranial hemorrhage or
involvement of greater than 33 % of the distribu-
tion of the middle cerebral artery (MCA) territory,
which would preclude thrombolytic therapy, iden-
tification of any early features of infarction, and
exclusion of other intracranial pathologies that
may mimic a stroke, such as a tumor.
CT Findings
Within the first few hours of an ischemic stroke,
imaging features on CT include loss of
gray–white matter differentiation and loss of the
ability to differentiate the basal ganglia or thala-
mus from the internal capsule. Mass effect is
typically absent. Areas of cortex with poor collat-
eral blood supply, such as the insular ribbon
(Fig. 1) are more vulnerable to ischemia [18] and
should be carefully inspected to detect early loss
of gray–white matter differentiation. The basal
ganglia should also be scrutinized for findings
associated with hyperacute stroke; relative
hypodensity in the lentiform nucleus (with loss
of contrast difference to the adjacent white matter
in the internal capsule) can be seen in 75 % of
patients at 3 h [18] (Fig. 2). The dense vessel sign
reflective of acute embolic occlusion can
16 Patterns of Ischemic Stroke: From Lacunar to Territorial to Multiple Embolic to Watershed. . .
Fig. 1 Hyperacute M2/M3 MCA distribution infarct;
insular ribbon sign. Axial CT of the brain demonstrates
marked hyperdensity in the frontal and temporal lobes
along the insular ribbon and cortex, consistent with acute
MCA distribution infarction
sometimes be seen on unenhanced CT and is
helpful in the setting of normal-appearing acutely
infarcted parenchyma. Evaluation of CT angio-
gram source images may make acute infarction
more apparent because of absent or decreased
enhancement of hypoperfused gray matter
(Fig. 3).
Over the first week, hypoattenuation and
edema associated with ischemia progressively
increase. In the case of a large infarct, substantial
mass effect is typical, often causing secondary
damage to other areas of the brain, such as in
cases of herniation (Fig. 4). Contrast-enhanced
scans may demonstrate parenchymal enhance-
ment, which is typically gyriform in cases of
cortical infarction. Approximately one week
after the onset of cerebral infarction, the swelling
starts to subside and there are new cortical pete-
chial hemorrhages and/or hyperemia, which result
in elevation of cortical attenuation (pseudonorma-
lization). In such cases, imaging of a stroke at this
time can be misleading as the affected cortex will
appear normal, known as the “CT fogging
333
phenomenon.” Administration of contrast will
demonstrate gyriform enhancement of the
“fogged” infarct area and can thus be performed
to prevent misinterpretation [19] (Fig. 5). Over the
weeks and months that follow, infarction gliosis
sets in, which appears as a region of low density
with focal volume loss (atrophy), such as in the
case of ex vacuo dilatation of the ventricles and
associated encephalomalacia (Fig. 6). Cortical
mineralization or laminar necrosis may also
occur, appearing as curvilinear hyperdensities fol-
lowing gyral anatomy.
Perfusion CT is superior in detecting the pres-
ence and extent of possible acute strokes that are
inconspicuous on unenhanced CT. It has emerged
as a critical tool for the selection of patients opti-
mal for reperfusion therapy, as it can identify the
penumbra surrounding the core of an infarct,
which is the region that is ischemic but has not
yet undergone permanent infarction and may thus
be possibly salvaged. The parameters associated
with cerebral hemodynamics include cerebral
blood flow (CBF), which represents instantaneous
capillary flow in tissue; cerebral blood volume
(CBV), which describes the blood volume of the
cerebral capillaries and venules per cerebral tissue
volume; the mean transit time (MTT), which mea-
sures the length of time a certain volume of blood
spends in the cerebral capillary circulation; and
the time to peak (TTP), which is inversely related
to CBF in that reduction of blood flow results in an
increase in the time needed for contrast to reach its
peak in the perfused volume of brain tissue. These
parameters are linked by the central volume prin-
ciple which state that CBF = CBV/MTT. Specif-
ically, areas which demonstrate matched defects
in CBVor CBF and MTT represent unsalvageable
infarct core, whereas those with prolonged MTT
but preserved CBV are considered to be the ische-
mic penumbra [20]. Recently the value of perfu-
sion imaging in determining success of stroke
intervention has come into question [21], leading
some investigators to state that, given the impor-
tance of time to treatment, a non-contrast head CT
is all that is needed to make a therapeutic decision
[22] (Fig. 2).
CT angiography (CTA) is utilized to
identify possible thrombus within an intracranial
334
Fig. 2 Hyperacute MCA infarct. (a) Axial CT of the brain
demonstrates a dense distal right internal carotid artery and
M1 portion of the MCA, a sign of acute thrombus. Note the
normal less dense left MCA. (b) A more superior image
demonstrates loss of normal gray matter density in the
entire right MCA distribution. Note the relative
hypodensity of the left and right sides of the ganglia and
cortex. (c) The corresponding axial CTA of the brain dem-
onstrates a filling defect in the distal right internal carotid
vessel and guide intra-arterial thrombolysis or
clot retrieval. CTA can also be applied to the
carotid and vertebral arteries in the neck to
establish stroke etiology, such as atherosclerosis
J. Rotman and R. Zimmerman
artery and M1 segment of middle cerebral artery due to
acute thrombosis. Note the normally enhancing patent left
MCA and normal enhancement of right M2 branches is
consistent with collateral supply. (d) CT perfusion demon-
strates decreased CBF and CBV and increased MTT
corresponding to the subtle hypoattenuation seen on
non-contrast CT. There is no mismatch
or carotid dissection, and identify any
possible endovascular treatment access limita-
tions, such as arterial tortuosity or stenosis
(Figs. 2 and 3).
16 Patterns of Ischemic Stroke: From Lacunar to Territorial to Multiple Embolic to Watershed. . . 335

Fig. 3 MCA/ACA
distribution hyperacute
infarction. (a) Unenhanced
CT in a patient with atrial
fibrillation demonstrates
subtle loss of normal gray
matter density involving
lateral (MCA) and medial
(ACA) frontal cortex. (b)
On source images loss of
normal cortical gray density
is more apparent, reflecting
decreased blood volume in
the region of infarction.
Axial (c) and coronal (d)
MIP CTA images
demonstrate lack of
enhancement of the A1 and
M1 segments of the right
MCA, indicative of acute
thrombus and loss of
normal cortical density in
the MCA and ACA
distributions. Note the
presence of filling of some
right M2 and A2 branches
due to collateral supply
from the posterior
circulation

Fig. 4 Subacute “malignant” MCA infarct. (a) Axial CT
of the brain 3 days after an acute MCA territory stroke
demonstrates marked hypodensity and mass effect with
uncal herniation, as a result of increased vasogenic edema
and intracranial pressure. (b) Image at the level of the
lateral ventricle demonstrates marked hypodensity with
transfalcine herniation. Punctate foci of hyperdensity and
mild gyriform hyperdensity within the infarct are consis-
tent with a combination of reperfusion hemorrhage and
hyperemia
336 J. Rotman and R. Zimmerman

Fig. 5 Subacute PCA P3

distribution infarct. (a)
Axial unenhanced CT of the
brain in a patient with a
subacute PCA territory
occipital lobe stroke, which
appears isodense to normal
gray matter due to
hyperemia and/or laminar
necrosis. (b) Contrast-
enhanced axial CT of the
brain demonstrated
gyriform enhancement that
confirms that non-enhanced
CT of the brain
demonstrated the “fogging
phenomenon”

Fig. 6 Chronic M2/M3

distribution infarct with
cystic encephalomalacia.
(a) Axial CT scan
demonstrates marked
hypodensity in the left
MCA distribution with ex
vacuo dilatation of the left
lateral ventricle. (b)
T1-weighted and (c) FLAIR
images demonstrate that
this region is fluid-filled
consistent with cystic
encephalomalacia. (d)
T2-weighted image at the
level of the mesencephalon
demonstrates marked
atrophy of the cerebral
peduncle consistent with
Wallerian degeneration
16 Patterns of Ischemic Stroke: From Lacunar to Territorial to Multiple Embolic to Watershed. . .
Table 4 Stages of ischemic stroke and their respective appearances on MRI
Stage Timing T1WI
Hyperacute 0–6 h Isointense, possible loss of
sulci
Acute 6 h–4 Hypointense, mass effect
days
Subacute 4–14 Hypointense
days
Chronic >14 Smaller area of
days hypointensity,
encephalomalacia
MRI magnetic resonance imaging, DWI diffusion-weighted imaging, ADC apparent diffusion coefficient
MRI Findings
The acuity or chronicity of a stroke is better elu-
cidated on MRI than CT due to different time-
sensitive signal characteristics. The typical MRI
findings that may help determine the relative ages
of cerebral infarcts are displayed in Table 4. MRI
is a much more sensitive modality than CT in the
identification of hyperacute infarction (<6 h)
[23]. In the first 3 h after infarction, CT does not
typically demonstrate any parenchymal abnor-
malities [23, 24]. T2-weighted fluid attenuation
inversion recovery (FLAIR) sequences are also
relatively insensitive during this hyperacute
period (and are less sensitive than CT in the
3–6 h range) but the overall sensitivity of MR is
82 % in the first 24 h, whereas detection rates for
acute infarction on CT are 58 % in the first 24 h
[25]. Even more impressive is the 99 % sensitivity
rate of diffusion-weighted imaging (DWI), asso-
ciated with detection of an ischemic stroke within
12 h of its onset (including the first 1 h), in
contradistinction to the 39 % sensitivity rate asso-
ciated with CT [26]. In addition, causes of stroke
such as venous thrombosis, vascular
malformations, infections, and tumors are better
identified and characterized with greater accuracy
on MRI than is possible with CT.
Described as the “light bulb sign,”
hyperintensity associated with acute infarction
usually occurs within 3 h and is easily detected
on DWI [7] (Figs. 7 and 8). This marked
hyperintensity is initially due to diffusion restric-
tion in the area of cerebral ischemia. Cytotoxic
edema occurs as a result of depletion of adenosine
337
T2WI/FLAIR DWI ADC
Isointense to Hyperintense Hypointense
mildly
hyperintense
Hyperintense Hyperintense Hypointense
Hyperintense Mildly hyperintense Isointense
from T2 shine through
Hyperintense Isointense Hyperintense
triphosphate, which in turn leads to breakdown of
normal cell membrane integrity. As a conse-
quence, there is intracellular swelling and damage
to intracellular proteins causing increased intra-
cellular viscosity. These processes are indicators
of cell death. After 4–6 h T2 increases within the
area of infarction due to vasogenic edema.
Restricted diffusion can be assessed independent
of T2 by evaluating apparent diffusion coefficient
(ADC) maps. Restricted diffusion usually peaks at
around 3 days and disappears by 5–10 days
(pseudonormalization) [27, 28]. Thereafter diffu-
sion increases in the area of infarction due to
removal of dead cells and development of gliosis.
Diffusion restriction and increased T2 combine to
produce hyperintensity on DWI images that lasts
for up to 2 weeks after infarction. In circum-
stances where diffusion is equal to normal brain
but T2 is increased, such as in the case of a
subacute infarction, the tissue will appear bright
on DWI and isointense on apparent diffusion
coefficient (ADC) maps, a phenomenon known
as T2 shine through (Fig. 9).
DWI nearly doubles the likelihood of detecting
acute ischemic strokes compared to FLAIR
[29]. Although T2-weighted imaging and FLAIR
are less sensitive in detecting parenchymal change
in the first few hours, loss of normal flow voids in
large arteries may be visible immediately
[30]. The MRI correlate of the early CT dense
vessel sign in acute embolic occlusion is
intraluminal hypointensity without associated
flow-related hyperintensity on gradient echo or
susceptibility-weighted images. Normally, large
vessels are centrally hypointense with associated
338
Fig. 7 Hyperacute PCA P2 distribution infarct on MR
(<4 h); FLAIR/DWI mismatch. (a) Hyperintensity on
DWI and (b) relative isointensity on FLAIR in the occipital
lobe and hippocampus. The presence of the diffusion-
Fig. 8 Acute infarct internal capsule – anterior choroidal
distribution. (a) CT reveals ill-defined hypodensity in left
posterior limb of internal capsule consistent with infarct of
indeterminate age. (b) DWI reveals marked hyperintensity.
(c) ADC map reveals hypointensity indicative restricted
J. Rotman and R. Zimmerman
FLAIR mismatch has been proposed as an indicator of
hyperacute infarction when the clinical timing of infarction
is uncertain (e.g., “wake-up strokes”). In many of these
cases, CT will be positive before FLAIR
diffusion. (d) FLAIR reveals subtle hyperintensity indica-
tive of acute infarct. (e) DWI at the level of the cerebral
peduncle reveals hyperintensity indicative of acute
Wallerian degeneration
16 Patterns of Ischemic Stroke: From Lacunar to Territorial to Multiple Embolic to Watershed. . .
Fig. 9 Subacute thalamic infarct – T2 shine through. (a)
Axial DWI MRI of the brain demonstrates hyperintensity
in the medial thalamus imparting mild mass on the third
ventricle. (b) ADC map demonstrates absence of
hyperintensity adjacent to the vessel due to signal
from spatially displaced blood secondary to flow
effects. In the presence of acute clot, however,
there is marked hypointensity that often extends
beyond the lumen of the vessel, referred to as
“blooming,” and there is no corresponding flow-
related hyperintensity [7] (Fig. 10).
Hyperacute infarction is T1 isointense and T2
isointense to mildly hypertense, which is best
appreciated on FLAIR and is usually confined to
the gray matter in cases of thromboembolic infarc-
tion. It typically takes 4–5 h for hyperintensity to
develop on FLAIR, which is actually less sensi-
tive than CT in the 3–6 h period [7]. The presence
of restricted diffusion and normal signal on
FLAIR (diffusion-FLAIR mismatch) may poten-
tially be used to determine the onset of infarction
in patients in whom the time on onset is unknown
(referred to as “wake-up” stroke), potentially
allowing for therapeutic intervention in these indi-
viduals [31] (Fig. 7).
After 6–12 h, infarcted tissue becomes high
signal on T2-weighted images, while low inten-
sity on T1-weighted images typically mirrors this
high T2/FLAIR signal. In thromboembolic
infarcts, T2 hyperintensity is confined to the
gray matter. Although the extent and degree of
T2 hyperintensity increase during the acute phase
339
hypointensity indicative of T2 shine through and not
restricted diffusion. (c) FLAIR MRI demonstrates marked
hyperintensity corresponding to region of hyperintensity
on DWI sequence
of infarction, the extent of the DWI abnormality
remains relatively unchanged (unless there has
been interval progression of the infarct)
(Fig. 10). In 50 % of cases, contrast enhancement
can be seen [2]. Enhancement can be intra-arterial
due to slow flow, in which case it can be seen at
the time of vascular occlusion. Subsequently tran-
sient leptomeningeal enhancement may occur and
usually resolves by the end of the first week.
Parenchymal enhancement due to blood–brain
barrier damage and ingrowth of new immature
vessels may occur by the end of the first week
and last for several weeks [2].
Blood flow to the affected portion of the brain
is typically reestablished 24–72 h after infarction,
during the early subacute phase. By day 3 or
4, new ingrowth of immature vessels with more
permeable blood–brain barriers result in increased
vasogenic edema, which is extracellular swelling
caused by leakage of fluid out of capillaries. In
large infarcts, mass effect typically peaks at
around 5 days and can lead to transfalcine or
transtentorial herniation. Infarcted brain is mildly
T1 hypointense and markedly T2 hyperintense,
involving both gray and white matters with
ill-defined margins. Intensity on DWI and ADC
maps is variable at this stage, reflecting the degree
of cytotoxic edema (decreased ADC) and
340
Fig. 10 Subacute deep and superficial MCA infarct sec-
ondary to carotid occlusion. (a) SWI image demonstrates
hypointensity in the M1 segment of the right MCA. Note
the normal hyperintense flow effects within the left MCA.
(b) MRA reveals absence of flow in the distal right internal
carotid artery and MCA. (c) DWI reveals marked
hyperintensity. (d) ADC reveals marked restricted
vasogenic edema (increased ADC) (Figs. 10 and 11).
If the infarct involves the corticospinal tract,
Wallerian degeneration may occur and manifest as
mild T2 hyperintensity and restricted diffusion
acutely (Fig. 8). In such cases, signal abnormality
will be present in the ipsilateral cerebral peduncle and
pons and should not be mistaken for an additional
area of infarction. In chronic infarcts Wallerian
degeneration is manifested by volume loss in the
ipsilateral cerebral peduncle without signal abnor-
mality (Fig. 6d).
Within 5–14 days, DWI demonstrates
isointensity to mild hyperintensity, while ADC
maps display hyperintensity reflective of
increased diffusion [7] (Fig. 9). T1 hypointensity
J. Rotman and R. Zimmerman
diffusion. (e) FLAIR reveals hyperintensity and mass
effect in the entire right MCA distribution consistent with
infarct of 2–3 days duration. (f) T1 reveals mild cortical
hypointensity and mass effect with gyral swelling and
sulcal obliteration. These findings are typical for an infarct
after 3–5 days
and T2 hyperintensity persist into the late sub-
acute and chronic healing phases. Cortical laminar
necrosis may be visualized as a ribbon of intrinsic
T1 hyperintensity and is usually apparent after
2 weeks [2] (Fig. 12). Cystic encephalomalacia
and lacunar infarcts appear as central regions of
T1 hypointensity, T2 hyperintensity, and FLAIR
hypointensity (similar to spinal fluid intensity)
surrounded by T2 hyperintensity and are best
seen on FLAIR imaging (Fig. 6). On DWI,
chronic infarction is isointense to mildly
hypointense, while the infarct appears
hyperintense on ADC maps due to increased dif-
fusion in the hypocellular infarcted brain
(Fig. 12). If contrast is administered, parenchymal
16 Patterns of Ischemic Stroke: From Lacunar to Territorial to Multiple Embolic to Watershed. . .
Fig. 11 Malignant MCA infarct MR. (a–c) FLAIR
images reveal an infarct involving the majority of the
deep and superficial distributions of the MCA. Note the
enhancement usually begins near the end of the
first week and lasts less than 12 weeks – if it lasts
longer, the presence of an underlying lesion
should be considered [2].
Vascular Imaging
Arterial and venous occlusion or slow flow can be
visualized on FLAIR and enhanced T1-weighted
images. Chronic vessel occlusion or extremely
slow flow in large vessels manifests with
isointensity to hyperintensity on T1-weighted
images (Fig. 13) and hyperintensity on FLAIR
because the intrinsic hyperintensity of blood can
be “captured” due to the absence of flow effects. If
contrast is administered, intraluminal
hyperintensity is more extensive distal to an
occlusion than in normal circulation because the
intrinsic intensity of the enhanced blood is detect-
able secondary to absent flow effects.
Direct visualization of vascular abnormalities
(stenosis occlusion and dissection) can be
obtained with CTA, magnetic resonance angiog-
raphy (MRA), and catheter angiography. The lat-
ter is usually reserved for cases where intervention
is contemplated or when diagnosis is in doubt.
Atherosclerotic narrowing and occlusion occur
in typical locations (e.g., carotid bifurcations in
the neck, M1 segments of middle cerebral
arteries, vertebral artery origins, and intracranial
341
marked mass effect with early medial displacement of the
uncus (a) and mass effect upon and displacement of the
lateral ventricle (b and c)
segments). Dissections of the extracranial internal
carotid artery produce four patterns of luminal
abnormality: smooth tapering to a pointed occlu-
sion, long segment asymmetric narrowing, double
lumen with an intimal flap, and pseudoaneurysm
formation [7]. CTA allows for the detection of
intimal flaps and double lumens and MRA is
excellent at identifying carotid dissection. MRI
imaging of the neck in particular with
fat-suppressed T1-weighted imaging can identify
the intramural hematoma that is the direct result of
the dissection (Fig. 13).
MR Perfusion
Perfusion imaging can be performed with MR as
well as CT. As with CT the brain is imaged as a
bolus of contrast is being injected. The contrast
agent is paramagnetic so it causes intensity in the
brain to decrease proportionate to the amount of
contrast in the vascular system. Whole brain cov-
erage is routinely obtained (as opposed to CT
perfusion). The perfusion parameters are similar
to those obtained from CT, but it is not possible
(or at least not easy) to obtain MTT measurements
and therefore quantification is difficult. However,
since the perfusion diffusion mismatch is qualita-
tively visible, there is no need for quantification to
delineate the core infarct and the penumbra
(Fig. 13).
342
Fig. 12 Chronic infarcts. (a) T1-weighted image demon-
strates cortical hyperintensity and atrophy in the M2 dis-
tribution of the MCA. (b) FLAIR demonstrates cortical
and subcortical T2 hyperintensity and volume loss. (c)
Gradient echo scan demonstrates no evidence of suscepti-
bility hyperintensity to suggest chronic hemorrhage. (d) T1
J. Rotman and R. Zimmerman
hyperintensity is due to pseudolaminar necrosis. (e) In an
example of a left chronic lacunar infarction, FLAIR images
demonstrate hyperintensity. T1-weighted images would
demonstrate corresponding minimal hypointensity. (f)
Lacunar infarction is isointense on DWI and (g)
hyperintense on ADC map
16 Patterns of Ischemic Stroke: From Lacunar to Territorial to Multiple Embolic to Watershed. . . 343

Fig. 13 Carotid dissection with slow flow and large per-
fusion diffusion mismatch. (a) T1- and (b) T2-weighted
images demonstrate the presence of peripheral
hyperintensity surrounding central hypointensity in the
left internal carotid artery at the skull base (“target sign”)
representing intramural clot surrounding the narrow but
patent artery. (c) MRA demonstrates the hyperintense
flow with the lumen surrounded by more amorphous
hyperintense clot in the vessel wall. (d) DWI image dem-
onstrates minimal restricted diffusion in the left corona
radiata (apparent hyperintensity overlying right hemi-
sphere is artifactual). (e) Contrast-enhanced T1-weighted
image demonstrates enhancement of MCA branches due to
slow flow. (f) MR perfusion demonstrates marked prolon-
gation of mean transit time and (g) moderate decrease in
cerebral blood flow
344
Localization of Ischemic Strokes
The intracranial circulation involves paired inter-
nal carotid arteries and vertebral arteries that
merge to form the basilar artery. An anastomotic
network connects the carotid and vertebrobasilar
systems at the circle of Willis provides defense
against major-vessel occlusive disease. However,
collateral circulation is more sparse distal to the
circle of Willis, which makes these areas more
susceptible to infarction. It is thus important to
understand the vascular anatomy and recognize
territorial arterial supply to localize blood vessels
with atherosclerotic disease/thrombosis.
Anterior Territory Infarcts
The internal carotid artery (ICA) has cervical,
petrous, lacerum (or pre-cavernous), and cavern-
ous portions prior to piercing the dura to enter the
intracranial space. The remainder of the internal
carotid is defined by the clinoid, ophthalmic, and
supraclinoid segments. Branches of the ophthal-
mic artery anastomose with those from the facial
and internal maxillary arteries to provide a rich
network of collateral supply should the proximal
carotid artery become occluded. The ophthalmic
artery supplies the orbit, globe, frontal scalp
region, frontal and ethmoidal sinuses, and upper
part of the nose. Occlusion of the ophthalmic
Fig. 14 Anterior choroidal distribution infarction. (a–c) Axial FLAIR images demonstrate acute infarct involving the
uncus, hippocampus, posterior thalamus, and posterior limb of internal capsule
J. Rotman and R. Zimmerman
artery and its supply to the central artery of the
retina can result in blindness.
Acute infarct involving the anterior choroidal
arteries, which arise from the ICA, may involve
the medial temporal cortex, posterior limb of the
internal capsule, posterior and medial aspect of
the lentiform nucleus, body of the caudate
nucleus, and periventricular white matter
[27–30] (Figs. 8 and 14). Patients with anterior
choroidal artery occlusion most commonly pre-
sent with contralateral hemiplegia [29, 31]. Left
lesions may be the source of aphasia, constructive
apraxia, and amnesia [32, 33], while right lesions
could be responsible for spatial neglect
[34–36]. Other possible manifestations include
impaired impulses, prosopagnosia, blindness
(optic tract and radiations), and paralysis (globus
pallidus and posterior limb of the internal cap-
sule). Large AchA territory ischemic lesions are
more associated with embolic pathologies and
portend worse prognoses than small AChA
strokes [37]. MRI findings include FLAIR and
T2-weighted hyperintensity and acute restricted
diffusion. Localization of small choroidal terri-
tories is easily missed on CT and in such cases
MRI with DWI proves very useful (Fig. 8).
In the setting of multiple small or subcortical
infarcts within the same arterial territory, fragmen-
tation of an embolus near its origin or within a
major proximal intracranial artery is implicated.
The main cause of acute multiple brain infarctions
16 Patterns of Ischemic Stroke: From Lacunar to Territorial to Multiple Embolic to Watershed. . . 345

in one cerebral hemisphere is typically MCA

and/or ICA disease (75 %) [6]. Nearly half of all
patients with ICA occlusion have territorial
infarction, although approximately 20 % of high-
grade symptomatic ICA territorial strokes are
associated with cardioembolic or lacunar
etiologies [6].
The ICA terminates at the bifurcation of the
anterior and middle cerebral arteries. Isolated
anterior cerebral artery (ACA) infarcts are rare
causes of ischemic stroke because extracranial
embolic material preferentially enters the middle
cerebral artery and the anterior communicating
artery protects the ACA distribution in the case
of cervical carotid occlusion. Isolated ACA distri-
bution infarction typically occurs as a result of
severe subfalcine herniation or intrinsic arterial
disease associated with occlusion, such as in
patients with diabetes, hypertension, vasospasm,
or vasculitis.
Fig. 15 ACA distribution infarction. FLAIR image dem-
onstrates an acute infarct in the right medial frontal lobe.
ACA Territory Note the hyperintensity in the ACA branches within the
The first portion of the ACA is termed the A1 infarction indicative of slow flow
segment, which turns along the genu of the corpus
callosum before forming the A2 segment. The function and sensation (medial precentral and
medial lenticulostriate arteries arise from the A1 postcentral gyri, respectively), memory, and emo-
segment to supply the anterior inferior basal tion (cingulate gyrus) [7].
ganglia and anterior limb of the internal capsule; On CT, ischemia or infarction in part or all of
specifically, a named branch called the recurrent the ACA vascular territory manifests with loss of
artery of Heubner supplies the head of the caudate gray–white matter differentiation (Fig. 3). Perfu-
nucleus, putamen, and anterior limb of the internal sion CT demonstrates decreased CBF/CBV ratio
capsule. A2 segment branches, such as the and increased MTT/TTP in the medial cerebral
pericallosal and callosomarginal arteries, primar- hemisphere. T2/FLAIR MRI exhibits medial
ily supply the corpus callosum and terminate as hemispheric hyperintensity, while DWI demon-
the splenial and cuneal arteries, respectively strates acute diffusion restriction and
(Fig. 15). corresponding hypointensity is present on the
The anterior communicating artery (ACoA) ADC map. Artery of Huebner infarcts produce
connects the ipsilateral A1 and contralateral A1 changes in the caudate head, anteriomedial puta-
segments of the ACA and should also always be men/globus pallidus, and anterior limb of the
evaluated for aneurysm. The ACA supplies the internal capsule. A2 distribution infarcts produce
medial anteroinferior frontal lobe, the anterior two changes in the parasagittal frontal cortex. There
thirds of the medial hemisphere surface, and a may be posterior extension into the parietal
variable amount of the cortical convexities. parasagittal region as well depending on the sup-
Since distal branches of the ACA supply the ply to this region by the posterior cerebral artery.
olfactory bulbs and tracts, corpus callosum, and
medial aspects of the frontal and parietal lobes, MCA Territory
infarctions affect olfaction, thought processes Infarcts involving the anterior territory are mostly
(medial inferior frontal lobe), lower limb motor related to the middle cerebral artery (MCA). MCA
346
occlusion on CT manifests as hypodensity involv-
ing the frontal, parietal, and temporal lobes and/or
basal ganglia and can occasionally be identified
by the hyperdense vessel sign, reflective of inter-
nal thrombus (Figs. 2 and 10). The MCA cortical
hemispheric supply includes the anterior temporal
lobe, insular, perisylvian, pre- and post-Rolandic
cortex.
The horizontal portion of the MCA is the M1
segment, which gives off lateral lenticulostriate
perforators to supply the putamen, globus
pallidus, superior internal capsule and the adja-
cent corona radiata, and body and posterior head
of the caudate. In cases of lateral lenticulostriate
infarctions, damage of the internal capsule typi-
cally results in hemiparesis and contralateral sen-
sory deficits. Speech can be affected when the left
frontal and temporal lobes are involved, resulting
in global aphasia due to cut off blood supply to
Broca (frontal) and Wernicke (superior temporal)
speech areas. Vision may be compromised by
contralateral hemianopsia/upper quandran-
tanopsia if the optic radiations that sweep from
the lateral geniculate nucleus to Meyer’s loops in
the temporal lobes are involved.
The M2 segment refers to the sylvian segment
of the MCA distal to its branching off into the
anterior division, posterior division, and anterior
temporal artery. The distal branches of the MCA
course lateral to the insula and loop around the
frontal operculum to form the “candelabra” effect
of the sylvian triangle and then become the corti-
cal M3 segment. The last branch of the MCA
sylvian vessels is the angular artery, which sup-
plies the angular gyrus just beyond the sylvian
fissure (Figs. 1 and 12).
Infarction of the MCA may affect motor and
sensory functions of the face, arm, and trunk
(lateral precentral and postcentral gyri, respec-
tively), speech (inferior lateral frontotemporal
gyri), thought processes (anteroinferolateral fron-
tal lobes), hearing (superior temporal gyri), mem-
ory and naming of objects (temporal lobe), and
taste (insular cortex) [7].
Large infarcts (Figs. 2, 4, 10, and 11) are
defined as those covering at least two of the
three MCA territories, including the deep, super-
ficial anterior, and superficial posterior regions.
J. Rotman and R. Zimmerman
Patients with large MCA infarcts demonstrate
higher frequencies of clinical deterioration and
poor prognoses, frequently with high mortality
rates. Large MCA infarctions are associated with
cardioembolism, ICA occlusion, and ICA
dissection.
Acute extracranial carotid occlusion may pro-
duce large areas of infarction involving the deep
basal ganglia and cortical MCA distribution; these
infarcts are typically the result of large distal
emboli associated with proximal occlusion
(Fig. 10). In such cases, the ACA territory is
typically spared because of collateral supply
from the contralateral ACA with the ACoA of
the circle of Willis. The ACA distribution may
be involved if the contralateral A1 segment of
ACA is absent or hypoplastic.
Malignant MCA infarction is a life-threatening
complete or nearly complete MCA infarction
associated with an 80 % mortality rate [6]
(Figs. 4 and 11). This type of infarction occurs in
up to 10 % of all stroke patients and is uniformly
characterized by clinical deterioration within the
first 2–3 days after stroke [6]. In the early phase of
large territory infarction, DWI can accurately pre-
dict transformation into malignant MCA infarc-
tion (MCA volume >145 cm3) with 100 %
sensitivity and 94 % specificity [32] in patients
with persistent arterial occlusion and signs of total
anterior circulation infarction. The main cause of
death is severe post-ischemic brain edema, which
leads to increased intracranial pressure; hence,
early detection is important since treatment with
hypothermia and/or hemicraniectomy can signifi-
cantly reduce mortality [33].
Limited infarcts only involve one of the three
MCA territories and are rarely associated with clin-
ical deterioration. These infarcts are caused by
large-artery atherosclerosis and cardioembolism
in equal numbers. Since cardiac thrombi are
generally larger than thrombi of the large vessels,
there is a lower incidence of cardioembolism in
limited MCA infarcts in contrast to large
infarcts [34].
Combined ACA/MCA
Involvement of both the ACA and MCA terri-
tories is rare and typically fatal. These infarcts
16 Patterns of Ischemic Stroke: From Lacunar to Territorial to Multiple Embolic to Watershed. . .
are typically cardioembolic and occur in the set-
ting of atrial fibrillation. Large clots typically fill
the distal internal carotid artery and A1 and M1
segments of anterior and middle cerebral arteries,
respectively. Combined ACA/MCA infarcts may
also occur in the context of cervical carotid occlu-
sion in the absence of the A1 segment of the ACA
contralateral to the carotid occlusion, thereby
eliminating protective effects of the Circle of
Willis (Fig. 3).
Posterior Territory Infarcts
Proximal disease of the vertebral artery is the most
common finding in large-artery disease involving
the posterior circulation. Vessels that are most
frequently thrombosed by intra-arterial embolism
are the intracranial vertebral artery and distal bas-
ilar artery, leading to infarctions of the posterior
inferior cerebellar (PICA), superior cerebellar
(SCA), and posterior cerebral arteries (PCA),
respectively. Posterior circulation infarcts are
associated with severe brain atrophy, compatible
with longstanding hypoperfusion.
Posterior circulation infarcts can involve indi-
vidual vertebrobasilar branches, perforating
branches or large portions of the posterior
circulation.
Branch Occlusions
The vertebral arteries arise from the subclavian
arteries and course superiorly before entering the
vertebral canal at the C6 transverse foramen (prior
to which they are known as V1 segments). The
arteries then continue upward in the vertebral
foramina (where they are referred to as V2 seg-
ments) before exiting at the C1–C2 level. In 75 %
of cases, the left vertebral artery is larger than the
right and is known as the “dominant” artery [35];
thus, in cases of posterior fossa aneurysm or bas-
ilar artery stenosis, the left side should be care-
fully evaluated. Branches of the vertebral artery
include muscular arteries to the neck, occipital
artery, segmental spinal arteries, and anterior spi-
nal artery, which supplies most of the spinal cord.
347
As the V3 segment travels from the atlas to pierce
the dura to enter the intracranial compartment via
the foramen magnum, posterior meningeal
branches are given off. In the intracranial com-
partment where it is known as the V4 segment, the
first branch is the PICA (although the PICA may
arise from the V3 extracranial segment in a sig-
nificant number of patients).
The PICA supplies the posterolateral medulla,
inferior cerebellar vermis, and inferior aspect of
the cerebellum. The inferior vermian artery and
tonsilohemispheric arteries are the terminating
branches of the PICA. Distal vertebral artery
and/or PICA occlusion can result in lateral med-
ullary (Wallenberg) syndrome (Fig. 16), which
causes loss of pain and temperature sensation on
the contralateral side of the body (lateral
spinothalamic tract) and face on the ipsilateral
side (descending trigeminothalamic tract), ataxia
(inferior cerebellar peduncle), ipsilateral
swallowing and taste disorders (CN IX), hoarse-
ness and ipsilateral paralysis of the vocal cord
(CN X), vertigo and nystagmus (CN VIII), and
ipsilateral Horner’s syndrome (hypothala-
mospinal tract) [36–41]. As the name implies,
the lateral medullary syndrome infarcts involve
the lateral medulla and inferior cerebellum. The
extent of infarction associated with a PICA occlu-
sion is quite variable given the reciprocal size of
PICA and AICA. It is common to have a dominant
PICA or AICA (with the other vessel being small
or absent). As a result, the vascular territories of
these vessels demonstrate considerable overlap,
and occlusion of either vessel can lead to infarc-
tion in the same portions of the brain stem and
cerebellum [42] (Fig. 17). Given the variability of
PICA and AICA, identification of an occluded
vessel may be difficult. This may be particularly
challenging when there is a nondominant verte-
bral artery that does not appear to connect with the
basilar artery. It is common for a nondominant
vertebral artery to end in PICA.
Non-visualization of the V4 segment of the verte-
bral artery in the absence of an ipsilateral PICA
should raise suspicion for a V4 and PICA
occlusion.
AICA infarcts are less common than PICA
infarcts. The AICA arises laterally from the
348
Fig. 16 Lateral medullary infarct in PICA distribution. (a)
DWI and (b) FLAIR images demonstrate an acute infarct
in the posterior lateral medulla and posterior medial infe-
rior cerebellum (lateral medullary syndrome) in the
Fig. 17 PICA/AICA distribution infarct. (a) DWI and (b)
T2-weighted images reveal a large acute infarct involving
the right inferior cerebellum. Note the compression of the
basilar artery to supply the inferolateral pons and
upper medulla via perforators as well as a small
region of the petrous surface of the cerebellum,
the brachium pontis, and the fourth ventricle fora-
men of Luschka and choroid plexus. After the
AICA loops in the auditory canal, it gives rise to
a labyrinthine branch, which supplies the inner ear
cranial nerves VII and VIII then continues on to
supply the anteroinferior aspect of the cerebellum.
Patients with AICA infarcts may present with
lateral inferior pontine syndrome, involving con-
tralateral loss of pain and temperature
(spinothalamic tract), ipsilateral facial paralysis
J. Rotman and R. Zimmerman
distribution of the posterior inferior cerebellar artery. (c)
MRA demonstrates occlusion of the V4 segment of the
vertebral artery
brain stem. (c) Follow-up CT scan 24 h later reveals devel-
opment of hydrocephalus due to compression of the fourth
ventricle
(CN VII), ipsilateral hearing loss and vertigo and
nystagmus (CN VIII), Horner’s syndrome
(hypothalamospinal tract), and ataxia (middle
and inferior cerebellar peduncles) [36,
42–44]. With acute AICA strokes, MRI demon-
strates diffusion restriction and hyperintensity
involving the inferolateral cerebellum and pons.
MRA or CTA may be used to evaluate the vessels.
While isolated unilateral AICA infarcts are likely
due to small artery atherosclerotic disease in dia-
betic patients, more widespread infarctions that
include AICA territory are typically due to basilar
artery occlusive disease.
16 Patterns of Ischemic Stroke: From Lacunar to Territorial to Multiple Embolic to Watershed. . .
Fig. 18 Pontine perforator
infarct. (a) FLAIR imaging
reveals an infarct in the left
pons. Note that the infarct
stops abruptly in the
midline reflecting the
vascular supply to the pons.
(b) MRA reveals
atherosclerotic irregularity
of the mid-basilar artery
Isolated superior cerebellar artery infarcts are
rare. The vessels are relatively large and often
duplicated. When they occur there is involvement
of the cerebellum hemisphere above the great
horizontal fissure and the superior vermis. The
SCA is the last infratentorial branch off the basilar
artery and supplies the upper aspect of the cere-
bellum prior to terminating as the superior
vermian artery. Fine SCA branches help supply
the pons, superior cerebellar peduncle, and infe-
rior colliculus. The superior vermian vessel anas-
tomoses with the inferior vermian artery of the
PICA. Predominantly caused by embolism,
infarcts in the territory of the SCA are uncommon
[45]. SCA occlusion can cause ipsilateral jaw
weakness and facial numbness (spinothalamic
tract and CN V), with the jaw deviating toward
the lesion due to unopposed action of the contra-
lateral medial pterygoid muscle.
Infarcts commonly arise from pontine perforat-
ing branches of the basilar artery. The pontine
perforators are paired paramidline structures that
extend posteriorly into the brain stem. Pontine
perforator infarcts therefore present as unilateral
paramidline lesions that extend posteriorly along
the course of the vessel into the posterior pons
(Fig. 18). These are essentially lacunar infarcts
(see below).
The basilar artery terminates in two PCAs as
well as a few small perforating vessels arising
from its vertical dome. The PCA distribution
includes the medial posterior third of the cerebral
hemispheres and infarcts commonly involve the
349
occipital or inferior temporal lobes including the
hippocampi, as well as regions supplied by the
small premammillary, posterior thalamoperforate,
thalamogeniculate, and peduncular perforating
arteries. These branches arise from the P2 seg-
ment to supply the thalamus, hypothalamus,
geniculate bodies, posterior limb of the internal
capsule, upper midbrain, and choroid plexus. The
PCA splenial branch supplies the splenium and
dorsal corpus callosum. There is variable supply
to the parietal parasagittal cortex. The extent of
infarction depends upon the site of obstruction.
With complete occlusion of the PCA, there may
be involvement of the entire PCA distribution,
including the upper brain stem, thalamus, hippo-
campi, posterior inferior temporal-occipital lobes,
paramidline occipital lobes (including the
calcarine cortex), and parietal lobes (Fig. 19).
More distal occlusions will produce cortical
occipital and parietal infarcts but will spare the
brain stem and thalamus (Figs. 5 and 7).
Although the posterior thalamoperforate arter-
ies are usually paired, the artery of Percheron
(a variant where there is bilateral supply to the
posterior superior brain stem and medial thala-
mus) and the thalamogeniculate branches arise
from the proximal PCA (P2 segment), the poste-
rior communicating artery (PCoA) may contribute
minimal supply to these areas as well (Fig. 20).
Occlusion of these vessels may affect memory
and emotion (anterior thalamus), endocrine func-
tion (hypothalamus), language (pulvinar), pain
sensation (thalami), sight (lateral geniculate),
350 J. Rotman and R. Zimmerman

Fig. 19 Basilar tip

occlusion. (a–c) FLAIR
image reveals an infarct
involving the superior brain
stem, right hippocampus,
and bilateral thalami. (d)
MRA reveals occlusion of
the distal basilar artery.
There is normal flow in the
P2 segment of the right
posterior cerebral artery due
to collateral supply from the
posterior communicating
artery

Fig. 20 Artery of
Percheron infarct. (a) DWI
and (b) FLAIR images
reveal bilateral medial
thalamic infarcts in the
distribution of the artery of
Percheron

and motor control (subthalamic nuclei). The P1
segment of the PCA joins the anterior circulation
via the PCoA. The PCoA thus connects the ante-
rior carotid circulation with the posterior
vertebrobasilar circulation, supplying parts of the
thalamus, hypothalamus, optic chiasm, and
mammillary bodies. It is important to note
that PCA ischemia may arise from anterior circu-
lation emboli if PCA of fetal origin is present and
that the PCoA allows emboli from both anterior
and posterior circulations to affect PCA
distribution.
16 Patterns of Ischemic Stroke: From Lacunar to Territorial to Multiple Embolic to Watershed. . .
The PCA continues in its P3 and P4 distal
segments, giving rise to anterior and posterior
temporal arteries and a posterior pericallosal
artery prior to terminating as parieto-occipital
branches and calcarine arteries to supply the
occipital lobe. PCA infarction most commonly
affects vision (occipital lobes) but can also affect
memory (posteroinferior temporal lobe), smell
(hippocampal region), and emotion (posterior
fornix).
Occlusion of the PCA can result in medial
midbrain (Weber) syndrome, characterized by
contralateral spastic paresis (corticospinal
and corticobulbar tracts), ipsilateral CN III
palsy, and contralateral homonymous hemianopsia
with macular sparing. Damage to the visual cortex
can result in alexia without agraphia, as the right
visual cortex is unable to communicate with Broca
and Wernicke areas. Large- and small-vessel dis-
eases are the usual culprits, whereas cardiogenic
embolism is rare [46]. PCA occlusion on CT dem-
onstrates low attenuation or loss of gray–white
margins in PCA territory, while MRI depicts dif-
fusion restriction and FLAIR and T2-weighted
hyperintensity (Fig. 7). MRI is best for localizing
small infarcts and subtle midbrain and/or temporal
lobe findings.
Fig. 21 Acute basilar
artery occlusion. (a and b)
CT scans reveal
hypodensity throughout the
left cerebellum, brain stem,
and bilateral occipital lobes
with a dense basilar artery.
Note the marked
compression of the
quadrigeminal plate cistern
and hydrocephalus
351
Basilar Occlusion
Occlusion of the basilar artery produces bilateral
infarction in all vascular territories superior to the
distribution of the PICA. Depending on the com-
petency of the posterior communicating arteries,
there may be partial or complete unilateral or
bilateral sparing of the posterior cerebral artery
distribution. The distal basilar artery provides per-
forators of rostral pontine, median, paramedian, and
lateral pontine branches and generates the SCA
prior to terminating as PCA branches. Pontine
artery occlusion may result in medial inferior
(Millard–Gubler) pontine syndrome, which is char-
acterized by contralateral motor and sensory deficits
(corticospinal tracts and medial lemniscus, respec-
tively), internuclear ophthalmoplegia (medial lon-
gitudinal fasciculus), ipsilateral gaze paralysis
(paramedian pontine reticular formation), ipsilateral
lateral rectus paralysis (CN VI), and ipsilateral facial
weakness and partial loss of taste (facial colliculus)
[44, 47–50] (Figs. 19, 21, and 22).
Acute occlusion of the basilar artery is a true
neurointerventional emergency. Patients with
acute occlusion of the basilar artery present with
sudden and dramatic neurological impairment
and, without early treatment, rapid deterioration.
352 J. Rotman and R. Zimmerman

Fig. 22 Acute basilar

artery occlusion. (a–c) DWI
images reveal restricted
diffusion in the cerebellum,
right hippocampus, right
occipital lobe, and bilateral
thalami. (d) T2-weighted
image reveals the absence
of a flow void in the distal
basilar artery. (e) MRA
reveals complete absence of
flow in the basilar artery and
its branches

Complete bilateral lesions to the pons will affect
the reticular activating substance, which affects
coordination and motor control and manifests as
locked-in syndrome with complete paralysis
[51]. Acute basilar artery occlusion can be due to
either thromboembolism or thrombosis from ath-
erosclerosis or propagation of an intracranial
dissection [52].
The earliest sign on CT of acute occlusion is a
hyperdense basilar artery (Fig. 21) analogous to
the dense MCA sign followed by the development
of hypodensity and loss of gray–white differenti-
ation in the bilateral temporal lobes, occipital
lobes, midbrain, thalami, and peduncles.
Diagnosis of hyperacute basilar thrombosis on
CT is challenging because calcifications within
the basilar artery are common (much more so
than within M1 segment of MCA) and density
changes within the brain are difficult to detect
because of beam-hardening artifact. MR is supe-
rior to CT allowing for detection of parenchymal
changes (hyperintensity on DWI and/or FLAIR)
and visualization of intraluminal clot (GrE and/or
SWI) (Fig. 22).
Large-vessel atherosclerosis of the
vertebrobasilar arteries has been demonstrated in
more than two thirds of patients with cerebellar
infarction [6], which only accounts for
16 Patterns of Ischemic Stroke: From Lacunar to Territorial to Multiple Embolic to Watershed. . .
approximately 1.5–2.3 % of all cerebral infarc-
tions [53, 54]. The cerebellum swells with an
infarction involving more than one third of its
volume, a basilar artery with poor collateral sup-
ply, an embolus with reperfusion, and/or a mas-
sive SCA infarction [7]. However, cerebellar
infarcts are often difficult to detect on CT due to
beam-hardening artifact or partial volume averag-
ing, which can mask subtle regions of low density.
Attention to subtle asymmetry involving the
fourth ventricle and quadrigeminal plate cistern
may aid in the identification of cerebellar edema.
Small cerebellar infarcts less than 2 cm are best
identified on MRI. Obstructive hydrocephalus
may occur, which manifests early as dilatation of
the temporal horns (Fig. 17). Minimal mass effect
and slight ventriculomegaly can rapidly evolve to
large-volume strokes with compression of the
brainstem and cerebellar herniation; the superior
vermis can herniate superiorly through the
tentorium, while the tonsils and inferior vermis
may herniate downward through the foramen
magnum.
Watershed Infarcts
Watershed, or border zone, infarctions occur at the
most distal areas between arterial territories and
are thought to be secondary to severe
hypoperfusion, such as in cases of severe stenosis
or occlusion of the carotid artery and prolonged
hypotension [55]. Also called hypotensive cere-
bral infarctions, they are typically seen at
gray–white matter junctions as well as between
perforating arteries in deep white matter. Reduc-
tion of flow primarily affects these zones to the
greatest extent due to insufficient CBF to meet
metabolic demands.
Two vascular border zone areas are present
within the cerebral hemispheres, the cortical and
the internal. Anterior cortical border zone areas
include the cortical blood supply of the ACA and
MCA, while the posterior cortical border zone
areas are located between the MCA and PCA.
Commonly associated with severe ICA or MCA
occlusion/stenosis, internal border zone areas are
located between the ACA, MCA, and PCA.
353
Watershed infarcts (WIs) that occur at the junction
between two or three arterial territories account
for approximately 10 % of ischemic strokes [6]
(Fig. 23).
In patients with ICA occlusion, hypotensive
cerebral infarcts account for 72 % of delayed
strokes, although they are rarely the initial mani-
festation of ICA occlusion [56]. Conversely, 50 %
of patients with high-grade ICA stenosis suffer
watershed infarcts [6]. Cerebrovascular investiga-
tion of patients with MCA watershed infarcts pre-
sumably due to atherosclerotic disease should
include evaluation of the large extra- and intracra-
nial vessels.
Bilateral hypotensive cerebral infarcts are
mainly attributed to global reduction in perfusion
pressure in the setting of significant hypoxia/
hypovolemia or hemodynamic impairment in
patients with diffuse cerebral vascular diseases,
such as sickle cell anemia (Fig. 24). Unilateral
presentations after global hypoxic-ischemic
events are possible in cases of focal underlying
vascular stenosis and relative hypoperfusion.
Decreased perfusion can also alter blood flow
and thus allow for microemboli to reach recipient
blood vessels; therefore, even in the absence of
severe stenosis, reduction of flow can result in
ischemia due to decreased circulation of distal
emboli, which can enable regional ischemia to
develop.
On CT, major arterial border zone infarcts pre-
sent with hypodensity at gray–white matter junc-
tions between vascular territories. In severe cases,
there is effacement of nearly all supratentorial
gray–white matter junctions. The basal ganglia
and thalami may also be affected. Deep white
matter hypotensive cerebral infarcts typically
manifest with more than three white matter lesions
within the centrum semiovale, similar in location
to areas affected by prolonged hypoxia, although
changes appear multifocal rather than diffuse. A
“string of pearls” appearance is used to describe
the linear orientation of multiple watershed
infarcts that occur parallel to the lateral ventricles
(Fig. 23). CTA can be performed to determine the
degree of severity in cases of ICA occlusion.
There is decreased CBF to affected areas on CT
perfusion studies.
354 J. Rotman and R. Zimmerman

Fig. 23 Watershed (border

zone) infarct. (a–c) DWI
reveals multiple foci of
restricted diffusion (“string
of pearls sign”) at the
borders between the
anterior and middle cerebral
artery territories. (d) MRA
reveals focal stenosis of the
supraclinoid internal carotid
artery with decreased flow
in the proximal internal
carotid artery

Fig. 24 Chronic lacunar infarct. (a) FLAIR image dem- differentiation from a dilated perivascular space. (b) On
onstrates a right periventricular infarct which has central DWI the infarct is hypointense, while on (c) ADC it is
fluid intensity and peripheral T2 hyperintensity. The pres- hyperintense
ence of peripheral hyperintensity on FLAIR allows for
16 Patterns of Ischemic Stroke: From Lacunar to Territorial to Multiple Embolic to Watershed. . .
The best imaging tool for detecting hypoten-
sive cerebral infarction is MRI, including SWI,
MRA, and DWI. Acute infarcts demonstrate
hypointensity and edematous gyri on
T1-weighted images, while subacute infarcts
may display gyriform cortical hyperintensity,
termed pseudolaminar necrosis (Fig. 12), and
enhance after contrast administration.
T2-weighted images show hyperintensity in
involved areas as well as cisternal/sulcal compres-
sion in severe cases. FLAIR images can demon-
strate intraluminal arterial hyperintensity in cases
of proximal vessel thrombosis or severe stenosis.
Lacunar Infarcts
The word “lacune,” translated as “hole” in Latin,
is used to describe small fluid-filled cavities
mostly present in the basal ganglia and thalami.
Lacunar infarcts (LIs) are deep cerebral infarcts
less than 1.5 cm in diameter, although this size
criteria does not apply in the acute phase due to
the presence of cellular swelling and extracellular
edema. LIs usually represent the healed stage of
small deep infarcts, which likely occurred second-
ary to occlusion of a single penetrator artery aris-
ing from the large arteries of the circle of Willis or
from the basilar artery. The most commonly
involved arteries involved include the
lenticulostriate, thalamoperforate, and anterior
choroidal artery perforators.
LIs are found in approximately 11 % of
patients admitted with stroke [9]. LIs can involve
the upper two thirds of the basal ganglia, the
internal capsule, the thalamus, or the paramedian
and lateral regions of the brainstem (Figs. 8, 9, and
12). Brainstem LIs are mainly found in the
paramidline region of the pons, which is supplied
from midline and anteromedial perforators arising
from the basilar artery (Fig. 18). LIs are less
commonly located in the medulla and midbrain,
as these areas are supplied by short arterioles and
are thus less vulnerable to aging and hypertension.
While symptomatic initial LIs are usually located
in the area supplied by anterior choroidal artery
[57] (Figs. 8 and 14), multiple asymptomatic LIs
are commonly located within the territory
355
supplied by the perforating arteries arising from
the MCA or ACA.
Patients with lacunar infarcts may have normal
CT examinations. MR is more sensitive and allows
for identification of most small lesions regardless
of their age (Figs. 8, 9, and 12). Most lacunar
infarcts are clinically silent, and subtle neurological
deficits may go unnoticed by patients and/or phy-
sicians. Clinical presentation depends on the size,
location, and number of lacunar infarcts. Pure
motor hemiparesis is the most common lacunar
syndrome (55 %) [9, 58]. Patients who present
with classic lacunar syndromes have high proba-
bilities of recovery and are typically treated with
antiplatelet therapy for secondary prevention [5].
As aforementioned, acute lacunar infarcts are
often undetected and are typically impossible to
distinguish from chronic lacunar infarcts on CT,
although the presence of distinct convex margins
is suggestive of acute disease. Because of their
small size, many lacunar infarcts are not visual-
ized on CT scans. When they are found on CT,
lacunar infarcts appear as small discrete round/
ovoid foci of hypodensity that may enhance in
the late acute or early subacute stage. The Massa-
chusetts General Hospital group recommends
narrowing CT window widths and levels to better
identify subtle hypoattenuating lesions that por-
tend infarct, although since the advent of
low-dose algorithms to limit radiation exposure,
the increased noise inherent to such scans makes
the detection of subtle infarcts more difficult [6].
On MRI, acute lacunar infarcts present as foci
of T1 iso- to hypointensity and T2 hyperintensity.
Similar to the case of CT, they remain difficult to
distinguish from chronic lacunar infarcts on these
sequences, particularly in patients with chronic
microvascular ischemia depicted as multiple foci
of T2 hyperintensity in the white matter. DWI
makes detection of acute lacunar infarcts easy
since these lesions are hyperintense, while chronic
white matter lacunar infarctions and ischemic
changes are DWI isointense (Fig. 12). Lacunar
infarcts often demonstrate peripheral
hyperintensity surrounding a fluid intensity core
on FLAIR thus allowing for differentiation from
large VR spaces, which lack the peripheral
hyperintensity of lacunar infarcts (Fig. 24).
356
Patients with thalamic lesions (Figs. 9, 11,
and 24) may present with sensorimotor and
behavioral symptoms due to interruption of recip-
rocal cortical-thalamic connections. There are
four major thalamic vascular territories, each
with a predilection for supplying particular groups
of nuclei: (1) tuberothalamic, (2) inferolateral,
(3) paramedian, and (4) posterior choroidal ves-
sels. Patients with tuberothalamic territory strokes
may display impaired arousal, orientation, learn-
ing, memory, and executive function as well as
superimposition of temporally unrelated informa-
tion and emotional facial paresis [59, 60].
Paramedian thalamic infarcts also cause
decreased arousal (especially if bilateral) and
impaired learning and memory [59]. Confusion,
agitation, aggression, and apathy may be persis-
tent features [61–63]. Left paramedian and left
tuberothalamic lesions that include the ventrolat-
eral nucleus produce language deficits, whereas
right thalamic lesions in these territories result in
visual-spatial deficits, including hemispatial
neglect [59]. Inferolateral territory strokes cause
contralateral hemisensory loss, hemiparesis
and hemiataxia, and pain syndromes that more
commonly occur with right thalamic lesions
[59, 64, 65].
Arising from the PCA, the medial and lateral
posterior choroidal arteries supply the trigone
region of the lateral ventricle. The medial poste-
rior choroidal arteries pass around the midbrain
and course medially toward the pineal gland.
They supply the tectum, choroid plexus of the
third ventricle, and thalami. The lateral posterior
choroidal arteries run behind the pulvinar into the
choroidal fissure to supply the choroid of the
lateral and third ventricles, posterior thalamus,
and fornix. The anterior choroidal artery territory
anastomoses with posterior medial and lateral
choroidal branches. Patients with posterior cho-
roidal lesions present with visual field deficits,
variable sensory loss, weakness, dystonia,
tremors, and occasional amnesia and language
impairment [59]. MRI findings are similar to
those with anterior choroidal artery infarcts, with
FLAIR and T2-weighted hyperintensity and acute
restricted diffusion.
J. Rotman and R. Zimmerman
Anoxia and Hypoxia
In anoxic injuries, the metabolically active areas of
the brain such as the basal ganglia, dentate nucleus,
and hippocampus are the most commonly affected.
Corresponding CT abnormalities may take as long
as 24 h to become visible. The earliest CT finding is
subtle loss of normal gray matter hyperdensity,
leading to the inability to discern the basal ganglia
and thalami from the internal capsules (Fig. 25).
Over time, as in the case with localized strokes,
hypodensity and mass effect become apparent. The
delayed CT appearance of anoxic insult may result
from maintenance of the CBV (which in contrast to
hyperacute infarction, for example, is diminished);
therefore, CT changes only become visible once
vasogenic edema has set in. In contradistinction,
anoxic brain injuries can be visualized within as
early as 3 h on MRI due to cytotoxic edema and cell
death, which result in DWI hyperintensity and
restricted diffusion on ADC maps (Fig. 26).
Changes that result from carbon monoxide toxicity
are similar to those seen in anoxia but are most
pronounced in the bilateral globus pallidi (Fig. 27).
If delayed encephalopathy occurs 2–3 weeks after
recovery, additional findings include T2-weighted
hyperintensity involving the corpus callosum, sub-
cortical U fibers, and internal and external capsules
and hypointensity in the thalamus and putamen
[7]. In patients who survive, chronic anoxic injury
results in basal ganglia and hippocampal atrophy
with secondary dilation of the frontal and temporal
horns of the lateral ventricles.
Prolonged hypoxia leads to hypoperfusion and
the portions of the brain furthest from the circle of
Willis are preferentially affected. Thus, the
lesions begin in the border zone lesions and at
least initially spare the more proximal but
more metabolically active deep portions of the
brain that are preferentially affected by
anoxia. On CT there may be subtle loss of
bilateral superficial gray–white differentiation
followed by diffuse brain swelling with sulcal
and ventricular compression. On MR there is dif-
fuse symmetric cortical/subcortical hyperintensity
on FLAIR and DWI may mimic a normal exam
(Fig. 28).
16 Patterns of Ischemic Stroke: From Lacunar to Territorial to Multiple Embolic to Watershed. . .
Fig. 25 Anoxia progressing to brain death. (a) Initial CT
scan performed 6 h after an episode of near drowning
demonstrates normal gray matter density in the deep gray
matter and cortex. (b) Follow-up scan at 18 h reveals loss
of normal density in the basal ganglia and thalami with
Fig. 26 Anoxia. (a) CT
demonstrates diffuse
hypodensity in the bilateral
basal ganglia with mass
effect on the lateral
ventricles. (b) FLAIR
image demonstrates diffuse
T2 hyperintensity in the
same regions
In severe cases of anoxia or hypoxia, CT find-
ings progress to diffuse cerebral edema with sulcal
and cisternal obliterations. The brain will appear
diffusely hypodense without any gray–white mat-
ter distinction and will be nearly completely fea-
tureless with the exception of visualization of
major arterial structures near the circle of Willis
and the dura. These findings may mimic subarach-
noid hemorrhage. Increased intracranial pressure
produces central and tonsillar herniations, leading
357
preserved density in the cortical gray matter. (c) Scan at
48 h reveals complete loss of normal gray matter density in
the basal ganglia and thalamus as well as the cortex. Sulci
have been obliterated and the ventricles are small, indica-
tive of brain death
to complete cessation of CBF and resultant brain
death (Fig. 25).
Infarcts related to fat emboli commonly affect
both gray and white matter and mimic thrombo-
embolic strokes. Diffuse white matter ischemia
can result directly from a massive fat embolism
or in combination with hypoxia from pulmonary
complications. A “hypodense MCA” sign may be
the only sign present on initial CT, related to the
presence of fat within the MCA. DWI will be
358 J. Rotman and R. Zimmerman

Fig. 27 Carbon monoxide

poisoning. (a and b) FLAIR
imaging demonstrates
hyperintensity in the
hippocampi and globus
palladi

Fig. 28 Severe prolonged

hypoxia. (a and b) Axial
DWI and (c and d) FLAIR
images demonstrate diffuse
gyral hyperintensity. Note
the sparing of the basal
ganglia. The global
abnormality creates a
“pseudonormal”
appearance
16 Patterns of Ischemic Stroke: From Lacunar to Territorial to Multiple Embolic to Watershed. . .
Fig. 29 Fat emboli. (a) Axial FLAIR and (b) DWI at the
level of the lateral ventricles demonstrate multiple foci of
T2 hyperintensity. (c) SWI demonstrates numerous foci of
positive in acute cases due to extensive small-
vessel ischemia and hypoxia; foci of cytotoxic
edema from multiple microemboli manifest as
bright spots on a dark background, which is
referred to as the “starfield” pattern [66]
(Fig. 29). T1-weighted images are usually
unremarkable, while T2 and FLAIR images may
demonstrate small, scattered foci of
hyperintensity, mimicking the pattern of water-
shed infarcts.
359
susceptibility hypointensity indicative of
microhemorrhage from microemboli, referred to as the
“starfield” pattern, consistent with fat embolization
Summary
Stroke is a leading cause of mortality and morbid-
ity in the developed world. The goals of imaging
in the setting of acute stroke are to establish a
diagnosis as early as possible and to deduce accu-
rate information regarding the intracranial vascu-
lature and brain perfusion to guide in the selection
of appropriate therapy. While CT has the
360
advantage of detecting acute hemorrhage, MRI
offers more accurate pathophysiological informa-
tion. Different imaging characteristics associated
with different stages of infarction significantly
affect the decision of whether or not to administer
thrombolytic therapy. For example,
hyperintensity on MRI diffusion-weighted imag-
ing may be the only sequence to identify
hyperacute infarct and thus indicate candidacy
for acute stroke management.
The most common causes of infarction include
large-artery atherosclerosis, cardioembolism, and
lacunes, and outcomes tend to differ depending on
subtype. For instance, large-artery lesions have
higher associated mortality rates than lacunar
infarcts, while cardioembolic strokes are associ-
ated with more frequent recurrence and higher
1-month mortality rates. Since characteristic
MRI imaging patterns can discern specific ische-
mic stroke etiologies, certain findings are able to
guide more personalized patient management. For
example, whereas large cortical territorial infarcts
are typically associated with cardioembolism and
warrant cardiac work-up, lacunar infarcts are typ-
ically treated with anticoagulation. Moreover,
understanding that intracranial emboli most com-
monly produce MCA territory infarcts cues the
physician to investigate the internal carotid artery,
since emboli that enter the carotid artery preferen-
tially lodge within MCA branches. Appreciating
this pattern also affects patient treatment, since
rather than medical management, carotid endar-
terectomy and stenting is indicated for large-
vessel disease.
References
1. Picanco MR et al (2014) Reperfusion after 4.5 hours
reduces infarct growth and improves clinical outcomes.
Int J Stroke 9(3):266–269
2. Allen LM et al (2012) Sequence-specific MR imaging
findings that are useful in dating ischemic stroke.
Radiographics 32(5):1285–1297; discussion 1297–9
3. Bamford J et al (1991) Classification and natural his-
tory of clinically identifiable subtypes of cerebral
infarction. Lancet 337(8756):1521–1526
4. Adams HP Jr et al (1993) Classification of subtype of
acute ischemic stroke. Definitions for use in a
J. Rotman and R. Zimmerman
multicenter clinical trial. TOAST. Trial of Org 10172
in Acute Stroke Treatment. Stroke 24(1):35–41
5. Kolominsky-Rabas PL et al (2001) Epidemiology of
ischemic stroke subtypes according to TOAST criteria:
incidence, recurrence, and long-term survival in ische-
mic stroke subtypes: a population-based study. Stroke
32(12):2735–2740
6. Rovira A, Grive E, Alvarez-Sabin J (2005) Distribution
territories and causative mechanisms of ischemic
stroke. Eur Radiol 15(3):416–426
7. Zimmerman RD (2010) Vascular diseases of the brain.
In: Yousem DM, Grossman RI (eds) Neuroradiology:
the requisites. Mosby, Philadelphia, pp 104–169
8. Jung JM et al (2010) Difference in infarct volume and
patterns between cardioembolism and internal carotid
artery disease: focus on the degree of cardioembolic
risk and carotid stenosis. Cerebrovasc Dis 29
(5):490–496
9. Arboix A, Alio J (2012) Acute cardioembolic cerebral
infarction: answers to clinical questions. Curr Cardiol
Rev 8(1):54–67
10. Ferro JM (2003) Cardioembolic stroke: an update.
Lancet Neurol 2(3):177–188
11. Hart RG, Halperin JL (2001) Atrial fibrillation and
stroke: concepts and controversies. Stroke 32
(3):803–808
12. Yesilot Barlas N et al (2013) Etiology of first-ever
ischaemic stroke in European young adults: the 15 cit-
ies young stroke study. Eur J Neurol 20(11):1431–1439
13. Timsit SG et al (1993) Brain infarction severity differs
according to cardiac or arterial embolic source. Neu-
rology 43(4):728–733
14. Horowitz DR et al (1992) Mechanisms in lacunar
infarction. Stroke 23(3):325–327
15. Lodder J et al (1990) Are hypertension or cardiac
embolism likely causes of lacunar infarction? Stroke
21(3):375–381
16. Tatu L et al (1996) Arterial territories of human brain:
brainstem and cerebellum. Neurology 47
(5):1125–1135
17. Tatu L et al (1998) Arterial territories of the human
brain: cerebral hemispheres. Neurology 50
(6):1699–1708
18. Nakano S et al (2001) Correlation of early CT signs in
the deep middle cerebral artery territories with angio-
graphically confirmed site of arterial occlusion. AJNR
Am J Neuroradiol 22(4):654–659
19. Becker H et al (1979) CT fogging effect with ischemic
cerebral infarcts. Neuroradiology 18(4):185–192
20. Hopyan J et al (2010) Certainty of stroke diagnosis:
incremental benefit with CT perfusion over
noncontrast CT and CT angiography. Radiology 255
(1):142–153
21. Wechsler LR (2011) Imaging evaluation of acute ische-
mic stroke. Stroke 42(Suppl 1):S12–S15
22. Meretoja A et al (2014) Stroke thrombolysis: save a
minute, save a day. Stroke 45(4):1053–1058
23. Gonzalez RG et al (1999) Diffusion-weighted MR
imaging: diagnostic accuracy in patients imaged within
16 Patterns of Ischemic Stroke: From Lacunar to Territorial to Multiple Embolic to Watershed. . .
6 hours of stroke symptom onset. Radiology 210
(1):155–162
24. Mitomi M et al (2014) Comparison of CT and DWI
findings in ischemic stroke patients within 3 hours of
onset. J Stroke Cerebrovasc Dis 23(1):37–42
25. Bryan RN et al (1991) Diagnosis of acute cerebral
infarction: comparison of CT and MR imaging. AJNR
Am J Neuroradiol 12(4):611–620
26. Brazzelli M (2009) Magnetic resonance imaging ver-
sus computed tomography for detection of acute vas-
cular lesions in patients presenting with stroke
symptoms. Cochrane Database Syst Rev (4):Cd007424
27. Huang IJ et al (2001) Time course of cerebral infarction
in the middle cerebral arterial territory: deep watershed
versus territorial subtypes on diffusion-weighted MR
images. Radiology 221(1):35–42
28. Warach S et al (1995) Acute human stroke studied by
whole brain echo planar diffusion-weighted magnetic
resonance imaging. Ann Neurol 37(2):231–241
29. Perkins CJ et al (2001) Fluid-attenuated inversion
recovery and diffusion- and perfusion-weighted MRI
abnormalities in 117 consecutive patients with stroke
symptoms. Stroke 32(12):2774–2781
30. Toyoda K, Ida M, Fukuda K (2001) Fluid-attenuated
inversion recovery intraarterial signal: an early sign of
hyperacute cerebral ischemia. AJNR Am J Neuroradiol
22(6):1021–1029
31. Emeriau S et al (2013) Can diffusion-weighted
imaging-fluid-attenuated inversion recovery mismatch
(positive diffusion-weighted imaging/negative fluid-
attenuated inversion recovery) at 3 Tesla identify
patients with stroke at <4.5 hours? Stroke 44
(6):1647–1651
32. Oppenheim C et al (2000) Prediction of malignant
middle cerebral artery infarction by diffusion-weighted
imaging. Stroke 31(9):2175–2181
33. Georgiadis D et al (2002) Hemicraniectomy and mod-
erate hypothermia in patients with severe ischemic
stroke. Stroke 33(6):1584–1588
34. Tei H et al (2000) Deteriorating ischemic stroke in
4 clinical categories classified by the Oxfordshire
Community Stroke Project. Stroke 31(9):2049–2054
35. Yousem DM, Grossman RI (2010) Cranial anatomy.
In: Yousem DM, Grossman RI (eds) Neuroradiology:
the requisites. Mosby, Philadelphia, pp 23–57
36. Ross MA et al (1986) Magnetic resonance imaging in
Wallenberg’s lateral medullary syndrome. Stroke 17
(3):542–545
37. Ogawa K et al (2013) Clinical study of medial area
infarction in the region of posterior inferior cerebellar
artery. J Stroke Cerebrovasc Dis 22(4):508–513
38. Peterman AF, Sieker TR (1960) The lateral medullary
(Wallenberg) syndrome: clinical features and progno-
sis. Med Clin North Am 44:887–896
39. Grant G (1966) Infarction localization in a case of
Wallenberg’s syndrome. A neuroanatomical investiga-
tion with comments on structures responsible for nys-
tagmus, impairment of taste and deglutition.
J Hirnforsch 8(5):419–430
361
40. Norrving B, Cronqvist S (1991) Lateral medullary
infarction: prognosis in an unselected series. Neurol-
ogy 41(2 (Pt 1)):244–248
41. Kim JS (2003) Pure lateral medullary infarction:
clinical-radiological correlation of 130 acute, consecu-
tive patients. Brain 126(Pt 8):1864–1872
42. Amarenco P et al (1993) Anterior inferior cerebellar
artery territory infarcts. Mechanisms and clinical fea-
tures. Arch Neurol 50(2):154–161
43. Oas JG, Baloh RW (1992) Vertigo and the anterior
inferior cerebellar artery syndrome. Neurology 42
(12):2274–2279
44. Kim JS, Choi-Kwon S (1999) Sensory sequelae of
medullary infarction: differences between lateral and
medial medullary syndrome. Stroke 30
(12):2697–2703
45. Kumral E, Kisabay A, Atac C (2005) Lesion
patterns and etiology of ischemia in superior cerebellar
artery territory infarcts. Cerebrovasc Dis 19
(5):283–290
46. Kim JS, Kim J (2005) Pure midbrain infarction: clini-
cal, radiologic, and pathophysiologic findings. Neurol-
ogy 64(7):1227–1232
47. Onbas O et al (2005) Millard-Gubler syndrome: MR
findings. Neuroradiology 47(1):35–37
48. Ahdab R et al (2013) Pure ipsilateral central facial
palsy and contralateral hemiparesis secondary to
ventro-medial medullary stroke. J Neurol Sci 332
(1–2):154–155
49. Toyoda K et al (1996) Medial medullary infarction:
analyses of eleven patients. Neurology 47
(5):1141–1147
50. Kim JS, Kim HG, Chung CS (1995) Medial medullary
syndrome. Report of 18 new patients and a review of
the literature. Stroke 26(9):1548–1552
51. Caplan LR (1980) “Top of the basilar” syndrome.
Neurology 30(1):72–79
52. Verdelho A, Pereira JG, Ferro JM (1999) Multiple
vertebro-basilar infarcts and cardio-embolism. Rev
Neurol 28(11):1027–1030
53. Tohgi H et al (1993) Cerebellar infarction. Clinical and
neuroimaging analysis in 293 patients. The Tohoku
Cerebellar Infarction Study Group. Stroke 24
(11):1697–1701
54. Macdonell RA, Kalnins RM, Donnan GA (1987) Cer-
ebellar infarction: natural history, prognosis, and
pathology. Stroke 18(5):849–855
55. Derdeyn CP et al (2001) Severe hemodynamic impair-
ment and border zone–region infarction. Radiology
220(1):195–201
56. Bogousslavsky J, Regli F (1986) Borderzone infarc-
tions distal to internal carotid artery occlusion: prog-
nostic implications. Ann Neurol 20(3):346–350
57. Boiten J, Lodder J, Kessels F (1993) Two clinically
distinct lacunar infarct entities? A hypothesis. Stroke
24(5):652–656
58. De Reuck J, De Groote L, Van Maele G (2008) The
classic lacunar syndromes: clinical and neuroimaging
correlates. Eur J Neurol 15(7):681–684
362
59. Schmahmann JD (2003) Vascular syndromes of the
thalamus. Stroke 34(9):2264–2278
60. Lisovoski F et al (1993) Left tuberothalamic artery
territory infarction: neuropsychological and MRI fea-
tures. Eur Neurol 33(2):181–184
61. Castaigne P et al (1981) Paramedian thalamic and
midbrain infarct: clinical and neuropathological study.
Ann Neurol 10(2):127–148
62. Graff-Radford NR et al (1984) Nonhemorrhagic infarc-
tion of the thalamus: behavioral, anatomic, and physi-
ologic correlates. Neurology 34(1):14–23
63. Graff-Radford NR et al (1985) Nonhaemorrhagic tha-
lamic infarction. Clinical, neuropsychological and
electrophysiological findings in four anatomical
J. Rotman and R. Zimmerman
groups defined by computerized tomography. Brain
108(Pt 2):485–516
64. Caplan LR et al (1988) Lateral thalamic infarcts. Arch
Neurol 45(9):959–964
65. Nasreddine ZS, Saver JL (1997) Pain after
thalamic stroke: right diencephalic predominance and
clinical features in 180 patients. Neurology 48
(5):1196–1199
66. Parizel PM et al (2001) Early diagnosis of cerebral fat
embolism syndrome by diffusion-weighted MRI
(starfield pattern). Stroke 32(12):2942–2944
67. Han SW et al (2007) Frequency and significance of
cardiac sources of embolism in the TOAST classifica-
tion. Cerebrovasc Dis 24(5):463–468
 Territorial Strokes as a Tool to Learn
Vascular Territories 17
Behroze Adi Vachha and Pamela Whitney Schaefer

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364 Stroke is the fourth leading cause of death in
the United States and a leading cause of seri-
General Overview of Intracranial Arteries . . . . . . . 364
ous, long-term adult disability. One of the key
Anterior Circulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365 tasks in establishing the diagnosis of acute
Intradural Internal Carotid Artery . . . . . . . . . . . . . . . . . . . 365 ischemic stroke and determining its appropri-
Anterior Cerebral Artery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
Middle Cerebral Artery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370 ate treatment is establishing the arterial terri-
tory affected. Knowledge of cerebral vascular
Posterior Circulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372
Intradural Vertebral Artery . . . . . . . . . . . . . . . . . . . . . . . . . . . 372
territories helps identify abnormal vessels on
Basilar Artery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374 CT and conventional angiograms, confirms
Posterior Cerebral Artery . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376 that a DWI hyperintense lesion represents an
Border Zones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378 acute arterial stroke, and guides further inves-
tigations and treatment. Factors contributing to
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379
understanding vascular territories include the
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379 anatomy of the intracranial circulation and its
normal variants and the intrinsic variability in
the extent of brain supplied by the main
branches of the intracranial circulation. This
chapter reviews the normal anatomy of intra-
cranial arteries and describes the vascular dis-
tribution as seen on CT and MRI using major
territorial strokes as a learning tool.

Keywords
Stroke • Infarct • Vascular territories

B.A. Vachha • P.W. Schaefer (*)

Neuroradiology, Massachusetts General Hospital, Boston,
MA, USA
e-mail: bvachha@mgh.harvard.edu; bvachha@partners.
org; pschaefer@mgh.harvard.edu; pschaefer@partners.org

# Springer Science+Business Media New York 2016 363

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_10
364
Introduction
Stroke is the fourth leading cause of death in the
United States and a leading cause of serious,
long-term adult disability [1]. Arterial
ischemic strokes account for 87 % of all
cases; the remaining strokes are mostly
hemorrhagic [2].
Readily available computed tomography
(CT) and magnetic resonance imaging (MRI) in
emergency settings have allowed clinicians to
improve correlations between clinical history
and precise anatomic lesion location in stroke
patients. One of the key tasks in establishing the
diagnosis of acute ischemic stroke and determin-
ing its appropriate treatment is establishing the
arterial territory affected. Knowledge of cerebral
vascular territories helps identify abnormal ves-
sels on CT and conventional angiograms, con-
firms that a DWI hyperintense lesion represents
an acute arterial stroke, and guides further inves-
tigations and treatment. Factors contributing to
understanding vascular territories include the
anatomy of the intracranial circulation and its
normal variants and the intrinsic variability in
the extent of brain that the main branches of the
intracranial circulation supply.
Fig. 1 Approximate vascular territories of the cerebral arteries and their branches as well as the posterior fossa vessels.
There is considerable variability in these territories particularly in the basal ganglia, brainstem, and posterior fossa
B.A. Vachha and P.W. Schaefer
This chapter reviews the normal anatomy of
intracranial arteries and describes the vascular
distribution as seen on CT and MRI using major
territorial strokes as a learning tool.
General Overview of Intracranial
Arteries
The intracranial circulation can be conveniently
divided into the anterior and posterior circulation.
The anterior circulation consists of the intradural
internal carotid artery (ICA) and its branches, as
well as the two terminal branches of the ICA – the
anterior cerebral artery (ACA) and the middle
cerebral artery (MCA). The anterior communicat-
ing artery (AComm) connecting the two ACAs
and the left and right posterior communicating
arteries (PComm) connecting the ipsilateral ICA
to the ipsilateral posterior communicating artery
(PCA) are also considered part of the anterior
circulation [3]. The posterior circulation includes
the vertebral arteries, the basilar artery and its
branches, and the terminal bifurcation of the bas-
ilar artery into the right and left PCAs [3]. Approx-
imate vascular territories of the major intracranial
arteries are depicted in Fig. 1.
17 Territorial Strokes as a Tool to Learn Vascular Territories
Fig. 2 Anatomic diagram
of the normal circle of
Willis
At the base of the brain, branches of the inter-
nal carotid arteries and the vertebrobasilar
trunk anastomose to form an arterial polygon
known as the circle of Willis [3]. The circle of
Willis is formed by two terminal ICAs, two
proximal or horizontal (A1) segments of the
ACA, the AComm, two PComms, two proximal
or horizontal (P1) segments of the PCAs, and the
top of the basilar artery (BA). The MCAs are not
considered part of the circle of Willis. The normal
circle of Willis is depicted diagrammatically in
Fig. 2.
The following sections discuss the
normal anatomy and vascular distributions of the
anterior and posterior circulations as seen on CT
and MRI.
365
Anterior Circulation
Intradural Internal Carotid Artery
The intracranial ICA is divided into six segments:
petrous (C2) segment, lacerum (C3) segment,
cavernous (C4), clinoid (C5) segment, ophthalmic
(C6) segment, and communicating (C7) segment
[4]. The ophthalmic (C6) segment is the first ICA
segment that lies within the subarachnoid space
and gives off two important branches: the oph-
thalmic artery which supplies the retina and the
optic nerve [4] and the superior hypophyseal
artery which supplies the anterior pituitary lobe,
the infundibular stalk, and the optic chiasm [5].
366
The communicating (C7) segment extends
from the PComm origin to the ICA terminus.
This segment has two important branches: the
PComm and the anterior choroidal artery.
Posterior Communicating Artery
The PComm joins the anterior and posterior cir-
culation. It extends from the ICA to the junction of
the P1 and P2 segments of the PCA. Perforating
arteries arise from the posterior PComm and sup-
ply the thalamus and hypothalamus [4].
The tuberothalamic artery or polar artery
arises from the caudal part of the PComm, which
is close to the PCA, or from the border between
the caudal and middle third of the PComm [6,
7]. It is absent in approximately one-third of the
normal population and is considered to have an
anatomically complementary relationship with
the paramedian artery originating from the P1
segment [7]. The tuberothalamic artery supplies
the ventral section of the thalamus including the
anterior thalamic nuclei (i.e., anteromedial,
anteroventral, anterodorsal nuclei), ventral ante-
rior nucleus, reticular nucleus, rostral portion of
the ventrolateral nucleus, ventral pole of the
medial dorsal nucleus, mammillothalamic tract,
ventral amygdalofugal pathway, and ventral por-
tion of the internal medullary lamina [6].
Infarction in the territory of the tuberothalamic
artery results in severe neuropsychological defi-
cits [7]. A characteristic feature of tuberothalamic
artery infarctions is the impairment of recent
memory, temporal disorientation, and new learn-
ing, which is more prominent with left-sided
lesions, while right-sided lesions result in visual
memory deficits, hemispatial neglect, and visual
spatial processing deficits [7, 6]. Language disor-
ders characterized by anomia, impaired fluency,
semantic and phonemic paraphasic errors, and
impaired comprehension are seen in left-sided
lesions. The “amnestic syndrome” is associated
with a disconnection between the anterior tha-
lamic nuclei and hippocampal formation, caused
by the disruption of the mammillothalamic tract,
as well as a disconnection between the amygdala
and anterior nuclei by disruption of the
amygdalothalamic projections passing through
the internal medullary lamina [8].
B.A. Vachha and P.W. Schaefer
Patients with infarctions confined to the ventral
anterior nucleus, anterior nuclei, paramedian
nuclei, mammillothalamic tract, and internal med-
ullary lamina manifest differently from those with
the previously described larger tuberothalamic
infarctions. In these patients, there is severe
impairment of memory and new learning with
relative sparing of language [6]. Palipsychism is
a phenomenon associated with anterior thalamic
infarcts in which patients demonstrate persevera-
tion which leads to overlap of sequential cognitive
processes [9].
Anterior Choroidal Artery
The anterior choroidal artery (AChA) arises from
the posterior wall of the ICA distal to the origin of
the PComm (although it occasionally arises from
the MCA) and courses laterally in the suprasellar
cistern to enter the choroidal fissure of the tempo-
ral horn of the lateral ventricle [3, 4].
Although the vascular territory of the AChA
shows large variations, the most consistently
documented area involves the amygdala, the lat-
eral geniculate body, posterior two-thirds of the
posterior limb of the internal capsule, most of the
globus pallidus, the origins of the optic radiations,
the middle one-third of the cerebral peduncles,
and the lateral thalamic border [3, 4, 10, 11]. The
AChA territory is reciprocal with those of the
posterolateral and posteromedial choroidal arter-
ies that arise from the PCA [4]. Figure 3 depicts an
infarct in the vascular territory supplied by the
AChA.
Interruption of blood flow from the AChA can
result in complete or partial manifestation of the
AChA syndrome which includes the triad of
hemiplegia (due to involvement of pyramidal
tracts in the posterior limb of the internal capsule),
hemisensory loss (due to involvement of the pos-
terolateral nucleus of the thalamus), and homon-
ymous hemianopia (due to involvement of the
lateral geniculate body) [10, 12, 13]. Many
AChA territory infarcts present as lacunar
syndromes [13].
Normal Variants of the ICA
Fetal origin of the posterior cerebral artery occurs
when the caliber of the PComm may be the same
17 Territorial Strokes as a Tool to Learn Vascular Territories
Fig. 3 Thirty-six-year-old
female with right
hemiparesis and word-
finding difficulties.
Diffusion-weighted MRI
(DWI) obtained 2 h after
onset of symptoms
demonstrated restricted
diffusion in the left mesial
temporal lobe (a) and in the
left posterior limb of the
internal capsule (b)
consistent with acute
infarction in the left anterior
choroidal artery distribution
as or larger than that of the ipsilateral P1 segment
of the posterior cerebral artery [14]. This variant
has clinical significance since simultaneous
infarcts in the vascular territories of both the ante-
rior and posterior circulation can occur from ICA
emboli in the presence of fetal circulation [15].
A “hyperplastic” AChA is defined as a normal
cerebrovascular variant in which the temporal-
occipital branches of the posterior cerebral artery
arise from the AChA [14]. The clinical signifi-
cance of this variant is the increased occurrence
of intracranial aneurysm formation [16].
In a 4-mm embryo, there are sites of anasto-
mosis between the paired dorsal aortic arches that
later form the internal carotid arteries and the
paired longitudinal neural arteries that later form
the vertebrobasilar system [17]. With the excep-
tion of the PComm, all other primitive arterial
connections regress when the definite circulation
develops. Failure of these vessels to regress
results in persistent carotid-vertebrobasilar anas-
tomoses. From cephalad to caudal, these are the
persistent trigeminal artery, persistent otic artery,
persistent hypoglossal artery, and proatlantal
intersegmental artery [14, 17].
The persistent trigeminal artery (PTA) is seen
in 0.1–0.2 % of patients and is the most common
of the carotid-vertebrobasilar anastomoses
[3]. Salas et al. classify the PTA into a medial
sphenoid variation which has an intrasellar or
367
trans-hypophyseal course and a lateral petrosal
variation in which the artery courses with the
sensory roots of the trigeminal nerve and exits
the Meckel cave below the petroclinoid ligament
[18]. Saltzman classifies the PTA into three types:
Type I is defined when the PTA supplies the distal
basilar artery, the proximal basilar artery is typi-
cally hypoplastic, and the ipsilateral PComm is
absent, Type II is defined when the PTA fills the
anterior superior cerebellar arteries only and the
posterior cerebral arteries are supplied by the
PComms, and Type III demonstrates a PTA that
unites with a remnant of the primitive paired lon-
gitudinal neural artery and supplies an ipsilateral
cerebellar artery (usually the anterior-inferior cer-
ebellar artery) [19]. Clinically, these anomalies
may be responsible for ischemia and trigeminal
neuralgia [20].
The persistent hypoglossal artery originates
from the ICA at the levels of the C1 through C3
vertebral bodies and courses through the hypo-
glossal canal to anastomose with the basilar artery
[3]. This variant is the second most common
carotid-vertebrobasilar anastomosis and has been
reported to result in glossopharyngeal neuralgia
and hypoglossal nerve paralysis [3, 14].
The proatlantal intersegmental artery arises
from the internal carotid artery (Type I) or the
external carotid artery (Type II) at the levels of
the C2 through C4 vertebral bodies and joins the
368
suboccipital vertebral artery before coursing
through the foramen magnum [3, 4].
The persistent otic artery is the least common
of the carotid-vertebrobasilar anastomoses. The
actual existence of this variant is controversial,
and it may represent overlapping vascular terri-
tories rather than persistence of an embryonic
vessel [4, 14].
Occlusions of the ICA may have clinical man-
ifestations related to embolism or low flow
(discussed in border zone infarctions) [15]. Epi-
sodes of transient monocular blindness due to
embolization to the retinal circulation are typical.
An ICA occlusion may be asymptomatic in the
presence of a competent circle of Willis. When a
thrombus propagates up to the top of ICA and
occludes both the MCA and the ACA, the occlu-
sion is often called a “T” occlusion [15]. It carries
a very poor prognosis unless complete recanaliza-
tion is achieved very early [21–23].
Anterior Cerebral Artery
The ACA is the smaller terminal branch of the
supraclinoid ICA. The ACA supplies the anterior
two-thirds of the medial surface of the cerebral
hemisphere extending from the frontal pole to the
parieto-occipital sulcus; this region includes the
motor and sensory cortical areas for the pelvis and
the lower extremities. The ACA vascular territory
includes the anterior four-fifths of the corpus
callosum, the anterior limb of the internal capsule,
the anterior-inferior head of the caudate nucleus,
and approximately 1 in. of the frontal and parietal
cortex of the cerebral convexity adjacent to the
interhemispheric fissure [3, 4]. Figure 4 demon-
strates an acute infarct with hemorrhagic conver-
sion in the expected vascular distribution of the
ACA. The ACA has three segments: horizontal
(A1), vertical (A2), and callosal (A3) segments.
Horizontal (A1) Segment
The horizontal/A1 segment extends from its ori-
gin to the midline where it communicates with the
contralateral ACA by the AComm. The ACA
B.A. Vachha and P.W. Schaefer
gives rise to the medial lenticulostriate arteries
which supply the medial basal ganglia and the
anterior limb of the internal capsule [3, 4]
(Fig. 1). The AComm has perforating arteries
that supply the anterior hypothalamus, optic chi-
asm, genu of the corpus callosum, cingulate
gyrus, and pillars of the fornix.
Vertical (A2) Segment
The vertical/A2 segment courses in the
interhemispheric fissure and gives rise to the
orbitofrontal and frontopolar arteries that sup-
ply the undersurface and the inferomedial aspect
of the frontal lobe [3, 4].
Callosal (A3) Segment
The callosal/A3 segment curves around the cor-
pus callosal genu and divides into the pericallosal
and callosomarginal arteries. The pericallosal
artery courses over the superior surface of the
corpus callosum in the pericallosal cistern and
gives off many small branches to the corpus
callosum [3]. The callosomarginal artery
courses within the cingulate sulcus over the cin-
gulate gyrus [3]. The cortical branches of the
pericallosal and callosal marginal arteries supply
the medial portions of the frontal lobes, superior
medial portions of the parietal lobes, and the ante-
rior corpus callosum [24]. Infarctions of these
vessels result in hemiparesis and hemianesthesia
of the contralateral leg due to involvement of the
medial precentral and postcentral gyri,
respectively.
Recurrent Artery of Heubner
The recurrent artery of Heubner arises from the
proximal A2 or the distal A1 segment [25, 26] and
supplies the anterior part of the caudate nucleus,
the anterior third of the putamen, the tip of the
outer segment of the globus pallidus, and the
inferior anterior limb of the internal capsule [25,
26]. Hemiparesis with faciobrachial predomi-
nance has been attributed to occlusions within
this artery [26]. Figure 5 depicts an acute infarct
in the vascular territory of the recurrent artery of
Heubner.
17 Territorial Strokes as a Tool to Learn Vascular Territories 369

Fig. 4 Sixty-year-old male

with weakness of the right
shoulder, leg, and foot.
Non-contrast-enhanced CT
of the brain (a) obtained 6 h
after onset of symptoms
demonstrates
hypoattenuation with
superimposed hyperdensity
consistent with
hemorrhagic conversion of
acute infarct in the left
anterior cerebral artery
distribution. Diffusion-
weighted (b) and FLAIR (c)
images obtained 1 h later in
the same patient show the
full extent of acute left
anterior cerebral artery
infarct

Normal Variants of the Anterior Cerebral
Arteries
The azygos anterior cerebral artery is a rare vari-
ant of the anterior cerebral artery involving a
common trunk in the A2 segment (above the
anterior communicating artery) with a prevalence
of 0.2–4.0 % [14, 27]. Baptista distinguished three
types of this anatomical variant: (1) “unpaired”
ACA or true azygos artery (Type I anomaly) in
which a single unpaired ACA provides branches
to both cerebral hemispheres; (2) “bihemispheric”
ACA (Type II anomaly) in which the A2 segment
of one ACA sends branches across the midline,
while the contralateral A2 segment either is hypo-
plastic or terminates early in its course toward the
genu of the corpus callosum; and (3) “accessory”
ACA (Type III anomaly) where there is an addi-
tional vessel arising from the anterior communi-
cating artery accompanied by two hypoplastic A2
segments [27]. The most common anomaly noted
by Baptista was the Type III anomaly, while the
Type I anomaly was observed in only 1 of the
381 brain specimens examined by him [27]. An
azygos anterior cerebral artery is associated with a
large number of cerebral anomalies including
agenesis of corpus callosum, porencephalic
cysts, hydranencephaly, lobular
holoprosencephaly, septo-optic dysplasia, and
saccular aneurysms. Being aware of the presence
of an azygos ACA is important, as an occlusion of
this vessel can result in infarcts involving the
bilateral medial cerebral hemispheres [14].
370
Fig. 5 Fifty-five-year-old male with history of atrial fibril-
lation presented with new onset seizures. Diffusion-
weighted MRI obtained 3 h after onset of symptoms dem-
onstrates restricted diffusion in the caudate nucleus, ante-
rior limb of the internal capsule, and anterior aspect of the
putamen and globus pallidus consistent with acute infarct
in the expected vascular distribution of the recurrent artery
of Heubner. There is an additional small acute infarct in the
right periatrial white matter
Trifurcation of the anterior cerebral artery,
defined as the presence of three A2 segments of
the ACA, is seen in 2–13 % cases and most likely
represents persistence of the median callosal
artery [14]. In the presence of thromboembolic
disease, hypoplasia or absence of the A1 segment
of the ACA typically results in a decreased collat-
eral supply and consequently an increased risk of
infarction [14].
Clinical manifestations of unilateral ACA ter-
ritory infarction are spastic hemiparesis of the
contralateral lower limb and to a lesser degree
paresis of the arm, with the face and tongue
largely spared [26, 28, 29]. Sensory dysfunction,
particularly proprioceptive and discriminative
impairment of the contralateral lower limb, may
be present but is usually mild. Other manifesta-
tions of unilateral ACA territory infarction
include speech disturbance with initial mutism,
B.A. Vachha and P.W. Schaefer
transcortical aphasia, apraxia, apathy, abulia,
incontinence, and disinhibition [26]. Bilateral
ACA territory infarction causes a neurologic syn-
drome characterized by profound akinetic mut-
ism, paraparesis, and poor recovery [26].
Middle Cerebral Artery
The MCA has the largest vascular territory of all
of the major intracranial arteries and is the vessel
most commonly affected by cerebrovascular acci-
dents [30]. Cortical branches of the MCA supply
two-thirds of the lateral surface of the cerebral
hemisphere with the exception of a thin 1-in.
strip near the vertex that is supplied by the ACA
and the occipital and posteroinferior parietal lobes
that are supplied by the PCA (Fig. 1). Infarcts that
occur in different segments of the MCA lead to
diverse neurologic outcomes. The following sub-
sections detail the specific MCA territories sup-
plied by the various segments of the MCA and
their correlation to specific neurologic deficits.
Horizontal (M1) Segment
The horizontal/M1 segment of the MCA arises
from the carotid terminus and extends laterally
where it bifurcates or trifurcates just before it
enters the Sylvian fissure. The anterior temporal
artery arises from the M1 segment and supplies
the tip of the temporal lobe [4]. Lateral
lenticulostriate arteries are deep penetrating
branches of the M1 segment that supply the supe-
rior part of the head and body of the caudate
nucleus, the putamen, and the external capsule
[4, 31].
Complete proximal M1 occlusions (Fig. 6)
result in sensory, motor, language, and executive
function deficits due to disruption of both MCA
cortical representations and basal ganglia circuit
structures, while distal MCA occlusions spare the
lateral lenticulostriate vascular territory but
involve the distributions of both the superior and
inferior division branches (see below).
Insular (M2) Segments
In the majority of cases, the MCA divides into two
trunks – superior and inferior – although
17 Territorial Strokes as a Tool to Learn Vascular Territories
Fig. 6 Sixty-seven-year-old female with left hemiplegia
last seen well 9 h prior to presentation in the emergency
room. Diffusion-weighted MRI (DWI) demonstrated
restricted diffusion within the right frontal and temporal
lobe consistent with acute infarct in the vascular distribu-
tion of the superior and inferior divisions of the right
middle cerebral artery
occasionally it trifurcates into superior, middle,
and inferior divisions [3, 4, 30, 31]. The superior
division supplies the frontal and peri-Rolandic
regions including the anterior parietal lobe
[30]. The inferior division supplies the lateral
temporal and inferior parietal lobes [30]. Figures 7
and 8 demonstrate acute infarcts in the vascular
territories of the superior and inferior divisions of
the MCA respectively.
Opercular (M3) Segments
The portions of the superior and inferior divisions
that course laterally under the frontal, parietal, and
temporal opercula are termed the opercular/M3
segments [30, 31].
Cortical (M4) Segments
The cortical/M4 segments are the portions of the
superior and inferior MCA divisions that exit the
Sylvian fissure and branch over the lateral surface
of the hemisphere [3, 4].
371
Fig. 7 Sixty-six-year-old male with right facial droop and
expressive aphasia. Diffusion-weighted MRI (DWI) of the
brain obtained 4 h after symptom onset shows an acute
infarct involving the left frontal operculum and left anterior
insula in the expected vascular distribution of the superior
division of the left middle cerebral artery
Normal Variants of the Middle Cerebral
Arteries
An early bifurcation or trifurcation of the M1
segment close to its origin at the ICA is a common
finding and is of no definite clinical significance
[14]. A middle cerebral artery branch arising from
the distal internal carotid artery is called a “dupli-
cated” MCA, while a vessel arising from the ACA
and coursing parallel to the M1 segment of the
MCA is called an “accessory MCA” [14]. The
frequency of accessory MCA is 0.3–0.4 %,
while that of MCA duplication is 0.2–2.9 %
[14]. The accessory MCA supplies the territory
of the orbitofrontal, prefrontal, precentral, and/or
central arteries [32]. The duplicated MCA sup-
plies the territory of the lateral part of the orbital
surface of the frontal lobe and the anterior and/or
middle temporal territories. In the setting of MCA
infarcts, the accessory MCA can provide collat-
eral flow to the anterior frontal lobe, but it cannot
372
Fig. 8 Eighty-nine-year-old female with new onset recep-
tive aphasia. Diffusion-weighted MRI (DWI) of the brain
obtained 12 h after symptom onset shows an acute infarct
involving the left temporal lobe in the expected vascular
distribution of the inferior division of the left middle cere-
bral artery
supply enough flow to the main MCA territory
[32]. Similarly, the duplicated MCA can provide
collateral flow to the anterior temporal lobe, but it
cannot supply enough flow to the main MCA
territory [32].
Clinical patterns of MCA territory infarcts
depend on the location, size, and side of the infarct
[30]. If the entire MCA is occluded at its origin,
clinical findings are contralateral hemiplegia,
hemianesthesia, homonymous hemianopia, and
conjugate deviation of the eyes toward the side
of the infarct in the acute phase [30, 33]. When the
speech/language-dominant hemisphere is
affected, severe global aphasia is present; when
the nondominant hemisphere is affected, the
patient presents with anosognosia, constructional
apraxia, and hemineglect [30]. A variant of MCA
stem infarction is malignant infarction, a term
used to describe those cases with subsequent
extensive brain swelling and herniation
(Fig. 9) [22].
B.A. Vachha and P.W. Schaefer
More distal superior division branch occlu-
sions may produce a clinical syndrome of contra-
lateral hemiparesis and hemianesthesia
affecting the lower face and upper extremity
more than the leg and contralateral visual
field deficits predominantly affecting the lower
fields [30]. For superior division infarcts
involving the dominant hemisphere, global apha-
sia with mutism (aphemia) is seen initially with
eventual progression to nonfluent (expressive)
aphasia [30], while nondominant hemisphere
infarcts are associated with an acute confusional
state and a variable degree of hemispatial
neglect [34].
Inferior division infarcts result in visual field
defects (contralateral homonymous
hemianopia/quadrantanopia) with relative
sparing of sensorimotor involvement [30].
Inferior division branch occlusion of the
MCA supplying the dominant superior
temporal lobe leads to fluent (receptive) aphasia
without weakness, while infarcts of the
nondominant hemisphere superior temporal lobe
lead to behavioral disturbances and impairment
of visuospatial skills [30]. Sensory aprosody,
probably as a result of damage to the right
posterosuperior temporal lobe, is a common find-
ing in nondominant inferior division territory
infarcts [35].
Posterior Circulation
Intradural Vertebral Artery
The intradural segment of the vertebral artery/V4
segment gives rise to the anterior and posterior
spinal arteries, the medullary perforating
branches, and the posterior inferior cerebellar
artery (PICA). Occlusions of the anterior spinal
artery or vertebral (V4) artery result in infarcts of
the medial aspect of the medulla (Fig. 10). The
classic triad of the medial medullary syndrome
(Déjerine-Roussy syndrome) includes contralat-
eral hemiplegia sparing the face, contralateral
loss of posterior column sensation, and ipsilateral
tongue paralysis due to involvement of the
17 Territorial Strokes as a Tool to Learn Vascular Territories 373

Fig. 9 Seventy-one-year-
old man with right
hemiplegia and receptive
aphasia. (a) Diffusion-
weighted MRI (DWI)
obtained 3 h after onset of
symptoms demonstrates an
acute infarct in the territory
of the superior and inferior
divisions of the left middle
cerebral artery.
Non-contrast axial (b) and
coronal (c) head CT
obtained 4 days later for
progressive worsening
mental status in the same
patient demonstrate right
midline shift with
subfalcine and uncal
herniation consistent with
malignant edema

Fig. 10 Fifty-five-year-old
male with acute onset of
nystagmus, right tongue
deviation, left-sided ataxia,
and sensory impairment.
Diffusion-weighted MRI
(DWI) obtained 6 h after
symptom onset
demonstrates restricted
diffusion in the right medial
medulla consistent with an
acute infarct in the expected
vascular distribution of the
intradural vertebral artery

corticospinal tract at the medullary pyramid, the
medial lemniscus, and the hypoglossal nucleus or
nerve [36, 37]. Frequently, however, the clinical
syndrome is not confined to the triad and is more
heterogeneous [38].
Posterior Inferior Cerebellar Artery
The PICA has a variable vascular territory
depending on the size of the anterior-inferior cer-
ebellar artery (AICA). PICA territory infarcts
involve the lateral medulla (Fig. 11a) as well as
374 B.A. Vachha and P.W. Schaefer

Fig. 11 (a) Thirty-seven-

year-old female with
Wallenberg syndrome.
Diffusion-weighted MRI
(DWI) obtained 3 h after
onset of symptoms
demonstrates restricted
diffusion in the left lateral
medulla consistent with
acute infarct. (b) Thirty-
three-year-old female with
acute onset of vomiting and
gait unsteadiness. DWI
demonstrates restricted
diffusion consistent with
acute infarct in the inferior
left cerebellar hemisphere.
Both case (a) and case (b)
represent acute infarcts in
the vascular territories of
the posterior inferior
cerebellar artery

the cerebellar tonsil and inferior cerebellar hemi-
sphere (Fig. 11b). Lateral medullary infarcts due
to occlusion of the PICA result in the lateral
medullary (Wallenberg) syndrome character-
ized by ipsilateral limb ataxia, ipsilateral loss of
pain and temperature sensation in the face, loss of
contralateral pain and temperature sensation in the
body, vomiting, vertigo, nystagmus away from
the lesion side, ipsilateral Horner syndrome, dys-
phagia, and hiccups [38, 39].
Occlusions of the PICA result in infarcts of the
inferior cerebellar hemisphere characterized by
vertigo, gait ataxia, limb dysmetria, dysarthria,
nausea, and vomiting [40]. If the infarct is large,
there may be mass effect on the brainstem and
hydrocephalus.
Normal Variants of the Vertebral Artery
Normal variations in the distal vertebral artery
have been described including C1 and C2 origins
of the PCIA, duplication of the distal vertebral
arteries, and an aberrant course of the distal ver-
tebral artery [41]. Knowledge of these variants is
important for both endovascular and surgical
planning.
Basilar Artery
The basilar artery is formed from the union of the
two vertebral arteries near the pontomedullary
junction. The basilar artery courses within the
prepontine cistern and bifurcates within the
interpeduncular fossa into the right and left pos-
terior cerebral arteries.
The basilar artery gives rise to several cerebel-
lar branches; occlusions in these vessels result in
territorial infarctions that are discussed below.
Superior Cerebellar Artery
The superior cerebellar arteries (SCA) arise from
the distal basilar artery and supply the superior
17 Territorial Strokes as a Tool to Learn Vascular Territories
Fig. 12 Fifty-seven-year-old male with endocarditis
presenting with right-sided ataxia, dysarthria, nystagmus,
nausea, and vomiting. Diffusion-weighted MRI (DWI)
obtained 3 h after onset of symptoms demonstrates
restricted diffusion in the right superior cerebellar hemi-
sphere consistent with an acute superior cerebellar artery
surface of the cerebellar hemispheres, the superior
vermis, and the dentate nuclei (Figs. 1 and 12a).
SCA occlusion causes ipsilateral ataxia, dysar-
thria, contralateral pain and temperature sensory
impairment, nystagmus, nausea, and vomiting
[40]. Horner syndrome may be caused by involve-
ment of the oculosympathetic fibers [40]. With
large infarcts, mass effect may lead to compres-
sion of the midbrain or cause hydrocephalus.
Sleep disorders due to lesions involving the
locus ceruleus have been documented [40].
Basilar Perforating Arteries
Basilar perforating arteries arise from the dorsal
surface of the midbrain to supply the pons and the
midbrain.
Anterior-Inferior Cerebellar Artery
The AICA supplies CN VII and CN VIII as well
as a narrow strip of the cerebellar hemisphere that
lies directly behind the petrous temporal bone
(Figs. 1 and 12b).
Normal Variants of the Basilar Artery
Basilar artery fenestration is noted in 0.6 % of
cases and is most commonly located close to the
vertebrobasilar junction [14].
375
infarct (a) and more inferiorly in the anterior right cerebel-
lar hemisphere posterior to the temporal bone consistent
with acute right anterior-inferior cerebellar artery infarct
(b). Foci of restricted diffusion are also noted within the
right brainstem and right occipital lobe (a) and in the left
cerebellum (b)
Infarctions of the basilar artery territory result
in a wide range of neurologic deficits. Complete
basilar artery occlusion leads to infarction of the
pons, midbrain, and thalamus. The medial por-
tions of the temporal, inferior parietal, and occip-
ital lobes also undergo infarction if the PCAs do
not receive sufficient collateralization from the
posterior communicating arteries. A combination
of dysarthria, pupillary disorders, lower cranial
nerve involvement, or reduced consciousness at
initial admission appears to be strongly associated
with severe disability or death in the absence of
reperfusion [42].
A constellation of clinical syndromes caused
by involvement of brainstem nuclei and ascending
or descending long tracts can be observed with
incomplete basilar artery occlusion and more
focal infarction. Several named syndromes are
commonly associated with basilar artery occlu-
sions. Occlusions of the proximal and middle
segments of the basilar artery result in infarction
of the basis pontis with sparing of the tegmen of
the pons (Fig. 13), manifesting as the locked-in
syndrome (quadriplegia with spared level of con-
sciousness, preserved eye movements, and
blinking) [43]. Top of the basilar syndrome
(characterized by visual, oculomotor, and
376
Fig. 13 Seventy-five-year-old male with acute onset of
right arm weakness and dysarthria. Diffusion-weighted
MRI (DWI) obtained 3 h after onset of symptoms demon-
strates a wedge-shaped acute infarct involving the left
paramedian basis pontis with sparing of the tegmen in the
expected vascular distribution of the branches of the basilar
artery
behavioral abnormalities, often without signifi-
cant motor dysfunction) occurs when there is
thromboembolic occlusion of the distal basilar
artery with resultant infarction of the rostral
brainstem and/or the posterior cerebral artery ter-
ritory [44]. Infarctions of the dorsolateral pons
usually result from occlusion of the AICA (caudal
pons) or SCA (rostral pons) and cause the lateral
pontine syndrome (ipsilateral limb ataxia, ipsi-
lateral loss of pain and temperature sensation of
the face, contralateral loss of pain and temperature
sensation of the body, ipsilateral Horner syn-
drome, and ipsilateral facial paralysis) [45].
The medial pontine syndrome (contralateral
spastic hemiparesis of the body, contralateral
loss of position and vibration of the body, and
medial strabismus) results from occlusion of the
paramedian perforating branches with infarction
of the medial pons [45].
Posterior Cerebral Artery
The PCAs are paired vessels arising from the
basilar artery. The PCAs supply parts of the
B.A. Vachha and P.W. Schaefer
midbrain, subthalamic nucleus, and thalamus.
The vascular territory of the PCA includes the
posterior two-thirds of the medial temporal lobe,
nearly the entire occipital lobe and the medial
inferior parietal lobes, as well as the posterior
one-fifth of the corpus callosum (Fig. 1). The
PComm, arising from the ICA, extends posteri-
orly to join the PCA at the junction of the P1 and
P2 segments, thereby completing the circle of
Willis and providing an important source of col-
lateral circulation for the MCA territory. Infarcts
in the region of the PCA result in homonymous
hemianopia of the contralateral visual field with
macular sparing. Proximal PCA strokes result in
infarcts in the thalamus and/or midbrain as well as
in the cortex, while distal PCA strokes involve
only cortical structures. The following subsec-
tions describe four distinct PCA segments and
their branches.
Precommunicating (P1) Segment
The precommunicating/P1 segment of the PCA
extends laterally from the basilar artery to the
junction with the PComm. Branches arising from
the P1 segment of the PCA include thalamoper-
forating or paramedian arteries to the thalamus
and brainstem branches [3, 4].
The paramedian arteries have great variability
in terms of the territorial contribution to the thal-
amus but principally supply the dorsomedial
nucleus and internal medullary lamina and
intralaminar nuclei (central lateral, centromedian,
and parafascicular nuclei) [6]. When the
tuberothalamic artery is absent, the paramedian
arteries can supply the anterior thalamic areas as
well [6, 7]. The artery of Percheron is an
anatomic variant of the paramedian artery in
which a single dominant paramedian artery sup-
plies the bilateral medial thalami with variable
contribution to the midbrain [4]. Specific ischemic
patterns and syndromes that result from occlusion
of the artery of Percheron are discussed in detail in
a later section.
Unilateral thalamic infarction in the territory
of the paramedian artery is characterized by
deficits in arousal and memory. In the early
17 Territorial Strokes as a Tool to Learn Vascular Territories 377

stages, patients demonstrate decreased and fluctu-

ating levels of consciousness, while impaired
arousal with confusion, agitation, and apathy are
seen in later stages of the infarct [7]. Left-sided
lesions result in language and speech impairments
characterized by hypophonia and dysprosody
with reduced verbal fluency and perseveration.
Syntactic structure is preserved with occasional
paraphasic errors and normal repetition [6]. Right-
sided infarcts result in visuospatial deficits.
Bilateral paramedian thalamic strokes are typ-
ically characterized by altered mental status,
memory impairment, and vertical gaze palsy
[6, 46]. Altered mental status may present on a
spectrum ranging from disorientation and
hypersomnolence to deep coma and akinetic mut-
ism [6]. Anterograde and retrograde memory Fig. 14 Fifty-year-old male presenting with
impairment with confabulation tends to resolve prosopagnosia. Diffusion-weighted MRI (DWI) obtained
with time but may be severe and persistent. In 15 h after symptom onset demonstrates restricted diffusion
in the right occipital and medial temporal lobes consistent
the late stages, the patient may demonstrate emo- with acute right posterior cerebral artery territory infarct
tional blunting, loss of initiative, and inappropri-
ate social behaviors.
Occlusion of the P1 segment of the PCA that The medial branch supplies the external half of
supplies the midbrain can result in a CN III palsy the medial geniculate nucleus. The principal
or contralateral ataxia or hemiplegia [47]. inferolateral arteries supply the major part of the
ventral posterior nuclei (lateral, medial, and infe-
Ambient (P2) Segment rior nuclei) as well as the ventral and lateral parts
The medial and lateral posterior choroidal of the ventrolateral nucleus. The inferolateral
arteries arise from the P2 segment. The lateral pulvinar arteries supply the rostral and lateral
posterior choroidal artery shares a reciprocal rela- parts of the pulvinar and lateral dorsal nucleus [6].
tionship with the anterior choroidal artery (which The Déjerine-Roussy syndrome or thalamic
arises from the ICA) [4]. pain syndrome occurs following infarctions
The anterior and posterior temporal arter- related to the inferolateral arteries and is charac-
ies supply the inferior surface of the temporal terized by contralateral sensory loss followed by
lobe, while the thalamogeniculate and pedun- agonizing burning pain [6, 47, 48]. This syndrome
cular perforating arteries supply the midbrain is usually associated with infarcts of the right
[4]. Occlusion of the P2 segment causes visual thalamus.
field defects and more complex visual changes
such as prosopagnosia secondary to infarction of Quadrigeminal (P3) Segment
the medial temporal and occipital lobes The quadrigeminal/P3 segment is a short segment
(Fig. 14) [47]. without any significant named branches [4].
The inferolateral arteries comprise five to ten
arteries that arise from the P2 segment of the PCA Calcarine (P4) Segment
and supply the thalamus [6]. There are three main The calcarine/P4 segment divides into the medial
groups: the medial geniculate, principal and lateral trunks within the calcarine fissure.
inferolateral, and inferolateral pulvinar arteries. The medial trunk gives off medial occipital,
378
Fig. 15 Sixty-nine-year-
old male with waxing and
waning mental status
progressively worsening
over the course of 3 days.
Diffusion-weighted MRI
demonstrates foci of
restricted diffusion within
the medial aspects of the
right and left thalami (a)
and medial midbrain (b)
consistent with an acute
artery of Percheron infarct
parieto-occipital, calcarine, and posterior
splenial arteries which supply the regions of the
brain represented in their names [4]. The lateral
trunk gives rise to the lateral occipital artery [4].
Artery of Percheron
The central artery of Percheron is a rare but clin-
ically important PCA variant. This occurs when
the bilateral, medial thalamic/rostral midbrain per-
forators arise from a single trunk from one P1
segment [4]. Occlusion may result in four distinct
patterns of infarction in descending order of fre-
quency: bilateral, paramedian, thalamic, and ros-
tral midbrain infarction (Fig. 15); bilateral,
paramedian, and thalamic without midbrain
involvement; bilateral, paramedian, and anterior
thalamic with midbrain involvement; and bilat-
eral, paramedian, and anterior thalamic without
midbrain involvement [46]. In addition to these
four distinct patterns, Lazzaro et al. described a
V-shaped hyperintense signal abnormality on
axial FLAIR images along the pial surface of the
midbrain in the interpeduncular fossa in their
series of 37 patients with occlusions of the artery
of Percheron [46]. Infarctions associated with
artery of Percheron occlusions produce long-
lasting memory dysfunction, deficits in executive
function, and mood changes [49].
Normal Variants of the Posterior Cerebral
Arteries
Fetal origin of the posterior cerebral artery has
been described under normal variants of the
ICA. Fenestration of the P1 and P2 segments of
B.A. Vachha and P.W. Schaefer
the PCA has been noted but is a very rare
occurrence [14].
Border Zones
Border zones occur at the junction between two
vascular territories. Two distinct border zone or
watershed territories are defined: (1) External
(cortical) border zone territories include the
frontal and parietal border zone territory between
the ACA and MCA territories and the posterior
parieto-occipital border zone territory between the
MCA and PCA territories. (2) Internal (subcor-
tical) border zone territories include the border
zones in the corona radiata and centrum semiovale
between the penetrating branches (lenticulostriate
and medullary perforating and anterior choroidal
arteries) and the main cerebral arteries (ACA,
MCA, and PCA) [4]. Figure 16 is an example of
diffusion-weighted MRI sequences demonstrat-
ing restricted diffusion consistent with border
zone infarcts.
Clinical manifestations of patients with border
zone infarcts depend upon the location of the
lesion. Patients with unilateral precentral cortico-
subcortical infarcts present with somnolence, bra-
chial monoparesis, disproportionate motor
hemiparesis with face mostly spared, transcortical
motor aphasia, and focal myoclonic jerks
[50]. Unilateral, postcentral cortico-subcortical
infarcts result in sensorimotor hemiparesis with
stereoagnosia, cortical hemihypesthesia, visual
and tactile hemispatial neglect, anosognosia,
17 Territorial Strokes as a Tool to Learn Vascular Territories
Fig. 16 Eighty-seven-year-old man with atrial fibrillation
with new onset right hemiplegia. Diffusion-weighted MRI
(DWI) obtained 4 h after symptom onset demonstrates
bilateral border zone infarcts between the anterior cerebral
artery and middle cerebral artery, presumed to be
cardioembolic in origin
hemianopia particularly involving the lower
quadrant, Wernicke-like aphasia, and focal myo-
clonic jerks [50]. Patients with deep border zone
infarcts in the dominant hemisphere develop
global aphasia, nonfluent aphasia with or without
comprehension and/or repetition deficits, ideomo-
tor apraxia, and acalculia. Patients with internal
border zone infarctions in the nondominant hemi-
sphere demonstrate anosognosia, neglect, or
dysprosody depending on the extent of the lesion.
Bilateral infarcts in the border zone regions
between the MCA and PCA can result in
asimultagnosia (difficulty recognizing several
objects at the same time), optical ataxia (difficulty
coordinating hand and eye movements), and
apraxia of gaze [50]. Bilateral infarcts in the bor-
der zone regions between the ACA and MCA can
result in selective weakness of the upper and
lower extremities with sparing of the face,
hands, and feet referred to as “a man in a barrel”
distribution [50].
379
Summary
Neuroimaging plays a key role in the initial
workup of stroke patients. Knowledge of arterial
territories and their associated stroke syndromes is
needed to achieve accurate localization of ische-
mic lesions to guide further management.
References
1. Go AS, Mozaffarian D, Roger VL, Benjamin EJ, Berry
JD, Blaha MJ, Dai S, Ford ES, Fox CS, Franco S,
Fullerton HJ, Gillespie C, Hailpern SM, Heit JA, How-
ard VJ, Huffman MD, Judd SE, Kissela BM, Kittner SJ,
Lackland DT, Lichtman JH, Lisabeth LD, Mackey RH,
Magid DJ, Marcus GM, Marelli A, Matchar DB,
McGuire DK, Mohler ER 3rd, Moy CS, Mussolino
ME, Neumar RW, Nichol G, Pandey DK, Paynter NP,
Reeves MJ, Sorlie PD, Stein J, Towfighi A, Turan TN,
Virani SS, Wong ND, Woo D, Turner MB (2014) Heart
disease and stroke statistics – 2014 update: a report from
the american heart association. Circulation 129(3):
e28–e292. doi:10.1161/01.cir.0000441139.02102.80,
01.cir.0000441139.02102.80 [pii]
2. Grysiewicz RA, Thomas K, Pandey DK (2008) Epide-
miology of ischemic and hemorrhagic stroke: inci-
dence, prevalence, mortality, and risk factors. Neurol
Clin 26(4):871–895. doi:10.1016/j.ncl.2008.07.003,
S0733-8619(08)00104-7 [pii]
3. Osborn A (1999) Diagnostic cerebral angiography.
Lippincott Williams & Wilkins, Philadelphia
4. Osborn A (2013) Osborn’s brain: imaging, pathology,
and anatomy. Amirsys, Manitoba
5. Joo W, Funaki T, Yoshioka F, Rhoton AL Jr (2012)
Microsurgical anatomy of the carotid cave. Neurosur-
gery 70(2 Suppl Operative):300–311. doi:10.1227/
NEU.0b013e3182431767; discussion 311–302
6. Schmahmann JD (2003) Vascular syndromes of the
thalamus. Stroke 34(9):2264–2278. doi:10.1161/01.
STR.0000087786.38997.9E, 01.STR.0000087786.
38997.9E [pii]
7. Bogousslavsky J, Regli F, Assal G (1986) The syn-
drome of unilateral tuberothalamic artery territory
infarction. Stroke 17(3):434–441
8. von Cramon DY, Hebel N, Schuri U (1985) A contri-
bution to the anatomical basis of thalamic amnesia.
Brain 108(Pt 4):993–1008
9. Ghika-Schmid F, Bogousslavsky J (2000) The acute
behavioral syndrome of anterior thalamic infarction: a
prospective study of 12 cases. Ann Neurol 48(2):
220–227
10. Hupperts RM, Lodder J, Heuts-van Raak EP, Kessels F
(1994) Infarcts in the anterior choroidal artery territory.
Anatomical distribution, clinical syndromes, presumed
pathogenesis and early outcome. Brain 117(Pt 4):
825–834
380
11. Ois A, Cuadrado-Godia E, Solano A, Perich-Alsina X,
Roquer J (2009) Acute ischemic stroke in anterior
choroidal artery territory. J Neurol Sci 281
(1–2):80–84. doi:10.1016/j.jns.2009.02.323, S0022-
510X(09)00055-0 [pii]
12. Foix C, Chavany J, Hillemand P, Schiff-Wertheimer S
(1925) Obliteration de l’arte’re choroidienne
ante’rieure: ramollissement de son territoire ce’re’bral:
he’miple’gie, he’mianesthe’sie, he’mianopsie. Bull
Soc Ophtalmol Paris 37:221–223
13. Helgason C, Caplan L (2012) Anterior choroidal artery
territory strokes. In: Caplan L, Van Gijn J (eds) Stroke
syndromes, 3rd edn. Cambridge University Press,
Cambridge, UK, pp 375–386
14. Dimmick SJ, Faulder KC (2009) Normal variants of
the cerebral circulation at multidetector CT angiogra-
phy. Radiographics 29(4):1027–1043. doi:10.1148/
rg.294085730, 29/4/1027 [pii]
15. Schwarz S, Schwab S, Hacke W (2002) Large and
panhemispheric infarcts. In: Bogousslavsky J, Caplan
L (eds) Stroke syndromes, 2nd edn. Cambridge Uni-
versity Press, Cambridge, UK, pp 490–498
16. Hyun DK, Shim YS, Park HS, Oh SY (2014) Throm-
boembolic complication following neurointervention
in ruptured anomalous hyperplastic anterior choroidal
artery aneurysm. Neuroradiol J 27(1):103–107, doi:IN.
v27.i1.p103 [pii]
17. Caldemeyer KS, Carrico JB, Mathews VP (1998) The
radiology and embryology of anomalous arteries of the
head and neck. AJR Am J Roentgenol 170(1):
197–203. doi:10.2214/ajr.170.1.9423632
18. Salas E, Ziyal IM, Sekhar LN, Wright DC (1998)
Persistent trigeminal artery: an anatomic study. Neuro-
surgery 43(3):557–561; discussion 561–552
19. Saltzman GF (1959) Patent primitive trigeminal artery
studied by cerebral angiography. Acta Radiol 51(5):
329–336
20. Morita A, Fukushima T, Miyazaki S, Shimizu T,
Atsuchi M (1989) Tic douloureux caused by primitive
trigeminal artery or its variant. J Neurosurg 70(3):
415–419. doi:10.3171/jns.1989.70.3.0415
21. Wijdicks EF, Diringer MN (1998) Middle cerebral
artery territory infarction and early brain swelling: pro-
gression and effect of age on outcome. Mayo Clin Proc
73(9):829–836. doi:10.4065/73.9.829, S0025-6196
(11)63788-9 [pii]
22. Hacke W, Schwab S, Horn M, Spranger M, De
Georgia M, von Kummer R (1996) ‘Malignant’ middle
cerebral artery territory infarction: clinical course and
prognostic signs. Arch Neurol 53(4):309–315
23. Krieger DW, Demchuk AM, Kasner SE, Jauss M,
Hantson L (1999) Early clinical and radiological pre-
dictors of fatal brain swelling in ischemic stroke.
Stroke 30(2):287–292
24. Bogousslavsky J, Caplan L (eds) (2002) Stroke syn-
dromes. Cambridge University Press, New York
25. Zunon-Kipre Y, Peltier J, Haidara A, Havet E,
Kakou M, Le Gars D (2012) Microsurgical anatomy
B.A. Vachha and P.W. Schaefer
of distal medial striate artery (recurrent artery of
Heubner). Surg Radiol Anat 34(1):15–20.
doi:10.1007/s00276-011-0888-5
26. Brust J (2012) Anterior cerebral artery. In: Caplan L,
Van Gijn J (eds) Stroke syndromes, 3rd edn. Cam-
bridge University Press, Cambridge, UK, pp 364–374
27. Baptista AG (1963) Studies on the arteries of the brain.
Ii. The anterior cerebral artery: some anatomic features
and their clinical implications. Neurology 13:825–835
28. Bogousslavsky J, Martin R, Moulin T (1992)
Homolateral ataxia and crural paresis: a syndrome of
anterior cerebral artery territory infarction. J Neurol
Neurosurg Psychiatry 55(12):1146–1149
29. Hiraga A, Uzawa A, Tanaka S, Ogawara K,
Kamitsukasa I (2009) Pure monoparesis of the leg
due to cerebral infarctions: a diffusion-weighted imag-
ing study. J Clin Neurosci 16(11):1414–1416.
doi:10.1016/j.jocn.2009.01.015, S0967-5868(09)
00156-8 [pii]
30. Mohr J, Kejda-Scharler J (2012) Middle cerebral artery
territory syndromes. In: Caplan L, Van Gijn J (eds)
Stroke syndromes, 3rd edn. Cambridge University
Press, Cambridge, UK, pp 344–363
31. Gibo H, Carver CC, Rhoton AL Jr, Lenkey C, Mitchell
RJ (1981) Microsurgical anatomy of the middle cere-
bral artery. J Neurosurg 54(2):151–169. doi:10.3171/
jns.1981.54.2.0151
32. Komiyama M, Nakajima H, Nishikawa M, Yasui T
(1998) Middle cerebral artery variations: duplicated
and accessory arteries. AJNR Am J Neuroradiol
19(1):45–49
33. Mohr JP, Foulkes MA, Polis AT, Hier DB, Kase CS,
Price TR, Tatemichi TK, Wolf PA (1993) Infarct topog-
raphy and hemiparesis profiles with cerebral convexity
infarction: the Stroke Data Bank. J Neurol Neurosurg
Psychiatry 56(4):344–351
34. Mori E, Yamadori A (1987) Acute confusional state
and acute agitated delirium. Occurrence after infarction
in the right middle cerebral artery territory. Arch
Neurol 44(11):1139–1143
35. Darby DG (1993) Sensory aprosodia: a clinical clue to
lesions of the inferior division of the right middle
cerebral artery? Neurology 43(3 Pt 1):567–572
36. Kim JS, Han YS (2009) Medial medullary
infarction: clinical, imaging, and outcome study in 86 con-
secutive patients. Stroke 40(10):3221–3225. doi:10.
1161/STROKEAHA.109.559864, STROKEAHA.109.
559864 [pii]
37. Dejerine J (1914) Semiologie des affections du
systeme nerveux. Masson, Paris
38. Kim J (2012) Medullary infarcts and hemorrhages. In:
Caplan L, Van Gijn J (eds) Stroke syndromes, 3rd edn.
Cambridge University Press, Cambridge, UK, pp
461–468
39. Kim JS (2003) Pure lateral medullary infarction:
clinical-radiological correlation of 130 acute, consecu-
tive patients. Brain 126(Pt 8):1864–1872. doi:10.1093/
brain/awg169, awg169 [pii]
17 Territorial Strokes as a Tool to Learn Vascular Territories
40. Mazighi M, Amarenco P (2012) Cerebellar infarcts. In:
Caplan L, Van Gijn J (eds) Stroke syndromes, 3rd edn.
Cambridge University Press, Cambridge, UK, pp
469–479
41. Siclari F, Burger IM, Fasel JH, Gailloud P (2007) Devel-
opmental anatomy of the distal vertebral artery in rela-
tionship to variants of the posterior and lateral spinal
arterial systems. AJNR Am J Neuroradiol 28
(6):1185–1190. doi:10.3174/ajnr.A0498, 28/6/1185 [pii]
42. Baird TA, Muir KW, Bone I (2004) Basilar artery
occlusion. Neurocrit Care 1(3):319–329. doi:10.1385/
NCC:1:3:319, NCC:1:3:319 [pii]
43. Smith E, Delargy M (2005) Locked-in syndrome. BMJ
330(7488):406–409. doi:10.1136/bmj.330.7488.406
330/7488/406, 330/7488/406 [pii]
44. Caplan LR (1980) “Top of the basilar” syndrome.
Neurology 30(1):72–79
45. Kumral E, Bayulkem G, Evyapan D (2002) Clinical
spectrum of pontine infarction. Clinical-MRI correla-
tions. J Neurol 249(12):1659–1670. doi:10.1007/
s00415-002-0879-x
381
46. Lazzaro NA, Wright B, Castillo M, Fischbein NJ,
Glastonbury CM, Hildenbrand PG, Wiggins RH,
Quigley EP, Osborn AG (2010) Artery of percheron
infarction: imaging patterns and clinical spectrum.
AJNR Am J Neuroradiol 31(7):1283–1289.
doi:10.3174/ajnr.A2044, ajnr.A2044 [pii]
47. Chaves C, Caplan L (2012) Posterior cerebral artery.
In: Caplan L, Van Gijn J (eds) Stroke syndromes, 3rd
edn. Cambridge University Press, Cambridge, UK, pp
405–418
48. Dejerine J, Roussy G (1906) Le syndrome thalamique.
Rev Neurol 14:521–532
49. Jodar M, Martos P, Fernandez S, Canovas D, Rovira A
(2011) Neuropsychological profile of bilateral paramedian
infarctions: three cases. Neurocase 17(4):345–352.
doi:10.1080/13554794.2010.509324, 931811693 [pii]
50. Ringelstein EB, Dittrich R, Stogbauer F (2012)
Borderzone infarcts. In: Caplan L, van Gijn J (eds)
Stroke syndromes, 3rd edn. Cambridge University
Press, Cambridge, UK, pp 480–500
 Hemorrhagic Stroke
18
Julius Griauzde, Elliot Dickerson, and Joseph J. Gemmete

Contents Magnetic Resonance Angiography

of Intracranial Hemorrhage . . . . . . . . . . . . . . . . . . . . . . . 391
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
Time of Flight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392
Clinical Presentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
Time of Flight: Pitfalls in Evaluating Causes
Non-contrast Computed Tomography in of Intracranial Hemorrhage . . . . . . . . . . . . . . . . . . . . . . . 393
Intracranial Hemorrhage . . . . . . . . . . . . . . . . . . . . . . . . . . 385
Contrast-Enhanced MRA . . . . . . . . . . . . . . . . . . . . . . . . . . 393
NCCT Technical Considerations and Pitfalls . . . . 386
Contrast-Enhanced MRA: Pitfalls in
Magnetic Resonance Imaging in Intracranial Evaluating Causes of Intracranial
Hemorrhage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387 Hemorrhage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394
Magnetic Susceptibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387 MRA: Aneurysms, AVMs, DAVF,
and Neoplasms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394
MRI of Intracranial Hemorrhage:
Biologic Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388 Imaging Patterns of Intraparenchymal
Hemoglobin Structure, Iron, and Oxygenation . . . . . . 388 Hemorrhage (ICH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394
Plasma, Red Blood Cell Membranes, and Blood Hemorrhage Etiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394
Breakdown Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389 Hypertensive Hemorrhage . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
Illicit Drugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
MRI of Intracranial Hemorrhage: Technical Cerebral Amyloid Angiopathy . . . . . . . . . . . . . . . . . . . . . . 396
Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390 Vascular Malformations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
Arteriovenous Malformations (AVMs) . . . . . . . . . . . . . . 397
CT Angiography of Intracranial Hemorrhage . . . 391
Cavernous Malformations (CMs) . . . . . . . . . . . . . . . . . . . 397
CTA: Pitfalls in Evaluating Causes Dural Arteriovenous Malformations (DAVFs) . . . . . . 397
of Intracranial Hemorrhage . . . . . . . . . . . . . . . . . . . . . . . 391 Aneurysm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
Hemorrhagic Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
Hemorrhagic Transformation of Ischemic Stroke . . . 398
Venous Thrombosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
Trauma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
Additional Clinical Topics in ICH . . . . . . . . . . . . . . . . . 400
J. Griauzde (*) • E. Dickerson Active Hemorrhage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
Department of Radiology, Radiology Resident, University Intraventricular Extension, Associated
of Michigan, Ann Arbor, MI, USA Subarachnoid Hemorrhage, and Hydrocephalus . . . . 400
e-mail: jgriauz@med.umich.edu; Hemorrhage Location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402
dickerse@med.umich.edu Predicting Clinical Outcome . . . . . . . . . . . . . . . . . . . . . . . . . 402
Evolution of the Intraparenchymal Hematoma . . . . . . 403
J.J. Gemmete
Department of Radiology, Neurosurgery, and
Otolaryngology, University of Michigan, Ann Arbor,
MI, USA
e-mail: gemmete@med.umich.edu

# Springer Science+Business Media New York 2016 383

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_46
384 J. Griauzde et al.

Extra-axial Central Nervous System Abstract

Hemorrhage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404 Neuroimaging is integral in the diagnostic
Subarachnoid Hemorrhage (SAH) . . . . . . . . . . . . . . . . . . . 404
Nonaneurysmal Perimesencephalic
algorithm and follow-up of patients with hem-
SAH (NPSAH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405 orrhagic stroke. The role of the neuroimager is
Subdural and Epidural Hematomas . . . . . . . . . . . . . . . . . . 405 to recognize and localize the hemorrhage and
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407 attempt to determine its etiology. Non-contrast
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407 computed tomography has long been the initial
imaging tool in the acute neurologic patient. As
MRI technology and angiographic imaging has
evolved, they too have proven to be beneficial
in narrowing the differential diagnosis and
triaging patient care. Several biological and
physical characteristics contribute significantly
to the appearance of blood products on neuro-
imaging. To adequately interpret images in the
patient with hemorrhagic stroke, the evaluator
must have a knowledge of the interplay
between imaging modalities and intracranial
blood products. Additionally, an understand-
ing of technical parameters as well as the lim-
itations of imaging modalities can be helpful in
avoiding pitfalls. Recognition of typical imag-
ing patterns and clinical presentations can fur-
ther aid the evaluator in rapid diagnosis and
directed care.

Keywords
Hemorrhagic stroke • Hemoglobin • Methemo-
globin • CT angiography • MR angiography •
Time of flight • Hypertension • Amyloid
angiopathy • Venous thrombosis • Aneurysm
• Arteriovenous malformation • Dural arterio-
venous fistula • Subarachnoid hemorrhage •
Subdural • Epidural

Introduction

Neuroimaging is undergoing a second revolution

in terms of understanding disease mechanisms
and therapeutics, fueled by rapid advancements
in advanced computed tomography (CT) and
magnetic resonance imaging (MRI). The accurate
identification and characterization of intracranial
hemorrhage is one of the most essential functions
of a neuroradiologist, because immediate
18 Hemorrhagic Stroke
recognition is important for further diagnostic
work-up, clinical management, and ultimately
patient outcome. CT and MRI are important for
the identification of intracranial hemorrhage, and
MRI can further characterize the age based on
specific signal characteristics. In this chapter, clin-
ical presentation will be discussed first, followed
by the physics and technical limitations of CT,
MRI, computed tomography angiography (CTA),
and magnetic resonance angiography (MRA) in
detecting intracranial hemorrhage. Then, specific
clinical patterns of intracranial hemorrhage will be
presented, so the neuroradiologist can make an
accurate diagnosis of the probable cause based on
clinical history and imaging findings. Finally, the
CT and MRI findings of extra-axial central nervous
system hemorrhage will be discussed, including
subarachnoid hemorrhage (SAH) and subdural
and epidural hematomas.
Clinical Presentation
Rapid identification of intracranial hemorrhage
(ICH) is essential for the clinical management of
ischemic and hemorrhagic stroke, since this will
affect treatment algorithms. Both ischemic and
hemorrhagic stroke are characterized by a sudden
onset of symptoms and neurological deficits. The
severity and type of symptoms depend on the
lesion type, location, and size.
Patients with subarachnoid hemorrhage (SAH)
will typically describe “the worst headache of my
life” prior to presenting with hemorrhage. One in
five may report a less severe sentinel headache in
hours or days leading up to the ictus. This also
may be accompanied by nausea/vomiting, sei-
zure, loss of consciousness, and focal neurologi-
cal signs. In contrast, ICH is classically described
as a gradual process with worsening symptoms
over minutes to hours. Headache and nausea/
vomiting are more frequent in ICH than ischemic
stroke. Given the overlap in the clinical presenta-
tion between ischemic and hemorrhagic stroke,
imaging is vital for determining the etiology and
clinical management of patients, particularly in
the early presentation [1, 2].
385
Non-contrast Computed Tomography
in Intracranial Hemorrhage
Since its inception as a diagnostic tool,
non-contrast computed tomography (NCCT) has
been a highly sensitive first-line screening exami-
nation in the diagnosis of intracranial hemorrhage
[3, 4]. The inherent contrast between the CT atten-
uation of acute blood products and brain paren-
chyma allows for a straightforward diagnosis for
the trained evaluator. Similarly, the speed, rela-
tively low cost, and widespread availability of CT
have preserved its place in the diagnostic algorithm
of the acute neurologic patient. Because no contrast
is needed to identify intracranial hemorrhage, there
is virtually no risk to the patient, although the small
associated radiation dose should be a cause of
additional consideration in pediatric patients.
The appearance of blood (or any other sub-
stance) on CT relates to its ability to attenuate
x-rays. This property is defined as substances’
attenuation coefficient. In clinical practice, the
CT attenuation of a substance is described by its
Hounsfield density. Hounsfield density values
(HU) range from
1,000 (air) to +1,000 (bone)
with a center at zero (water) [5]. The attenuation
of white matter ranges between approximately
25 and 34 HU and that of gray matter from 30 to
40 HU [6]. Gray matter values can normally
extend as high as 45 HU. The CT attenuation of
intracranial blood varies linearly with the hemat-
ocrit value, and hemoglobin concentration is the
most significant contributor to the attenuation of
an intracranial hematoma [7, 8]. Intracranial blood
in a patient with normal hematocrit can be
expected to range between 30 and 60 HU imme-
diately after the hemorrhagic event (Fig. 1). As
time progresses, the attenuation of intracranial
blood products will evolve in a somewhat predict-
able manner. In the case of intraparenchymal
hematoma, the Hounsfield attenuation of the col-
lection will increase (particularly centrally) as clot
forms and serum is extruded [9]. This process will
occur in the first hours and days after the hemor-
rhagic event with HU values ranging from 60 to
80. Higher values extending up to 90–100 HU can
be expected in the center of the hematoma with
386
Fig. 1 Hounsfield densities. Cerebrospinal fluid has an
HU of 0 (white arrow). White matter has an HU of 25–34
(open white arrow). Gray matter has an HU of 30–40
(white arrowhead). Acute blood has an HU of 60–80
(black arrowhead). Bone has an HU near 1000 (open
black arrow)
low attenuation surrounding the hematoma
related to extruded serum and edema.
At the midpoint of the first week after the hem-
orrhagic event, the blood collection will begin to
decrease in attenuation converging from the periph-
ery centrally. The density is estimated to decrease by
approximately 1–2 HU per day [10]. This decrease
in density is attributable to the breakdown of red
blood cells and subsequent removal of the compo-
nents of the hematoma, most importantly hemoglo-
bin, by phagocytic macrophages [10]. This process
leads to an inhomogeneous appearance which is
mildly hyperdense to isodense to brain parenchyma
(HU approximately 30–50) in the subacute period
(up to about 30 days) [11, 12]. After 30 days, the
collection is expected to be hypodense to brain
parenchyma (<30 HU) [10, 11].
In addition to the initial identification of intra-
cranial hemorrhage, NCCT helps triage the care of
the patient presenting with intracranial hemor-
rhage. Identification of the location of the hema-
toma, when taken together with the clinical
picture, aids in narrowing the differential diagno-
sis and deciding on additional imaging [13]. Eval-
uating the extent of the hematoma and the
J. Griauzde et al.
associated mass effect influences immediate treat-
ment planning, and the volume of the hematoma
has been identified as an important predictor of a
30-day patient outcome [14].
NCCT Technical Considerations
and Pitfalls
A simplified description of technical imaging con-
siderations is included to highlight solutions for
some common pitfalls that can be encountered.
Axial images using 5 mm reconstructions are gen-
erally used in NCCT of intracranial hemorrhage, as
this thickness has an adequate signal-to-noise ratio
to allow for evaluation of the brain parenchyma,
identification of intracranial hemorrhage, and
assessment of intracranial mass effect. Unfortu-
nately, in some cases intracranial hematoma can
be obscured by nearby skeletal structures or be
mistaken for dural prominence or calcifications
[15]. Reviewing a wider “blood” window can
improve the detection of intracranial hemorrhage
which layers along the skull (Fig. 2a, b). Review of
coronal and sagittal reformatted images is also par-
amount in all cases but particularly when findings
of hemorrhage are equivocal. Reformatted images
allow for visualization parallel and/or perpendicu-
lar to the long axis of the hematoma allowing for
improved diagnostic accuracy [16] (Fig. 3a, b).
Similarly, thin-cut images may help resolve any
questionable areas (Fig. 2c). In the interest of
time, thin section cuts should only be created and
reviewed when there is a specific area of concern.
Review of bone algorithm reconstructions is also
suggested (particularly in cases of trauma) for iden-
tification of skull fractures which will heighten the
readers’ suspicion of underlying intracranial hem-
orrhage and help to exclude mimickers of hemor-
rhage like small calcifications (Fig. 3c).
It is important to note that in the first minutes
after hemorrhage, there is a partial overlap
between the HU of blood products and that of
brain parenchyma, creating a potential pitfall in
diagnosis. Similarly, intracranial hemorrhage can
remain occult in patients with low hematocrit
because of isodensity to brain parenchyma [17].
18 Hemorrhagic Stroke
Fig. 2 (a) Focus of extra-axial hemorrhage layering along
the left frontal region is poorly seen with standard
windowing secondary to adjacent skull (black arrow). (b)
Focus of hemorrhage is more conspicuous with wider
Fig. 3 (a) Subtle focus of hemorrhage along the posterior
skull is very difficult to identify (black arrow). (b) Sagittal
image in wider blood window allows visualization parallel
to the long axis of the hemorrhage increasing conspicuity
In such cases, and in those where there is a high
index of suspicion for intracranial hemorrhage
without definite CT abnormality, MRI should be
recommended for further evaluation.
Magnetic Resonance Imaging
in Intracranial Hemorrhage
The evaluation of blood products on magnetic
resonance imaging (MRI) is much more involved
than on CT. The appearance depends on a myriad
of biologic factors which evolve with time and
vary based on operator-dependent factors. A basic
387
“blood” window (black arrow). (c) Focus of hemorrhage
is made even more conspicuous using thin-cut slices (black
arrow). Note the increased noise associated with the
thinner cut
(white arrow). (c) Bone algorithm window shows a skull
fracture, heightening the reader’s attention for subtle hem-
orrhage (black arrow)
understanding of magnetic properties as well as
the biologic and technical factors which affect the
appearance of intracranial blood products aids in
overcoming confusion and avoiding potential pit-
falls. What follows is a description of the key
factors involved and their effects on the appear-
ance on intracranial blood products.
Magnetic Susceptibility
Magnetic susceptibility refers to the measure of
substances’ interaction and response to an applied
magnetic field [18, 19]. In simplified terms, the
388
magnetic susceptibility of a substance can be
attributed to the pairing of its valence electrons
[19]. Those substances which have no unpaired
valence electrons are termed diamagnetic. Paired
electrons in an orbital have opposing spins which
cancel each other, resulting in no magnetic dipole
when placed in a magnetic field [19]. When a
diamagnetic substance is placed in an applied
magnetic field, its valence electrons do experience
minimal alterations in their orbital motion, how-
ever [19]. According to Lenz’s law, this results in
a weak induced magnetic field which opposes the
applied field [20].
In contrast to diamagnetic substances, para-
magnetic substances have unpaired valence elec-
trons. When placed in a magnetic field, the spins
of the valence electrons align with the magnetic
field creating a magnetic dipole which augments
the applied field [19, 20]. In human tissues, the
magnetic dipole of paramagnetic substances
allows them to undergo proton-electron dipole
interactions with local protons (i.e., those in
water molecules) [21]. These interactions have a
profound effect on the magnetic microenviron-
ments of human tissues, most significantly affect-
ing the longitudinal relaxation of nearby protons,
resulting in T1 shortening [20, 22]. For this dis-
cussion, it is important to note that the strength of
dipole interactions varies inversely with the sixth
power of the distance between the dipoles [23].
Also, as magnetic field strength increases,
paramagnetic properties become more pro-
nounced because paramagnetic substances aug-
ment the applied field more so than diamagnetic
substances oppose it [24]. Relevant compounds
are classified based on their magnetic susceptibil-
ity in Table 1.
MRI of Intracranial Hemorrhage:
Biologic Factors
Several biologic factors contribute to the charac-
teristic appearance of intracranial blood products
on MRI. As time passes, intracranial blood prod-
ucts are exposed to metabolic processes and var-
iations in their local environment which alter their
chemical and magnetic properties. What follows
J. Griauzde et al.
Table 1 Purely diamagnetic and paramagnetic substances
Purely diamagnetic Paramagnetic
Water Deoxyhemoglobin
Oxyhemoglobin Methemoglobin
Most soft tissue Hemosiderin/ferritin
Bone Gadolinium
is a brief description of these processes and the
resultant effects they have on the MRI signal
produced.
Hemoglobin Structure, Iron,
and Oxygenation
Hemoglobin is a metalloprotein composed of four
protein moieties, each of which contains a heme
subunit. Each heme subunit is composed of a
porphyrin ring with a central iron atom which
can reversibly bind oxygen [25] (Fig. 4). In oxy-
hemoglobin, an iron atom is reversibly bound to
oxygen in the ferrous state (+2). Oxyhemoglobin
has no unpaired outer electrons and is therefore
diamagnetic [26].
In oxygen-poor environments (i.e., the intra-
cranial hematoma), hemoglobin releases its bound
oxygen leaving deoxyhemoglobin with four
unpaired electrons and the paramagnetic proper-
ties that come with them [26]. From the discussion
above, this would lead the reader to believe that
the presence of deoxyhemoglobin in human tis-
sues would result in dipole interactions with con-
sequential T1 shortening. The magnetic properties
of deoxyhemoglobin are more complex, however.
For dipole interactions to occur between the
valence electrons of iron and the protons of
water molecules, they must be able to approach
within 3 Å of each other [23]. The loss of oxygen
from hemoglobin causes a shift in its three-
dimensional configuration, displacing the iron
atom of the heme subunit out of the plane of the
porphyrin ring [25, 27]. This shift effectively
increases the dipolar distance between iron and
nearby water proton dipoles, preventing any sig-
nificant dipolar interactions [21, 27] (Fig. 5). The
result is a lack of marked T1 shortening in the
presence of deoxyhemoglobin.
18 Hemorrhagic Stroke
With further exposure to an oxygen-poor
environment, deoxyhemoglobin is oxidized into
methemoglobin whose iron atom is in the Ferric
state (+3) and unable to bind oxygen [25]. The
associated conformational change decreases the
dipolar distance between the unpaired valence
electrons of iron and nearby water protons,
allowing for proton-electron dipole interactions
[22, 26, 28]. Therefore, when methemoglobin is
Fig. 4 The heme subunit of hemoglobin
Fig. 5 (a) In plane view of the oxyhemoglobin molecule.
The oxygen-iron bond is indicated by an arrow. (b) When
oxygen is released by hemoglobin, the iron atom shifts out of
plane from the heme subunit. This shift prevents water mol-
ecules from approaching within 3 Å of the valence electrons
of iron, preventing dipolar interactions. This figure was
389
present, there is a significant T1 shortening in the
intracranial hematoma.
As time progresses, methemoglobin is further
oxidized, metabolized, and phagocytosed yield-
ing residual iron atoms which are scavenged and
stored as hemosiderin and ferritin [29]. The scav-
enged iron is paramagnetic, but because of the
structures of hemosiderin and ferritin, water can-
not closely approach it and proton-electron dipo-
lar interactions are not possible [20, 30–32].
Plasma, Red Blood Cell Membranes,
and Blood Breakdown Products
Freshly extruded plasma (prior to clot formation)
has high water content, admixed serum proteins,
and intact red blood cells. The high water content
of extruded plasma causes T2 prolongation com-
pared to brain parenchyma [33]. As time pro-
gresses, clot begins to form, water is resorbed,
and serum proteins are concentrated with subse-
quent clot retraction. These processes all have
minor contributions to T2 shortening of the evolv-
ing hematoma [34]. Red blood cell (RBC)
created using graphics reproduced under the GNU Free Doc-
umentation License (Citation: http://commons.wikimedia.
org/wiki/File:Heme_deoxy_vs_oxygenated.jpg created by
user Mrbean427 and http://commons.wikimedia.org/wiki/
File:Dipole_H2O.svg user NEUROtiker)
390
Table 2 Biologic factors affecting T1 and T2
Field strength dependent
Biologic factors T1 T2 (T2)
Fresh plasma
/ " No
#
Deoxyhemoglobin
# Yes
Methemoglobin " # Yes
Hemosiderin/
# Yes
ferritin
RBC membrane
# Yes
membranes remain intact early on after extrusion,
sequestering hemoglobin. In the oxygen-poor
environment of the hematoma, oxyhemoglobin
is converted to deoxyhemoglobin and then later
to methemoglobin. In this way, the RBC effec-
tively creates a paramagnetic microenvironment
in contrast to the diamagnetic microenvironment
external to it. Diffusing water molecules experi-
ence the gradient created by this magnetic field
inhomogeneity, which dephases their protons and
results in T2 shortening [24, 35, 36]. When the
RBCs lyse, the deoxy- and methemoglobin are no
longer sequestered, and the field inhomogeneity is
lost. Finally, after progressive metabolism, para-
magnetic iron atoms are taken up and sequestered
by hemosiderin and ferritin. This again results in
magnetic field inhomogeneity which dephases
diffusing water protons with resultant T2 shorten-
ing [30]. The effects of biological materials on T1
and T2 in intracranial hemorrhage are listed in
Table 2.
MRI of Intracranial Hemorrhage:
Technical Considerations
The appearance of intracranial blood products can
be significantly altered depending on the
magnetic field strength and the pulse sequence
used. Magnetic field strength imparts its influence
by altering susceptibility-induced T2 effects.
The inverse of the T2 relaxation rate varies qua-
dratically with the field strength [24, 37, 38].
This relationship is a product of the relatively
increased magnetization of paramagnetic
J. Griauzde et al.
How does it work on How does it work on
T1? T2?
High protein content High water content
N/A Susceptibility
Dipole interactions Susceptibility
N/A Susceptibility
N/A Susceptibility
compounds in relation to diamagnetic compounds
at higher field strengths [24]. This exaggerated
difference in magnetization results in greater
magnetic field inhomogeneity between paramag-
netic and diamagnetic microenvironments
causing more pronounced T2 shortening
(T2 darkness).
Gradient recall echo (GRE) sequences are
highly sensitive to susceptibility effects caused
by magnetic field in homogeneities
[38–40]. GRE sequences are, in turn, highly sen-
sitive for the detection of intracranial hemorrhage
and highly accurate at predicting its extent
[38, 41]. This makes GRE particularly valuable
in the hyperacute and chronic periods, when blood
products might otherwise remain inconspicuous
[38, 39]. As a result, GRE sequences have become
widely incorporated in standard MRI protocols
for the evaluation of intracranial hemorrhage.
Fast spin echo (FSE) sequences have become a
mainstay of standard MRI protocols, widely
replacing conventional spin echo. By using
shorter interecho times and closely spaced pulses,
FSE allows for significant reductions in scan times
[42, 43]. These shorter echo times have been
shown to cause increased T2 prolongation by a
variety of mechanisms [42]. An untoward result of
this T2 prolongation is a decreased sensitivity
to T2 susceptibility effects, although the clinical
significance of this has been questioned [43].
Nevertheless, it is recommended that GRE
sequences be included in standard MRI protocols
to ensure that high sensitivity for small amounts
of blood products is maintained despite the usage
of FSE.
18 Hemorrhagic Stroke
CT Angiography of Intracranial
Hemorrhage
CT angiography (CTA) is a useful adjunct to both
NCCT and MRI in the evaluation of patients with
acute intracranial hemorrhage. CTA images can
be rapidly acquired and reconstructed to allow
intricate visualization of the intracranial vessels
in multiple planes and projections. It has high
sensitivity, specificity, and diagnostic accuracy in
identifying vascular causes of intraparenchymal
hemorrhage and can be beneficial in evaluating
the vascularity of hemorrhagic neoplasms
[44–48].
In addition to its role in diagnosis, CTA can
provide valuable prognostic information in the
patient with intracranial hemorrhage. The CTA
“spot sign” is a high-attenuation focus within an
intracerebral hematoma which represents a site of
extravasation [49] (Fig. 6). In patients with intra-
cerebral hematoma, the “spot sign” has been iden-
tified as a predictor of hematoma progression,
longer mean hospital stay, in-hospital mortality,
and poor overall outcomes in survivors [50–53].
Fig. 6 The CTA “spot sign”: a high-attenuation focus
along the medial aspect of an intracranial hematoma
(arrow)
391
CTA: Pitfalls in Evaluating Causes
of Intracranial Hemorrhage
Although CTA plays an important role in the
evaluation of patients with intracranial hemor-
rhage, it has several notable limitations.
Reconstructed images (i.e., volume-rendered and
maximum-intensity-projection images) can
enhance artifacts and are susceptible to data loss,
potentially confusing the diagnosis of vascular
pathology [54]. Pitfalls such as this can be
avoided by reviewing the source images in all
cases and using multiplanar reformatted images
(which contain all source data), rather than relying
on reconstructed images [54]. Acute extraluminal
blood products can also obscure evaluation of the
arteries, although this can be minimized with
proper arterial enhancement, as the Hounsfield
density of intravenous contrast is significantly
higher than that of acute extraluminal blood.
Unfortunately, other artifacts are not as easily
avoided as these.
Evaluation of patients with previous place-
ment of embolic materials, endovascular coils,
and surgical aneurysm clips is often
suboptimal secondary to marked streak artifact.
Also, CTA is limited in evaluating small arter-
ies as well as arteries near bony structures,
secondary to limited spatial resolution, poor
contrast between arteries and adjacent bones,
and beam-hardening artifact [55, 56]. Because
of these limitations, catheter-based digital sub-
traction angiography (DSA) remains the gold
standard in the evaluation of intracranial hem-
orrhage [48, 57, 58].
Magnetic Resonance Angiography
of Intracranial Hemorrhage
Magnetic resonance angiography (MRA) is an
extensive topic whose breadth is beyond the
scope of this chapter. There are, however, impor-
tant concepts which will allow for a more thor-
ough understanding of its value and shortcomings
in the diagnostic algorithm of intracranial hemor-
rhage. The three main types of magnetic
392
resonance angiography are time-of-flight MRA
(TOFMRA), phase-contrast MRA (PCMRA),
and contrast-enhanced MRA (CEMRA).
PCMRA allows for evaluation of hemodynamic
flow, but its use in the anatomic evaluation of
intracranial vasculature is at present dwarfed by
TOFMRA. Therefore, this discussion will focus
on the important strengths and weaknesses of
non-contrast TOFMRA and CEMRA in the eval-
uation of intracranial hemorrhage.
Time of Flight
TOF imaging, in its simplest form, takes advan-
tage of the flow of blood. In TOF, radiofrequency
(RF) pulses are applied at short TRs to a given
slice thickness resulting in the saturation of all
the magnetic spins within the slice and suppres-
sion of their longitudinal magnetization
[59]. Spins that are moving perpendicularly into
and out of the slice are not saturated, however. In
a given length of time, the blood that has been
saturated by the RF pulses will move out of the
slice and be replaced by “new” unsaturated
blood. When exposed to an excitatory pulse,
this “new” blood is magnetized and can undergo
longitudinal relaxation with resultant bright sig-
nal (Fig. 7). The relative signal of the flowing
blood will be directly related to the slice
Fig. 7 New blood flows
into the imaging slice
displacing saturated blood. Saturating RF pulses
When the “new” blood
arrives, it can be excited and
undergoes longitudinal
relaxation with resultant
bright signal
Stationary Tissue in Slice
J. Griauzde et al.
thickness and the velocity of the blood within
the vessel. Maximal signal is obtained when all
saturated blood is replaced with unsaturated
blood. This occurs when the velocity of flow is
greater than or equal to the slice thickness
divided by the repetition time, represented by
the equation v  z/TR, where v is the velocity
of flow, z is the slice thickness, and TR is the
pulse repetition time [60, 61]. Additional factors
that affect the signal of flowing blood include the
flip angle of the pulse, the T1 of the tissue in the
slice, and the orientation of the vessel to the
imaging slice [59, 61]. In the case of the intracra-
nial circulation, flow signals are optimized by
selecting axial slice acquisition, which is perpen-
dicular to the average overall intracranial arterial
flow. Similarly, optimal TR and slice thickness
can be selected to ensure that the signal of the
static tissue is suppressed and that the signal in
the vessels is maximally enhanced.
TOF imaging identifies any flow within an
image slice; therefore, both arterial and venous
structures will produce signal. To selectively
highlight the vascular structure of interest (artery
or vein) and to minimize contamination of the
other, saturation bands are employed. These
pre-saturation bands are positioned a certain dis-
tance from the imaging slice to suppress the
undesired flow signal before it arrives within the
imaging slice [62] (Fig. 8).
Saturated Blood
When V > z /TR
Flow
18 Hemorrhagic Stroke
Fig. 8 Blood within a
vessel is pre-saturated
upstream to the imaging
slice. When the blood
arrives in the imaging slice
(asterisk), it has no signal
because of pre-saturation
Saturated Imaging Slice
Saturating Pulses
Pre-Saturated Band
Time of Flight: Pitfalls in Evaluating
Causes of Intracranial Hemorrhage
Because of its dependence on flow and spin satu-
ration, TOF is susceptible to many artifacts. When
flow within a vessel is slow or turbulent, or the
orientation of flow is parallel (or near parallel) to
the imaging slice, the blood spins have a longer
“dwell time” within the image slice [63]. Increased
“dwell time” exposes the spins to more saturating
pulses, leading to suppressed signal in these por-
tions of the vessel, which can mimic vessel
narrowing or occlusion [63]. Additional sources
of artifact include bright T1 substances, tissue inter-
faces, pulsation, and foreign bodies. Methemoglo-
bin (T1 bright) presents a particular challenge in
TOF because it is often present in hemorrhage
and/or vessel thrombus. The T1 signal from met-
hemoglobin can be difficult to suppress, and it
recovers its T1 signal rapidly, decreasing the con-
spicuity of vessels from background tissue and
even mimicking flow where there is none
[58, 64]. Metallic foreign bodies (surgical clips)
and tissue interfaces can dephase spins in nearby
vessels causing alteration in their apparent config-
uration and diameter, although these are a less
significant source of artifact [65, 66].
The artifacts associated with TOF can make
proper evaluation of aneurysms, arteriovenous
malformations (AVM), and dural venous fistulas
(DAVF) very challenging. Aneurysms can contain
393
Flow
complex flow patterns which vary based on their
size and morphology. Turbulent and slow flow
within the aneurysm results in loss of signal on
TOF, leading to underestimation of the aneurysm
volume and neck size [67]. Also, the presence of
methemoglobin in thrombosed aneurysms can
lead to false-positive results [66]. Similarly,
AVMs and DAVFs pose a formidable diagnostic
challenge on TOFMRA because of their vessel
tortuosity, complex angioarchitecture, and slow-
flow draining veins. These properties cause
TOFMRA to overestimate the size of small AVMs
and underestimate the size of large AVMs [68].
Additionally, TOF is limited in correctly
interpreting small feeding arteries and in identifying
draining veins of AVMs and DAVFs, which can
have significant treatment implications [68, 69].
The use of higher field strength has been shown to
improve characterization, but many shortcomings
remain [68].
Contrast-Enhanced MRA
CEMRA employs fast spoiled gradient-recalled
echo-based sequences (FSPGR) and the paramag-
netic properties of gadolinium to intensify the
signal within vessels. FSPGR is used to spoil the
transverse magnetization of background tissues
which accentuates the marked T1 shortening of
intravascular gadolinium [70]. FSPGR also
394 J. Griauzde et al.

employs short RF pulses allowing for fast Table 3 MRA and CTA: sources of artifact and limitation
sequence times [70]. Administration of gadolin- in intracranial hemorrhage
ium allows for improved signal-to-noise ratio and Source CTA TOFMRA CEMRA
minimizes flow-related artifacts [70]. Acute blood products +


Methemoglobin
+ +
Tissue interfaces
+

Contrast-Enhanced MRA: Pitfalls Skeletal structures +


in Evaluating Causes of Intracranial Metal/embolic ++ + +
materials
Hemorrhage Tortuous vessels
++

Turbulent/slow flow
++

The saturation and spin dephasing effects that Complex
++ +
occur with TOF are of no consequence in angioarchitecture
CEMRA. It does, however, have its own limita- Venous enhancement + + +
tions. Correct bolus timing is paramount to creat- Small arteries + + +
ing a diagnostic study but unfortunately can be ++ significant effect, + some effect,
no significant effect
very challenging to obtain [70]. When bolus
timing is poor, significant artifact can limit evalu-
ation of vascular pathology [71]. Artifacts includ- combination of TOFMRA and CEMRA can reli-
ing distortion of the size of the vessel, “ringing” ably diagnose moderate- and high-flow DAVF, and
within the vessel, and nontarget vessel enhance- CEMRA has the added ability to identify the pres-
ment can obscure pathology [70, 72]. In addition ence of cortical venous drainage [69]. MRA plays a
to bolus timing, the marked T1 shortening of limited role in the evaluation of intracranial neo-
methemoglobin can obscure contrast between plasms. It can identify encasement and mass effect
vessels and background tissue, as it does in TOF of large arteries but does not reliably identify small
images. Finally, limited spatial resolution in arterial feeders [77].
CEMRA makes it insensitive in the evaluation of
small aneurysms (<3 mm) and in the evaluation
of complex AVMs [73]. Artifacts and limitations Imaging Patterns of Intraparenchymal
of CTA and MRA are summarized in Table 3. Hemorrhage (ICH)

MRA: Aneurysms, AVMs, DAVF,
and Neoplasms
Despite the limitations noted above, TOFMRA and
CEMRA play a role in the diagnostic work-up of
patients with intracranial hemorrhage. MRA has
been shown to adequately depict aneurysms greater
than 3 mm, which encompasses most clinically
significant aneurysms [66, 74, 75]. Additionally,
MRA has been shown to be sensitive in the detec-
tion of AVMs greater than 3 cm [48]. Furthermore,
the addition of contrast allows for improved delin-
eation of aneurysm anatomy and AVM nidus as
well as improved assessment of AVM drainage
pattern and small arterial feeders [71, 73, 76]. The
Hemorrhage Etiology
There are numerous etiologies of ICH including
both traumatic and nontraumatic. Among
nontraumatic etiologies, about 80–90 % are “pri-
mary” ICH which is related to hypertension
and/or cerebral amyloid angiopathy and the asso-
ciated changes of vessel walls which leads to
weakening and rupture [78–80]. Less common
but still clinically relevant “secondary” causes of
ICH include ruptured vascular malformations,
hemorrhagic transformation of ischemia, tumor,
vasculitis, anticoagulation, and cerebral venous
thrombosis. Determining an exact etiology for an
ICH can be challenging because multiple
dynamic factors may be present in the acute
18 Hemorrhagic Stroke 395

neurologic patient; for instance, ICH can arise

from hemorrhagic transformation of infarcted
tissue, but a large ICH may also cause mass effect
on surrounding tissue that restricts the perfusion
gradient of the tissue and leads to surrounding
infarction. Familiarity with characteristic imag-
ing factors can improve the interpreter’s ability to
troubleshoot in challenging cases.

Hypertensive Hemorrhage

Hypertension induces changes of intracranial arte-

rial vessel walls, weakening them and allowing
them to rupture leading to hypertensive ICH.
These changes include smooth muscle cell hyper-
plasia, smooth muscle cell death, collagen
replacement of the vascular walls, and arteriolar
ectasia [78, 79]. The most classic locations of
hypertensive ICH are the basal ganglia, thalami,
pons, and cerebellum; the cerebral lobes are some-
what less common (Fig. 9) [80]. A relatively
homogeneous single blood collection in these
locations, without imaging evidence of an under-
lying etiology such as an aneurysm, is a
classical appearance of these hemorrhages.
Patient clinical factors include hypertension as
well as other cardiovascular risk factors such as
old age, smoking, and excessive alcohol use
[79]. Imaging markers of diffuse microvascular
disease can suggest long-standing vascular
changes which predispose to hypertensive ICH.
These include periventricular low attenuation on
CT or FLAIR hyperintensity on MRI and numer-
ous foci of susceptibility artifact on T2* MR
sequences in characteristic locations which repre-
sent evidence of prior clinically silent
“microbleeds” [81].
In the absence of other recognized etiologies,
ICH is typically assumed to be related to hyper-
tension. It is possible that some of the ICHs attrib-
uted to hypertension are in fact occult
presentations of other etiologies. Nevertheless,
the pathologic effects of hypertension on other
risk factors for ICH (i.e., vascular malformations)
cannot be dismissed, and control of blood
Fig. 9 Hypertensive hemorrhage into the thalamus. A
62-year-old female presenting with altered mental status
related to hypertensive ICH. The hematoma is centered in
the left thalamus, a classic location for hypertensive ICH
(solid arrow). Note that blood has dissected into the lateral
ventricle (open arrow). Also, there is a mild edema around
the hematoma and lateral ventricle
pressure after an ICH is of critical importance as
discussed in the American Stroke Association’s
most recent guidelines on intracranial hemorrhage
(Class 1, Level A recommendation) [82].
Illicit Drugs
Although far less common than hypertensive
ICH, illicit drug-induced ICH is presented here
because its imaging findings are similar to hyper-
tensive hemorrhage in terms of bleed location.
Commonly implicated drugs include those caus-
ing hypertension, vasospasm, or even cerebral
vasculitis such as cocaine, amphetamines, and
MDMA (“ecstasy”). Differentiation for hyperten-
sive ICH is typically based on clinical factors
including a history of illicit drug use, younger
age, and absence of classical cardiovascular risk
396
Fig. 10 A 66-year-old female with mental status changes
from hemorrhages related to cerebral amyloid angiopathy.
T2* GRE images show multifocal hemorrhages based in a
factors. For young adults, drug use may surpass
other biologic risk factors in causing ICH,
although these patients often have a more favor-
able clinical course compared to other patients
with ICH [83].
Cerebral Amyloid Angiopathy
Amyloid angiopathy is a common etiology of
nontraumatic ICH, perhaps second only to hyper-
tension [84]. Amyloid angiopathy is characterized
by the deposition of β-amyloid protein into small-
and medium-sized cerebral arteries [79]. The vas-
cular insults related to cerebral amyloid
J. Griauzde et al.
subcortical location (solid arrows) and scattered spots of
susceptibility artifact from remote microhemorrhage
angiopathy may play an important role in several
dementias, and most patients presenting with ICH
related to amyloid are elderly (usually >70 years).
The imaging hallmark of cerebral amyloid
angiopathy is hemosiderin staining secondary to
multiple prior microbleeds, particularly in the cere-
bral lobes (Fig. 10). This is in contrast to the tha-
lamic and basal ganglia microbleeds seen in
hypertensive angiopathy. These microbleeds are
most readily detected as scattered areas of suscepti-
bility artifact on T2* images [85, 86]. The charac-
teristic finding of multiple lobar microhemorrhages
suggests the diagnosis of cerebral amyloid
angiopathy even for patients who have not
presented with a major ICH [85, 86].
18 Hemorrhagic Stroke
Vascular Malformations
Vascular malformations are the most common
cause of ICH which are secondary to specific
anatomic abnormalities and are a leading cause
of ICH in young adults. Vascular malformations
often include vessel wall components which are
not capable of withstanding the high pressure
exerted within the vascular malformation, leading
to ICH [87]. By far the most clinically relevant
vascular malformations are arteriovenous
malformations (AVMs), which lead to about
80 % of ICH due to vascular malformations
[88]. These are followed in importance by cavern-
ous malformations (CMs) and dural arteriovenous
fistulas (DAVFs). Developmental venous anoma-
lies (DVAs) and capillary telangiectasias usually
do not lead to ICH unless they occur in association
with other vascular anomalies. Identification of
the causative vascular malformation is rarely pos-
sible using non-contrast examination, and CTA,
contrast-enhanced MRI/MRA, and conventional
angiography are much more sensitive.
Arteriovenous Malformations (AVMs)
Arteriovenous malformations (AVMs) are develop-
mental lesions in which pial arteries connect directly
into pial veins without an intervening capillary bed,
allowing for rapid shunting of blood through the
malformation and elevating pressures on the venous
side. AVMs often have a characteristic “bag of
worms” appearance both on imaging and gross
examination due to the high flow of blood through
the feeding and draining vessels (Fig. 11) [87]. A
central nidus at the actual site of connection
between the arterial and venous vessels is a charac-
teristic feature although the size can vary from
microscopic (nearly impossible to detect on most
imaging modalities) to several centimeters in size
(potentially detectable on even non-contrast CT).
The gold standard for the treatment of AVMs
remains surgical excision, but therapeutic guide-
lines endorse a role for endovascular and radiation
treatments prior to surgery for some patients.
When AVMs are detected incidentally, the care
team will usually opt for treatment, as they have a
397
significant risk of future hemorrhage (approxi-
mately 2 % per year) [89]. Findings which predict
that an AVM will lead to future bleeding include
presenting with hemorrhage, deeper location, an
associated aneurysm (reflecting multifocal weak-
ening of vessel walls), and drainage exclusively
into deep cerebral veins (veins which drain into
the vein of Galen and inferior sagittal sinus as
opposed to superficial veins) [90, 91].
Cavernous Malformations (CMs)
Cavernous malformations are benign low-flow
malformations comprised of endothelial-lined
blood vessels not separated by brain parenchyma.
CMs are characterized by a heterogeneous central
appearance including regions of high T1 and T2
signal with a surrounding area of hemosiderin
staining most obvious as signal loss on T2*
images. CMs can be found in association with
DVAs, and several hereditary syndromes of mul-
tiple CMs have been described [92].
Dural Arteriovenous Malformations
(DAVFs)
DAVFs are fistulas from meningeal arteries to a
meningeal vein or dural venous sinus. DAVFs can
be either sequela of prior trauma such as a fracture
or craniotomy or be developmental. Just as with
developmental AVMs in the brain, DAVFs are
characterized by large feeding and draining ves-
sels, which are best demonstrated by angiographic
imaging. In addition to ICH, symptoms can
include neurologic deficits and tinnitus related to
venous hypertension. Treatment often centers
around curative endovascular techniques without
the need for later surgery [89].
Aneurysm
Aneurysmal hemorrhage most typically causes
subarachnoid hemorrhage (discussed below), but
intraparenchymal hemorrhage can also occur if
the aneurysm is enclosed in brain parenchyma
[50, 79, 93].
398
Fig. 11 AVM with hemorrhage. CT of a 43-year-old
woman presenting with headache and mental status
changes with a large hemorrhage in the left cerebral lobe
(a); areas of irregular high attenuation are presented ante-
rior to hematoma (solid arrow). Further evaluation with
MRI reveals large flow voids with a “bag of worms”
Hemorrhagic Tumors
Both primary and metastatic CNS tumors can be
complicated by clinically significant hemorrhage.
Identification of contrast enhancement can help
delineate the underlying lesion from surrounding
blood products, and the presence of multiple sim-
ilar appearing lesions is highly suggestive of met-
astatic disease [94]. About 5 % of CNS
malignancies will be complicated by macroscopic
bleeding although the risk varies dramatically by
the type of tumor. For instance, among metastatic
masses, about a third of melanoma metastasis will
be complicated by hemorrhage (Fig. 12). Meta-
static adenocarcinomas and squamous cell carci-
nomas have a lower risk of bleeding on a per lesion
basis but are likely etiologies for hemorrhage
J. Griauzde et al.
appearance on T2 image (b, open arrow). The arteriove-
nous malformation which led to the hemorrhage is best
shown on DSA injection of the left internal carotid artery
(ICA) (c–e) including early filling of draining veins
(arrowheads)
because they are more common tumors overall.
Among primary brain tumors, glioblastoma
multiforme, astrocytoma, oligodendroglioma, and
mixed astrocytoma/oligodendroglioma all bleed
with significant frequency, but glioblastoma
multiforme is the most common etiology due to
its high overall prevalence [94, 95].
Hemorrhagic Transformation
of Ischemic Stroke
After an ischemic stroke, bleeding can occur
within the area of infarcted tissue. Hemorrhagic
transformation of ischemic stroke which is symp-
tomatic occurs in about 6 % of patients treated
with thrombolytic therapy and about 1 % of
18 Hemorrhagic Stroke
Fig. 12 Hemorrhagic melanoma metastasis. A 59-year-
old man with history of melanoma and headache. Note the
enhancement of the mass between pre-contrast (a) and
post-contrast (b) axial T1 images as well as the fluid
collection just lateral to the mass (solid arrow). This fluid
patients treated without thrombolysis. It often
carries a dismal clinical prognosis with estimates
of severe morbidity and/or mortality exceeding
90 % [96].
Imaging findings which predict an increased
risk of hemorrhagic transformation include the
CT findings of early stroke (low attenuation in
the infarcted area, loss of gray-white differentia-
tion, and parenchymal swelling), perhaps because
areas of ischemia visible on CT tend to be further
progressed than those visible only with MRI
(Fig. 13). The size of these abnormalities is also
important; several authors have found that CT
changes suggesting stroke over a region larger
than a third of the middle cerebral artery (MCA)
distribution are more likely to hemorrhage
[97]. Studies with MRI have also found that
greater volumes of restricted diffusion also predict
a greater risk of hemorrhage [97, 98].
Venous Thrombosis
Cerebral venous thrombosis can present in a num-
ber of prothrombotic clinical scenarios including
pregnancy, recent surgery, oral contraceptives,
coagulation disorders, dehydration, and infection
(both systemic and local) and present as either
acute or chronic symptoms including headache,
neurologic deficits, mental status changes, and
399
collection demonstrates intrinsic high T1 (a) and high T2
signal (c) compatible with a late subacute hemorrhage.
Also note vasogenic edema surrounding the hematoma
(c, open arrow)
seizure [99]. The impaired venous drainage
reduces the perfusion pressure of the brain paren-
chyma which in turn can lead to ischemia and
subsequent hemorrhagic transformation. ICH
complicates about a third of cerebral venous
thrombosis (Fig. 14). The most characteristic find-
ing of ICH related to venous thrombosis is irreg-
ular or flame-shaped hemorrhage that occurs in the
region of parenchyma drained by the thrombosed
cerebral vein. These findings are almost always
accompanied by other parenchymal abnormalities
caused by the thrombosis such as edema and per-
fusion/diffusion restriction [100].
Trauma
When the head is exposed to external trauma,
common sites of injury are at the site of external
force (coup) and at the area of the brain opposite to
the external force (contrecoup). Intraparenchymal
hemorrhage has a propensity to occur at the
contrecoup site due to anchoring structures
such as falx cerebri and bridging cortical
veins pulling on the parenchyma as the brain is
accelerated. Other areas which are at particular
risk of ICH following trauma are the areas of the
brain adjacent to the jagged portions of the skull
base such as the petrous ridges and the ethmoid
plate.
400 J. Griauzde et al.

Fig. 13 Hemorrhagic
transformation of ischemia.
A 69-year-old male with
stroke demonstrated on
diffusion-weighted images
(a) and non-contrast CT (b)
(solid arrows). Repeat
non-contrast CT at 1 day (c)
and 1 month (d)
demonstrates hemorrhagic
transformation of the
infarction (open arrows)

Small, punctate regions of parenchymal hem-
orrhage can also be seen at sites of other traumatic
injuries such as shearing and diffuse axonal inju-
ries, although low attenuation from edema may
predominate at these sites [101, 102].
Additional Clinical Topics in ICH
Active Hemorrhage
Observing active extravasation of blood into the
area of hemorrhage is an ominous radiologic find-
ing that suggests both ongoing expansion of the
hematoma and a poor clinical outcome [103]. On
non-contrast examinations, freshly extravasated
blood will have attenuation characteristics differ-
ent from the blood which has been present in the
hematoma for a longer period, and these hetero-
geneous groups of blood products can circle
around one another to produce a “swirl sign”
which has also been associated with hemorrhage
growth and poor outcomes [103]. If the patient
receives a CTA study, active extravasation can
present as a tiny spot on arterial phase images
(the “spot sign”) which can rapidly expand on
more delayed phase images (Fig. 15) [50]. Even
when a spot of precise extravasation is not identi-
fied on arterial phase images, more delayed
images can directly demonstrate extravasated con-
trast indicating ongoing hemorrhage [104–106].
Intraventricular Extension, Associated
Subarachnoid Hemorrhage,
and Hydrocephalus
Although intraparenchymal hemorrhage begins
within the brain parenchyma, blood can dissect
through the relatively pliant tissues of the central
18 Hemorrhagic Stroke 401

Fig. 14 Venous sinus

thrombosis with
hemorrhage. A 13-year-old
female that recently
initiated oral contraceptive
pills presenting with
headache. MR venogram
(a) demonstrates occlusion
of the straight sinus and left
transverse sinus
(arrowheads). Diffusion
restriction (b) in the right
thalamus related to the
thrombosis (closed arrow).
Rim of hemorrhage in the
area of the prior diffusion
restriction is demonstrated
on non-contrast CT (c, open
arrow). Follow-up MRI
3 months later (d) after
resolution of thrombosis
shows (not shown)
susceptibility artifact from
hemosiderin staining at the
site of prior hemorrhage
(arrowhead)

Fig. 15 A 51-year-old male with a hypertensive hemor- aneurysm or vascular malformation reveals a tiny high-
rhage in the left basal ganglia seen on non-contrast CT (a, attenuation spot compatible with active extravasation
solid arrow). CT angiography performed to screen for (b and c, open arrows)

nervous system and breech the ependymal linings that it will evolve to include an intraventricular
of the cerebral spinal fluid spaces to present as component also increases. This volume also
intraventricular and subarachnoid hemorrhage. appears to depend upon the location of the hem-
As the size of the ICH increases, the probability orrhage as certain regions of the brain may permit
402
Infratentorial
Brainstem
Pons Medulla
Fig. 16 Classification of areas of hemorrhage within the brain
more hemorrhage volume before forcing blood to
decompress into the ventricular system; for
instance, many more lobar ICHs achieve a volume
over 50 ml without breeching the ventricles than
thalamic ICHs [107].
Just as in primary subarachnoid hemorrhage,
hydrocephalus is a common complication of
intraparenchymal hemorrhage (about half of
patients with ICH). Hydrocephalus is also an
independent predictor of mortality [108].
Hemorrhage Location
ICH may occur almost anywhere within the intra-
cranial parenchyma. The most basic distinction in
hemorrhage location is above or below the
tentorium cerebelli (supratentorial versus
infratentorial). Supratentorial hemorrhagic is
most common and is estimated to occur in
approximately 90 % of cases [109]. About half
to two thirds of the supratentorial ICHs are
based in the deep cerebral structures including
the basal ganglia and thalami with the
remaining supratentorial ICHs based in the
cerebral lobes [109, 110]. Infratentorial hemor-
rhage is less common, accounting for approxi-
mately 10 % of ICH [109]. Brainstem
J. Griauzde et al.
Means of
describing ICH
location
Supratentorial
Cerebellar Deep Lobar
Basal ganglia Thalami
hemorrhages account for about a third of
infratentorial ICH with the remainder occurring
in the cerebellum [109] (Fig. 16). Multifocal ICH
is well described but uncommon (about 1 % of
ICHs) [109, 110].
Studies of the prognostic implications of
ICH location as an independent risk factor
have shown inconsistent results, and
hematoma volume and growth have been identi-
fied as more significant prognostic factors [84,
109–111].
Predicting Clinical Outcome
As alluded to above, a variety of radiologic and
clinical factors can assist in prognosticating for
patients with ICH. Among the most widely used is
the validated “ICH Score,” described by Hemphill
et al. in 2001 [84, 112]. Radiologic predictors of
poor clinical outcome in the ICH Score include an
infratentorial origin of the ICH, an ICH volume
above 30 ml, and an intraventricular extension of
the hemorrhage, while clinical predictors include
the Glasgow Coma Scale and age at presentation.
For instance, approximately 50 % of patients scor-
ing as the lowest risk category on the ICH Score
will have favorable clinical outcomes at 12 months
18 Hemorrhagic Stroke
(total or near total recovery, “patient is able to
carry out all usual duties and activities”), while
those in the highest categories will have virtually
no chance of recovering to ambulatory status
(moderate to severe disability or death)
[112]. Other radiologic factors discussed above
such as active extravasation or growth of ICH
provide additional important prognostic informa-
tion; for example, active extravasation on a
contrast-enhanced CT such as a CTA raises the
risk of a 30-day mortality by an odds ratio of
4.7 [103].
Evolution of the Intraparenchymal
Hematoma
A more detailed description of the biophysical
factors involved in the CT and MRI appearance
of intracranial blood products has been provided
above; this section aims to provide a clinically
oriented guide to aid in understanding the evolv-
ing appearance of intracranial hemorrhage over
time. Recognizing different phases of hemorrhage
can be useful for identifying ongoing hemorrhage
requiring further management and correlating the
onset of neurologic deficits and occasionally in
medicolegal settings such as non-accidental
trauma in children.
In the minutes immediately following
extravasation, the CT and MRI appearance of
hemorrhage is similar to intravascular
blood excepting flow-related artifacts that
can cause signal void in patent blood vessels
on T2 images. On CT, this is usually manifest
as fluid collections of 40–60 HU which can be
difficult to differentiate from the adjacent
brain, although early clot products including clumps
of fibrin, platelets, and white blood cells will intro-
duce heterogeneous attenuation [113].
The hours and days following extravasation,
the “hyperacute” and “acute” phases, are charac-
terized by clot formation where the water compo-
nent is squeezed out leaving behind a matrix of
red and white blood cells, platelets, and clotting
proteins. The attenuation of these protein-
rich components is higher than water, leading to
a hyperdense collection on CT reaching up to
403
60–80 HU which gradually fades to approach
water attenuation over the coming months of
evolution. The MRI appearance is more
complex: the central portion of the hematoma
containing residual oxyhemoglobin will be T2
hyperintense and T1 hypointense, and the periph-
ery containing deoxyhemoglobin will be T2
hypointense and T1 isointense. As the proportion
of oxyhemoglobin drops and deoxyhemoglobin
prevails, the hematoma will become more uni-
formly T2 hypointense and T1 isointense to
hyperintense, especially as protein content pre-
dominates in the extracellular space favoring T1
signal gain and further T2 signal loss. Because of
mass effect, T2 hyperintense surrounding
vasogenic edema is usually observed until the
brain parenchyma has remodeled around the
mass effect [114].
As the clot continues to contract and the hema-
toma progresses into the “early subacute” phase
(typically several days), deoxyhemoglobin is oxi-
dized into methemoglobin within the still-intact
red blood cells; the presence of methemoglobin
dramatically increases the T1 signal of the clot,
but because the methemoglobin is sequestered in
intact red blood cells, the field inhomogeneity
between intracellular and extracellular spaces
leads to marked T2 hypointensity. In the “late
subacute” phase several days to a month follow-
ing the initial bleed, the red blood cells finally
lyse, allowing the methemoglobin to distribute
evenly throughout the clot, and finally both T1
and T2 hyperintensity prevail.
In the months and years of the “chronic” phase,
macrophages will gradually digest the clot, reduc-
ing its volume and transforming extracellular met-
hemoglobin into hemosiderin; in this phase, the
clot is hypointense on T1 and markedly
hypointense on T2 due to susceptibility effects
of the hemosiderin (Fig. 17).
As evidenced above, the appearance of the
ICH at different periods of time depends consid-
erably upon a number of factors. For instance, in
early phases, the hematocrit and protein levels of
the hematoma will dramatically alter the CT atten-
uation in the hematoma. In later phases, factors
such as oxygen tension at the hematoma will
determine how quickly deoxyhemoglobin
404 J. Griauzde et al.

Late subacute (days-month)

hypointense ← T2 signal → hyperintense

Extracellular met-Hb
Lysed RBC’s

Hyperacute (<12 h)
oxy-Hb
Intact RBC’s

Acute (hours-days)
deoxy-Hb
Intact RBC’s
Early subacute (days)
Chronic (month-years) Intracellular met-Hb
Hemosiderin and ferritin Intact RBC’s
Macrophage digested clot

hypointense ← T1 signal → hyperintense

Fig. 17 Evolution of blood products on MRI (Derived from a figure created by Dr. Frank Gaillard as presented on http://
radiopaedia.org/articles/ageing-blood-on-mri, with permission)

transitions into methemoglobin and how quickly
red blood cells finally lyse and decrease the field
inhomogeneity effects of sequestered methemo-
globin. The integrity of the blood-brain barrier
also helps to determine the degree to which
hemosiderin-laden macrophages remain trapped
in the parenchyma causing hemosiderin staining
long after the vast majority of the hematoma mass
has been resorbed [113].
Extra-axial Central Nervous System
Hemorrhage
Subarachnoid Hemorrhage (SAH)
The most common cause of atraumatic SAH in the
adult population is a ruptured intracranial aneu-
rysm. Other causes include secondary leakage of
blood from a primary intraparenchymal hemor-
rhage, trauma, intracranial AVM, hemorrhagic
tumor, dural AVM, intracranial dissection, vascu-
litis, mycotic aneurysm, blood dyscrasias, amy-
loid angiopathy, moyamoya disease,
complications of pregnancy, and spinal cord
AVM [115, 116]. Fifteen percent of patients with
SAH have no aneurysm found and no other cause
noted after four vessel angiography [117].
SAH is best identified on non-enhanced CT of
the brain. This shows hyperdensity in the CSF
spaces surrounding the brain parenchyma and
may often be seen layering in the ventricles
when there is intraventricular extension
(Fig. 18). High-density blood products in acute
SAH are identified within cisterns and subarach-
noid spaces in 90–97 % of patients within the first
24 h following SAH [118, 119]. The sensitivity
for SAH decreases to 80 % within 3 days, 70 %
within 5 days, 50 % within 1 week, and 30 %
within 2 weeks. CT is useful in directing which
blood vessel should be injected first for the
angiogram.
18 Hemorrhagic Stroke
Fig. 18 A 45-year-old female who presents with worst
headache of her life. Axial NCCT of the brain shows
diffuse high-attenuation material with the subarachnoid
space surrounding the midbrain, within the sylvian fis-
sures, and interhemispheric fissure
The diagnosis of SAH on MRI may be first
identified using fluid-attenuated inversion recov-
ery (FLAIR) or susceptibility-weighted (SW)
imaging [120, 121]. On FLAIR images, SAH
can be seen as high signal intensity in the sulci
due to blood protein shortening the T2 of the CSF
[122, 123]. On conventional T1-WI and T2-WI,
acute SAH is not visible. The oxygen pressure of
CSF is approximately 43 mmHg. Given the sig-
moid shape of the oxyhemoglobin dissociation
curve, this gives a saturated hemoglobin of
approximately 72 %. Therefore, only 28 % of
the hemoglobin is in the deoxygenated form.
The hematocrit of a SAH is usually about 5 %,
and thus the low hematocrit together with the
small amount of deoxyhemoglobin makes acute
hemorrhage difficult to detect on MR [124]. Oxy-
hemoglobin has no unpaired electrons and there-
fore behaves like other nonparamagnetic
substances, being isointense to the brain on
T1-WI and T2-WI. A majority of the time,
RBCs are cleared from the CSF so that normally
405
one does not detect subarachnoid methemoglobin
if the patient is imaged a few days after an ictus.
MRI and MRA can be helpful in a patient with a
SAH with multiple aneurysms, since the aneu-
rysm that bled may have high intensity of methe-
moglobin in the adjacent sulci or aneurysm wall
[125–127].
The rate of rehemorrhage from a ruptured
aneurysm has been reported to be between 4 %
and 20 % within the first 24 h [128, 129]. The rate
of rupture then decreases to between 1 % and 2 %
per day for the next 2 weeks. Vasospasm occurs at
3–15 days after SAH and is the chief cause of
significant morbidity and mortality in SAH.
Symptomatic vasospasm can be treated with
hypervolemia, hypertension, hemodilution,
nimodipine, intra-arterial nicardipine, and balloon
angioplasty [130, 131]. Cerebral aneurysms can
be treated by surgical clip placement of
endovascular coil embolization [132].
Nonaneurysmal Perimesencephalic
SAH (NPSAH)
Patients with NPSAH have a similar presentation to
traditional SAH patients with sudden onset of the
“worst headache of their life.” The radiologic
criteria to diagnose NPSAH consist of (1) hemor-
rhage centered immediately anterior to the mid-
brain with or without extension of blood into the
anterior part of the ambient cistern or to the basal
part of the sylvian fissure, (2) no blood filling the
anterior interhemispheric fissure and no extension
to the lateral sylvian fissure (except for minute
amounts of blood), and (3) the absence of intraven-
tricular hemorrhage [133]. These cases have an
uncomplicated clinical course [134]. The cause of
NPSAH is not known although the current belief is
that it may result from a capillary or venous
rupture [135].
Subdural and Epidural Hematomas
Subdural and epidural hematomas can be readily
identified by T1-WI and T2-WI MR images,
because the contrast between the calvarium and
406
Fig. 19 Axial NCCT shows a (a) crescentic extra-axial
high-attenuation area along the right cerebral hemisphere,
acute subdural hematoma; (b) crescentic extra-axial inter-
mediate attenuation area along the right cerebral
Fig. 20 Axial T1-WI (a)
and T2-WI (b) MR images
show crescentic extra-axial
material predominantly
high signal on T1, with
areas of low and high signal
of T2 over the left cerebral
hemisphere, consistent with
predominantly subacute
subdural hematoma
blood is high [136]. Precise compartmentalization
of these extra-axial collections on MR images can
be determined by traditional criteria that are based
on CT (Fig. 19). Isodense subdural hematomas on
CT can be easily deciphered on MR because these
subacute hematomas typically contain methemo-
globin which is hyperintense on MR (Fig. 20). The
signal intensity patterns of evolving subdural
hematomas on conventional MR have been
reviewed in an earlier portion of this chapter and
appear similar to intraparenchymal hematomas in
acute and subacute stages [137, 138]. Chronic sub-
dural hematomas are markedly hyperintense on
MR due to the methemoglobin content. As chronic
J. Griauzde et al.
hemisphere consistent, subacute subdural hematoma; and
(c) crescentic extra-axial low-attenuation material (HU =
2) along the right cerebral hemisphere, chronic subdural
hematoma
subdural hematomas evolve, there is breakdown of
the methoglobin with a gradual decrease in the
signal intensity on T1-WI images. Chronic sub-
durals can be only minimally hyperintense to CSF
on T1-WI MR, reflecting either a very low concen-
tration of paramagnetic methemoglobin or a high
concentration of nonparamagnetic protein.
Rebleeding into a subdural collection should be
suspected if irregular hypointense membranes and
loculating areas of different signal are found or if
there are debris-fluid levels seen within the collec-
tion. In cases of suspected child abuse, subdural
hematomas typically have a crescentic appearance
and cross suture lines extending across a greater
18 Hemorrhagic Stroke
Fig. 21 (a) Axial non-contrast head CT shows a convex
high-attenuation area within the right posterior temporal
lobe, acute epidural hematoma, and (b) convex
Table 4 Causes of subdural and epidural hematomas
Trauma
Dural-based metastases
Meningioma
Dural intracranial (or intraspinal) vascular malformation
Superficial intraparenchymal hemorrhage from any cause
Bleeding diathesis due to disease
Iatrogenic coagulopathy
Peripheral aneurysm
area of the brain surface. In contrary, epidural
hematomas have a convex appearance and do not
cross suture lines involving a smaller portion of the
brain surface (Fig. 21). MR plays an important role
in detecting multiple sites of intracranial hemor-
rhage at different stages of evolution. The clinical
etiologies of subdural and epidural hematomas are
shown in Table 4.
Summary
The role of the neuroimager in evaluating CT and
MR studies in suspected cases of hemorrhagic
stroke is to recognize and localize the hemorrhage
and attempt to determine its etiology. Understand-
ing the limitations of CT in the evaluation of
hemorrhagic stroke and how signal intensity
varies on MRI based on the stage of hemorrhage
407
intermediate attenuation area within the right posterior
temporal lobe, subacute epidural hematoma
is important in determining the age and etiology of
the hemorrhage. In the future, MR will play a
larger role in the evaluation of acute hemorrhagic
stroke.
References
1. Broderick J, Connolly S, Feldmann E et al (2007)
Guidelines for the management of spontaneous intra-
cerebral hemorrhage in adults: 2007 update: a guide-
line from the American Heart Association/American
Stroke Association Stroke Council, High Blood Pres-
sure Research Council, and the Quality of Care and
Outcomes in Research Interdisciplinary Working
Group. Stroke 38(6):2001–2023
2. Adams HP Jr, del Zoppo G, Alberts MJ et al (2007)
Guidelines for the early management of adults with
ischemic stroke: a guideline from the American Heart
Association/American Stroke Association Stroke
Council, Clinical Cardiology Council, Cardiovascu-
lar Radiology and Intervention Council, and the Ath-
erosclerotic Peripheral Vascular Disease and Quality
of Care Outcomes in Research Interdisciplinary
Working Groups: the American Academy of Neurol-
ogy affirms the value of this guideline as an educa-
tional tool for neurologists. Stroke 38(5):1655–1711
3. Scott WR, New PF, Davis KR, Schnur JA (1974)
Computerized axial tomography of intracerebral and
intraventricular hemorrhage. Radiology 112(1):73–80
4. Greenberg JO, Skubick DL (1977) Unexpected brain
hemorrhages and the value of computerized tomogra-
phy. Comput Tomogr 1(4):349–357
5. Hounsfield GN (1973) Computerized transverse axial
scanning (tomography). 1. Description of system. Br J
Radiol 46(552):1016–1022
408
6. Weinstein MA, Duchesneau PM, MacIntyre WJ
(1977) White and gray matter of the brain differenti-
ated by computed tomography. Radiology 122(3):
699–702
7. New PFJ, Aronow S (1976) Attenuation measure-
ments of whole-blood and blood fractions in com-
puted tomography. Radiology 121(3):635–640
8. Norman D, Price D, Boyd D, Fishman R, Newton TH
(1977) Quantitative aspects of computed tomography
of blood and cerebrospinal-fluid. Radiology
123(2):335–338
9. Bergstrom M, Ericson K, Levander B, Svendsen P,
Larsson S (1977) Variation with time of the attenua-
tion values of intracranial hematomas. J Comput
Assist Tomogr 1(1):57–63
10. Dolinskas CA, Bilaniuk LT, Zimmerman RA, Kuhl
DE (1977) Computed tomography of intracerebral
hematomas. I. Transmission CT observations on
hematoma resolution. AJR Am J Roentgenol 129(4):
681–688
11. Tans JT (1977) Computed tomography of intracere-
bral hematoma. Clin Neurol Neurosurg 79(4):
285–295
12. Messina AV, Chernik NL (1976) Computed tomogra-
phy – resolving intracerebral hemorrhage. Radiology
118(3):609–613
13. Zhu XL, Chan MS, Poon WS (1997) Spontaneous
intracranial hemorrhage: which patients need diag-
nostic cerebral angiography? A prospective study of
206 cases and review of the literature. Stroke
28(7):1406–1409
14. Broderick JP, Brott TG, Duldner JE, Tomsick T,
Huster G (1993) Volume of intracerebral hemorrhage.
A powerful and easy-to-use predictor of 30-day mor-
tality. Stroke 24(7):987–993
15. Gado M, Phelps M (1975) The peripheral zone of
increase density in cranial computed tomography.
Radiology 117(1):71–74
16. Zacharia TT, Nguyen DT (2010) Subtle pathology
detection with multidetector row coronal and sagittal
CT reformations in acute head trauma. Emerg Radiol
17(2):97–102
17. Smith WP Jr, Batnitzky S, Rengachary SS (1981)
Acute isodense subdural hematomas: a problem in
anemic patients. AJR Am J Roentgenol 136(3):
543–546
18. Schenck JF (1996) The role of magnetic susceptibility
in magnetic resonance imaging: MRI magnetic com-
patibility of the first and second kinds. Med Phys
23(6):815–850
19. Saini S, Frankel RB, Stark DD, Ferrucci JT Jr (1988)
Magnetism: a primer and review. AJR Am J
Roentgenol 150(4):735–743
20. Thulborn KR, Brady TJ (1989) Iron in magnetic
resonance imaging of cerebral hemorrhage. Magn
Reson Q 5(1):23–38
21. Bloembergen N (1948) Nuclear magnetic relaxation.
M. Nijhoff, The Hague
J. Griauzde et al.
22. Koenig SH, Brown RD 3rd, Lindstrom TR (1981)
Interactions of solvent with the heme region of met-
hemoglobin and fluoro-methemoglobin. Biophys J
34(3):397–408
23. Gomori JM, Grossman RI (1988) Mechanisms
responsible for the MR appearance and evolution of
intracranial hemorrhage. Radiographics 8(3):
427–440
24. Thulborn KR, Waterton JC, Matthews PM, Radda GK
(1982) Oxygenation dependence of the transverse
relaxation time of water protons in whole blood at
high field. Biochim Biophys Acta 714(2):265–270
25. Ingram VM (1961) Hemoglobin and its abnormali-
ties. Thomas, Springfield, Ill
26. Pauling L, Coryell CD (1936) The magnetic proper-
ties and structure of hemoglobin, oxyhemoglobin and
carbonmonoxyhemoglobin. Proc Natl Acad Sci
U S A 22(4):210–216
27. Dickerson RE, Geis I (1983) Hemoglobin: structure,
function, evolution, and pathology. Benjamin/Cum-
mings Publication, Menlo Park
28. Di Chiro G, Brooks RA, Girton ME et al (1986)
Sequential MR studies of intracerebral hematomas
in monkeys. AJNR Am J Neuroradiol 7(2):193–199
29. Matz P, Turner C, Weinstein PR, Massa SM, Panter
SS, Sharp FR (1996) Heme-oxygenase-1 induction
in glia throughout rat brain following experimental
subarachnoid hemorrhage. Brain Res 713(1–2):
211–222
30. Thulborn KR, Sorensen AG, Kowall NW et al (1990)
The role of ferritin and hemosiderin in the MR appear-
ance of cerebral hemorrhage: a histopathologic bio-
chemical study in rats. AJNR Am J Neuroradiol
11(2):291–297
31. Bizzi A, Brooks RA, Brunetti A et al (1990) Role of
iron and ferritin in MR imaging of the brain: a study in
primates at different field strengths. Radiology
177(1):59–65
32. Gillis P, Koenig SH (1987) Transverse relaxation of
solvent protons induced by magnetized spheres:
application to ferritin, erythrocytes, and magnetite.
Magn Reson Med 5(4):323–345
33. Zimmerman RD, Heier LA, Snow RB, Liu DP, Kelly
AB, Deck MD (1988) Acute intracranial hemorrhage:
intensity changes on sequential MR scans at 0.5 T.
AJR Am J Roentgenol 150(3):651–661
34. Clark RA, Watanabe AT, Bradley WG Jr, Roberts JD
(1990) Acute hematomas: effects of deoxygenation,
hematocrit, and fibrin-clot formation and retraction on
T2 shortening. Radiology 175(1):201–206
35. Atlas SW, DuBois P, Singer MB, Lu D (2000) Diffu-
sion measurements in intracranial hematomas: impli-
cations for MR imaging of acute stroke. AJNR Am J
Neuroradiol 21(7):1190–1194
36. Gomori JM, Grossman RI, Yu-Ip C, Asakura T (1987)
NMR relaxation times of blood: dependence on field
strength, oxidation state, and cell integrity. J Comput
Assist Tomogr 11(4):684–690
18 Hemorrhagic Stroke
37. Gomori JM, Grossman RI, Goldberg HI, Zimmerman
RA, Bilaniuk LT (1985) Intracranial hematomas:
imaging by high-field MR. Radiology 157(1):87–93
38. Weingarten K, Zimmerman RD, Deo-Narine V,
Markisz J, Cahill PT, Deck MD (1991) MR imaging
of acute intracranial hemorrhage: findings on sequen-
tial spin-echo and gradient-echo images in a dog
model. AJNR Am J Neuroradiol 12(3):457–467
39. Edelman RR, Johnson K, Buxton R et al (1986) MR
of hemorrhage: a new approach. AJNR Am J
Neuroradiol 7(5):751–756
40. Atlas SW, Mark AS, Grossman RI, Gomori JM
(1988) Intracranial hemorrhage: gradient-echo MR
imaging at 1.5 T. Comparison with spin-echo imaging
and clinical applications. Radiology 168(3):803–807
41. Patel MR, Edelman RR, Warach S (1996) Detection
of hyperacute primary intraparenchymal hemorrhage
by magnetic resonance imaging. Stroke 27(12):
2321–2324
42. Melki PS, Mulkern RV, Panych LP, Jolesz FA (1991)
Comparing the FAISE method with conventional
dual-echo sequences. J Magn Reson Imaging 1(3):
319–326
43. Jones KM, Mulkern RV, Mantello MT et al (1992)
Brain hemorrhage: evaluation with fast spin-echo and
conventional dual spin-echo images. Radiology
182(1):53–58
44. Delgado Almandoz JE, Schaefer PW, Forero NP,
Falla JR, Gonzalez RG, Romero JM (2009) Diagnos-
tic accuracy and yield of multidetector CT angiogra-
phy in the evaluation of spontaneous
intraparenchymal cerebral hemorrhage. AJNR Am J
Neuroradiol 30(6):1213–1221
45. Wong GK, Siu DY, Abrigo JM et al (2011) Computed
tomographic angiography and venography for young
or nonhypertensive patients with acute spontaneous
intracerebral hemorrhage. Stroke 42(1):211–213
46. Tsuchiya K, Hachiya J, Mizutani Y, Yoshino A (1996)
Three-dimensional helical CT angiography of skull
base meningiomas. AJNR Am J Neuroradiol 17(5):
933–936
47. Hirai T, Korogi Y, Ono K, Uemura S, Yamashita Y
(2004) Preoperative embolization for meningeal
tumors: evaluation of vascular supply with angio-
CT. AJNR Am J Neuroradiol 25(1):74–76
48. Gross BA, Frerichs KU, Du R (2012) Sensitivity of
CT angiography, T2-weighted MRI, and magnetic
resonance angiography in detecting cerebral arterio-
venous malformations and associated aneurysms.
J Clin Neurosci 19(8):1093–1095
49. Dowlatshahi D, Wasserman JK, Momoli F
et al (2014) Evolution of computed tomography angi-
ography spot sign is consistent with a site of active
hemorrhage in acute intracerebral hemorrhage. Stroke
45(1):277–280
50. Wada R, Aviv RI, Fox AJ et al (2007) CT angiography
"spot sign" predicts hematoma expansion in acute
intracerebral hemorrhage. Stroke 38(4):1257–1262
409
51. Delgado Almandoz JE, Yoo AJ, Stone MJ et al (2010)
The spot sign score in primary intracerebral hemor-
rhage identifies patients at highest risk of in-hospital
mortality and poor outcome among survivors. Stroke
41(1):54–60
52. Demchuk AM, Dowlatshahi D, Rodriguez-Luna D
et al (2012) Prediction of haematoma growth and
outcome in patients with intracerebral haemorrhage
using the CT-angiography spot sign (PREDICT): a
prospective observational study. Lancet Neurol 11(4):
307–314
53. Delgado Almandoz JE, Kelly HR, Schaefer PW
et al (2012) CT angiography spot sign predicts
in-hospital mortality in patients with secondary intra-
cerebral hemorrhage. J Neurointerv Surg 4(6):442–447
54. Tomandl BF, Kostner NC, Schempershofe M
et al (2004) CT angiography of intracranial aneu-
rysms: a focus on postprocessing. Radiographics
24(3):637–655
55. Ng SH, Wong HF, Ko SF et al (1997) CT angiography
of intracranial aneurysms: advantages and pitfalls.
Eur J Radiol 25(1):14–19
56. Tanabe S, Uede T, Nonaka T, Ohtaki M, Hashi K
(1998) Diagnosis of cerebral arteriovenous
malformations with three-dimensional CT angiogra-
phy. J Clin Neurosci 5(Suppl):33–38
57. Chappell ET, Moure FC, Good MC (2003) Compar-
ison of computed tomographic angiography with dig-
ital subtraction angiography in the diagnosis of
cerebral aneurysms: a meta-analysis. Neurosurgery
52(3):624–631; discussion 630–621
58. Tanaka H, Numaguchi Y, Konno S, Shrier DA,
Shibata DK, Patel U (1997) Initial experience with
helical CT and 3D reconstruction in therapeutic plan-
ning of cerebral AVMs: comparison with 3D time-of-
flight MRA and digital subtraction angiography.
J Comput Assist Tomogr 21(5):811–817
59. Gullberg GT, Wehrli FW, Shimakawa A, Simons MA
(1987) MR vascular imaging with a fast gradient
refocusing pulse sequence and reformatted images
from transaxial sections. Radiology 165(1):241–246
60. Bradley WG Jr, Waluch V (1985) Blood flow: mag-
netic resonance imaging. Radiology 154(2):443–450
61. Keller PJ, Drayer BP, Fram EK, Williams KD,
Dumoulin CL, Souza SP (1989) MR angiography
with two-dimensional acquisition and three-
dimensional display. Work in progress. Radiology
173(2):527–532
62. Felmlee JP, Ehman RL (1987) Spatial presaturation: a
method for suppressing flow artifacts and improving
depiction of vascular anatomy in MR imaging. Radi-
ology 164(2):559–564
63. van Tyen R, Saloner D, Jou LD, Berger S (1994) MR
imaging of flow through tortuous vessels: a
numerical simulation. Magn Reson Med 31(2):
184–195
64. Huston J 3rd, Rufenacht DA, Ehman RL, Wiebers DO
(1991) Intracranial aneurysms and vascular
410
malformations: comparison of time-of-flight and
phase-contrast MR angiography. Radiology 181(3):
721–730
65. Siegelman ES, Charafeddine R, Stolpen AH, Axel L
(2000) Suppression of intravascular signal on
fat-saturated contrast-enhanced thoracic MR arterio-
grams. Radiology 217(1):115–118
66. Deutschmann HA, Augustin M, Simbrunner J
et al (2007) Diagnostic accuracy of 3D time-of-flight
MR angiography compared with digital subtraction
angiography for follow-up of coiled intracranial aneu-
rysms: influence of aneurysm size. AJNR Am J
Neuroradiol 28(4):628–634
67. Bosmans H, Wilms G, Marchal G, Demaerel P, Baert
AL (1995) Characterisation of intracranial aneurysms
with MR angiography. Neuroradiology 37(4):
262–266
68. Heidenreich JO, Schilling AM, Unterharnscheidt F
et al (2007) Assessment of 3D-TOF-MRA at 3.0
Tesla in the characterization of the angioarchitecture
of cerebral arteriovenous malformations: a prelimi-
nary study. Acta Radiol 48(6):678–686
69. Noguchi K, Melhem ER, Kanazawa T, Kubo M,
Kuwayama N, Seto H (2004) Intracranial dural arte-
riovenous fistulas: evaluation with combined 3D
time-of-flight MR angiography and MR digital sub-
traction angiography. AJR Am J Roentgenol 182(1):
183–190
70. Zhang H, Maki JH, Prince MR (2007) 3D contrast-
enhanced MR angiography. J Magn Reson Imaging
25(1):13–25
71. Isoda H, Takehara Y, Isogai S et al (2000) MRA of
intracranial aneurysm models: a comparison of
contrast-enhanced three-dimensional MRA with
time-of-flight MRA. J Comput Assist Tomogr
24(2):308–315
72. Maki JH, Prince MR, Londy FJ, Chenevert TL (1996)
The effects of time varying intravascular signal inten-
sity and k-space acquisition order on three-
dimensional MR angiography image quality. J Magn
Reson Imaging 6(4):642–651
73. Unlu E, Temizoz O, Albayram S et al (2006)
Contrast-enhanced MR 3D angiography in the assess-
ment of brain AVMs. Eur J Radiol 60(3):367–378
74. White PM, Wardlaw JM, Easton V (2000) Can non-
invasive imaging accurately depict intracranial aneu-
rysms? A systematic review. Radiology 217(2):
361–370
75. Locksley HB (1966) Natural history of subarachnoid
hemorrhage, intracranial aneurysms and arteriove-
nous malformations. Based on 6368 cases in the
cooperative study. J Neurosurg 25(2):219–239
76. Suzuki M, Matsui O, Kobayashi K et al (2003)
Contrast-enhanced MRA for investigation of cerebral
arteriovenous malformations. Neuroradiology 45(4):
231–235
77. Wilms G, Bosmans H, Marchal G et al (1995) Mag-
netic resonance angiography of supratentorial
J. Griauzde et al.
tumours: comparison with selective digital subtrac-
tion angiography. Neuroradiology 37(1):42–47
78. Sessa M (2008) Intracerebral hemorrhage and hyper-
tension. Neurol Sci 29(Suppl 2):S258–S259
79. Qureshi AI, Tuhrim S, Broderick JP, Batjer HH,
Hondo H, Hanley DF (2001) Spontaneous intracere-
bral hemorrhage. N Engl J Med 344(19):1450–1460
80. Mutlu N, Berry RG, Alpers BJ (1963) Massive cere-
bral hemorrhage. Clinical and pathological correla-
tions. Arch Neurol 8:644–661
81. Nighoghossian N, Hermier M, Adeleine P et al (2002)
Old microbleeds are a potential risk factor for cerebral
bleeding after ischemic stroke: a gradient-echo T2*-
weighted brain MRI study. Stroke 33(3):735–742
82. Morgenstern LB, Hemphill JC 3rd, Anderson C
et al (2010) Guidelines for the management of spon-
taneous intracerebral hemorrhage: a guideline for
healthcare professionals from the American Heart
Association/American Stroke Association. Stroke
41(9):2108–2129
83. McEvoy AW, Kitchen ND, Thomas DG (2000) Intra-
cerebral haemorrhage and drug abuse in young adults.
Br J Neurosurg 14(5):449–454
84. Hemphill JC 3rd, Bonovich DC, Besmertis L, Manley
GT, Johnston SC (2001) The ICH score: a simple,
reliable grading scale for intracerebral hemorrhage.
Stroke 32(4):891–897
85. Vernooij MW, van der Lugt A, Ikram MA et al (2008)
Prevalence and risk factors of cerebral microbleeds:
the Rotterdam Scan Study. Neurology 70(14):
1208–1214
86. Greenberg SM, Vernooij MW, Cordonnier C
et al (2009) Cerebral microbleeds: a guide to detection
and interpretation. Lancet Neurol 8(2):165–174
87. McCormick WF (1966) The pathology of vascular
(“arteriovenous”) malformations. J Neurosurg
24(4):807–816
88. Al-Shahi R, Bhattacharya JJ, Currie DG et al (2003)
Prospective, population-based detection of intracra-
nial vascular malformations in adults: the Scottish
Intracranial Vascular Malformation Study (SIVMS).
Stroke 34(5):1163–1169
89. Meyers PM, Schumacher HC, Higashida RT
et al (2010) Indications for the performance of intra-
cranial endovascular neurointerventional procedures.
A scientific statement from the American Heart Asso-
ciation Council on cardiovascular radiology and inter-
vention, Stroke council, council on cardiovascular
surgery and anesthesia, interdisciplinary council on
peripheral vascular disease, and interdisciplinary
council on quality of care and outcomes research.
J Neurointerv Surg 2(2):177–188
90. Stapf C, Mast H, Sciacca RR et al (2006) Predictors of
hemorrhage in patients with untreated brain arterio-
venous malformation. Neurology 66(9):1350–1355
91. Kidwell CS, Jahan R, Gornbein J et al (2013) A trial
of imaging selection and endovascular treatment for
ischemic stroke. N Engl J Med 368(10):914–923
18 Hemorrhagic Stroke
92. Zabramski JM, Wascher TM, Spetzler RF et al (1994)
The natural history of familial cavernous
malformations: results of an ongoing study.
J Neurosurg 80(3):422–432
93. Huisman TA (2005) Intracranial hemorrhage: ultra-
sound, CT and MRI findings. Eur Radiol 15(3):
434–440
94. Salmaggi A, Erbetta A, Silvani A, Maderna E, Pollo B
(2008) Intracerebral haemorrhage in primary and met-
astatic brain tumours. Neurol Sci 29(Suppl 2):
S264–S265
95. Kondziolka D, Bernstein M, Resch L et al (1987)
Significance of hemorrhage into brain tumors:
clinicopathological study. J Neurosurg 67(6):
852–857
96. Lansberg MG, Albers GW, Wijman CA (2007)
Symptomatic intracerebral hemorrhage following
thrombolytic therapy for acute ischemic stroke: a
review of the risk factors. Cerebrovasc Dis 24(1):
1–10
97. Lansberg MG, Thijs VN, Bammer R et al (2007) Risk
factors of symptomatic intracerebral hemorrhage
after tPA therapy for acute stroke. Stroke 38(8):
2275–2278
98. Singer OC, Humpich MC, Fiehler J et al (2008) Risk
for symptomatic intracerebral hemorrhage after
thrombolysis assessed by diffusion-weighted mag-
netic resonance imaging. Ann Neurol 63(1):52–60
99. Bousser MG (2000) Cerebral venous thrombosis:
diagnosis and management. J Neurol 247(4):252–258
100. Leach JL, Fortuna RB, Jones BV, Gaskill-Shipley MF
(2006) Imaging of cerebral venous thrombosis: cur-
rent techniques, spectrum of findings, and diagnostic
pitfalls. Radiographics 26(Suppl 1):S19–S41; discus-
sion S42-13
101. Cianfoni A, Colosimo C (2011) Problem solving in
neuroradiology: expert consult, 1st edn. Elsevier,
Philadelphia, pp 427–472
102. Tong KA, Ashwal S, Holshouser BA et al (2004)
Diffuse axonal injury in children: clinical correlation
with hemorrhagic lesions. Ann Neurol 56(1):36–50
103. Kim J, Smith A, Hemphill JC 3rd et al (2008) Contrast
extravasation on CT predicts mortality in primary
intracerebral hemorrhage. AJNR Am J Neuroradiol
29(3):520–525
104. Goldstein JN, Fazen LE, Snider R et al (2007) Con-
trast extravasation on CT angiography predicts hema-
toma expansion in intracerebral hemorrhage.
Neurology 68(12):889–894
105. Ederies A, Demchuk A, Chia T et al (2009)
Postcontrast CT extravasation is associated with
hematoma expansion in CTA spot negative patients.
Stroke 40(5):1672–1676
106. Becker KJ, Baxter AB, Bybee HM, Tirschwell DL,
Abouelsaad T, Cohen WA (1999) Extravasation of
radiographic contrast is an independent predictor of
death in primary intracerebral hemorrhage. Stroke
30(10):2025–2032
411
107. Hallevi H, Albright KC, Aronowski J et al (2008)
Intraventricular hemorrhage: anatomic relationships
and clinical implications. Neurology 70(11):848–852
108. Diringer MN, Edwards DF, Zazulia AR (1998)
Hydrocephalus: a previously unrecognized predictor
of poor outcome from supratentorial intracerebral
hemorrhage. Stroke 29(7):1352–1357
109. Sacco S, Marini C, Toni D, Olivieri L, Carolei A
(2009) Incidence and 10-year survival of intracerebral
hemorrhage in a population-based registry. Stroke
40(2):394–399
110. Zahuranec DB, Gonzales NR, Brown DL et al (2006)
Presentation of intracerebral haemorrhage in a com-
munity. J Neurol Neurosurg Psychiatry 77(3):
340–344
111. Davis SM, Broderick J, Hennerici M et al (2006)
Hematoma growth is a determinant of mortality and
poor outcome after intracerebral hemorrhage. Neurol-
ogy 66(8):1175–1181
112. Hemphill JC 3rd, Farrant M, Neill TA Jr (2009) Pro-
spective validation of the ICH Score for 12-month
functional outcome. Neurology 73(14):1088–1094
113. Parizel PM, Makkat S, Van Miert E, Van Goethem
JW, van den Hauwe L, De Schepper AM (2001)
Intracranial hemorrhage: principles of CT and MRI
interpretation. Eur Radiol 11(9):1770–1783
114. Linfante I, Llinas RH, Caplan LR, Warach S (1999)
MRI features of intracerebral hemorrhage within
2 hours from symptom onset. Stroke 30(11):
2263–2267
115. Kim YW, Lawson MF, Hoh BL (2012)
Nonaneurysmal subarachnoid hemorrhage: an
update. Curr Atheroscler Rep 14(4):328–334
116. Carvi y Nievas MN, Archavlis E (2009) Atypical
causes of nontraumatic intracranial subarachnoid
hemorrhage. Clin Neurol Neurosurg 111(4):354–358
117. Delgado Almandoz JE, Jagadeesan BD, Refai D
et al (2012) Diagnostic yield of repeat catheter angi-
ography in patients with catheter and computed
tomography angiography negative subarachnoid
hemorrhage. Neurosurgery 70(5):1135–1142
118. Gee C, Dawson M, Bledsoe J et al (2012) Sensitivity
of newer-generation computed tomography scanners
for subarachnoid hemorrhage: a Bayesian analysis.
J Emerg Med 43(1):13–18
119. Mark DG, Hung YY, Offerman SR et al (2013)
Nontraumatic subarachnoid hemorrhage in the setting
of negative cranial computed tomography results:
external validation of a clinical and imaging predic-
tion rule. Ann Emerg Med 62(1):1–10 e11
120. Verma RK, Kottke R, Andereggen L et al (2013)
Detecting subarachnoid hemorrhage: comparison of
combined FLAIR/SWI versus CT. Eur J Radiol
82(9):1539–1545
121. Wu Z, Li S, Lei J, An D, Haacke EM (2010) Evalu-
ation of traumatic subarachnoid hemorrhage using
susceptibility-weighted imaging. AJNR Am J
Neuroradiol 31(7):1302–1310
412
122. Spickler E, Lufkin R, Teresi L et al (1990) MR imag-
ing of acute subarachnoid hemorrhage. Comput Med
Imaging Graph 14(1):67–77
123. Ibrahim GM, Weidauer S, Macdonald RL (2011)
Interobserver variability in the interpretation of
computed tomography following aneurysmal sub-
arachnoid hemorrhage. J Neurosurg 115(6):
1191–1196
124. Atlas SW, Thulborn KR (1998) MR detection of
hyperacute parenchymal hemorrhage of the brain.
AJNR Am J Neuroradiol 19(8):1471–1477
125. Sailer AM, Wagemans BA, Nelemans PJ, de Graaf R,
van Zwam WH (2014) Diagnosing intracranial aneu-
rysms with MR angiography: systematic review and
meta-analysis. Stroke 45(1):119–126
126. Cirillo M, Scomazzoni F, Cirillo L et al (2013) Com-
parison of 3D TOF-MRA and 3D CE-MRA at 3T for
imaging of intracranial aneurysms. Eur J Radiol
82(12):e853–e859
127. Pierot L, Portefaix C, Rodriguez-Regent C, Gallas S,
Meder JF, Oppenheim C (2013) Role of MRA in the
detection of intracranial aneurysm in the acute phase
of subarachnoid hemorrhage. J Neuroradiol 40(3):
204–210
128. Roos YB, Beenen LF, Groen RJ, Albrecht KW,
Vermeulen M (1997) Timing of surgery in patients
with aneurysmal subarachnoid haemorrhage:
rebleeding is still the major cause of poor outcome
in neurosurgical units that aim at early surgery.
J Neurol Neurosurg Psychiatry 63(4):490–493
129. Nieuwkamp DJ, de Gans K, Algra A et al (2005)
Timing of aneurysm surgery in subarachnoid
haemorrhage–an observational study in The Nether-
lands. Acta Neurochir (Wien) 147(8):815–821
130. Velat GJ, Kimball MM, Mocco JD, Hoh BL (2011)
Vasospasm after aneurysmal subarachnoid hemor-
rhage: review of randomized controlled trials and
meta-analyses in the literature. World Neurosurg
76(5):446–454
J. Griauzde et al.
131. Kimball MM, Velat GJ, Hoh BL (2011) Participants
in the international multi-disciplinary consensus con-
ference on the critical care management of subarach-
noid H. Critical care guidelines on the endovascular
management of cerebral vasospasm. Neurocrit Care
15(2):336–341
132. Bakker NA, Metzemaekers JD, Groen RJ, Mooij JJ,
Van Dijk JM (2010) International subarachnoid aneu-
rysm trial 2009: endovascular coiling of ruptured
intracranial aneurysms has no significant advantage
over neurosurgical clipping. Neurosurgery 66(5):
961–962
133. Brinjikji W, Kallmes DF, White JB, Lanzino G, Mor-
ris JM, Cloft HJ (2010) Inter- and intraobserver agree-
ment in CT characterization of nonaneurysmal
perimesencephalic subarachnoid hemorrhage. AJNR
Am J Neuroradiol 31(6):1103–1105
134. Kong Y, Zhang JH, Qin X (2011) Perimesencephalic
subarachnoid hemorrhage: risk factors, clinical pre-
sentations, and outcome. Acta Neurochir Suppl
110(Pt 1):197–201
135. Buyukkaya R, Yildirim N, Cebeci H et al (2014) The
relationship between perimesencephalic subarach-
noid hemorrhage and deep venous system drainage
pattern and calibrations. Clin Imaging 38(3):226–230
136. Gentry LR, Godersky JC, Thompson B, Dunn VD
(1988) Prospective comparative study of intermediate-
field MR and CT in the evaluation of closed head
trauma. AJR Am J Roentgenol 150(3):673–682
137. Ganz JC, Thuomas KA, Vlajkovic S et al (1993)
Changes in intracranial morphology, regional cerebral
water content and vital physiological variables during
epidural bleeding. An experimental MR study in
dogs. Acta Radiol 34(3):279–288
138. Han JS, Kaufman B, Alfidi RJ et al (1984) Head
trauma evaluated by magnetic resonance and com-
puted tomography: a comparison. Radiology 150(1):
71–77
 Medical Therapy of Acute Stroke
19
Keith G. DeSousa and Albert S. Favate

Contents Management of Body Temperature . . . . . . . . . . . . . . . 420

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414 Neuroprotection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420
Pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414 Deep Venous Thrombosis (DVT) Prophylaxis . . . 420
Initial Assessment of the Stroke Patient . . . . . . . . . . 415 Acute Stroke Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
Acute Therapy: Hemorrhagic Stroke . . . . . . . . . . . . . 421
Acute Therapy: Ischemic Stroke . . . . . . . . . . . . . . . . . . 416
Management of Intracranial Pressure . . . . . . . . . . . . 421
Intravenous Thrombolysis . . . . . . . . . . . . . . . . . . . . . . . . . 416
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422
Antiplatelet Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422
Anticoagulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418
Statin Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419
Blood Pressure Management
in Ischemic Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419
Fluid Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419
Management of Hyperglycemia . . . . . . . . . . . . . . . . . . . 420

K.G. DeSousa (*)

Department of Neurology, NYU Langone Medical Center,
New York, NY, USA
Vascular Neurology, University of Miami, Miami,
FL, USA
e-mail: Keith.DeSousa@nyumc.org;
drkeith5@gmail.com; k.desousa@med.miami.edu
A.S. Favate
NYU Langone Medical Center, New York, NY, USA
e-mail: Albert.Favate@nyumc.org

# Springer Science+Business Media New York 2016 413

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_40
414
Abstract
Approximately 800,000 strokes occur in the
United States each year. However, first and
recurrent strokes are preventable. Stroke is
the fourth leading cause of death and a major
cause of disability in the United States (Go AS
et al, Heart disease and stroke statistics – 2014
update: a report from the American Heart
Association. Circulation 129(3):e28–e292,
2014). Recent advances in medical therapy of
stroke have resulted in greatly improved sur-
vival rates from stroke. Early treatments as
well as preventative therapies can help mitigate
the brain damage that could be caused by
stroke.
The treatment of stroke varies by the type of
stroke, i.e., ischemic or hemorrhagic. The time
since symptom onset is also critical to stroke
management. Risk factor management and
reduction also plays a key factor in stroke
prevention. This chapter aims to discuss cur-
rent medical therapy of acute stroke.
Keywords
Stroke/diagnosis • Stroke/therapy tissue plas-
minogen activator • Telemedicine • Evidence-
based medicine • Practice guidelines as topic •
Quality assurance health care/standards
Introduction
Stroke is a term given to any brain damage caused
by dysfunction in blood supply to the brain. The
term applies generally when symptoms begin
abruptly. Stroke refers to an extremely broad
range of cerebral pathology that vary consider-
ably. Rupture of a blood vessel and hemorrhage
into the cranial vault and occlusion of a miniscule
blood vessel with a small amount of tissue damage
in an important area are both classified as strokes
but the treatment varies considerably. C. Miller
Fisher, one of the pioneers of vascular neurology,
said “neurology is learned stroke by stroke.” To
treat a stroke patient, the clinician must be inti-
mately familiar with the anatomy, physiology, and
pathology of not only the brain but the spinal cord,
K.G. DeSousa and A.S. Favate
heart, and vascular structures. Every stroke patient
is an individual, and treatment must be guided
with an understanding of the basic mechanism
that leads to the incident.
Pathophysiology
Ischemic stroke results when there is a decrease in
blood flow to the brain resulting in tissue damage.
Criteria for classifying strokes by their underlying
etiology are available. The stroke subtype classi-
fication includes large artery atherosclerosis,
cardioembolism, small vessel occlusion, other
unusual but determined etiology, and
undetermined etiology or cryptogenic [1]. The
TOAST (Trial of Org 10172 in Acute Stroke
Treatment) classification system is still used in
clinical practice as well as for studies.
At a molecular level, a process called
“excitotoxicity” is responsible for cellular death
in ischemic tissue. Glutamate, which is normally
present in synaptic terminals, is cleared from the
extracellular space in an energy-dependent pro-
cess. Lack of blood flow results in decreased
levels of ATP (adenosine triphosphate) available
for clearing glutamate. The high levels of gluta-
mate cause calcium channels to remain open and
continuous membrane depolarization. There is an
influx of calcium, sodium, and chloride ions as
well as an efflux of potassium [2]. The elevated
calcium causes activation of destructive enzymes.
The osmotic gradient leads to cellular swelling.
All these events lead to a loss of cellular integrity.
This correlates with MRI (magnetic resonance
imaging) diffusion-weighted imaging (DWI) as
this sequence helps demonstrate increased intra-
cellular water. As a result of the aforementioned
cellular cascade of events, the DWI image shows
high signal within minutes of onset of an acute
ischemic stroke (Fig. 1).
The brain does not, for the most part, have its
own energy stores and relies on blood flow for its
survival. Thus, it is susceptible to damage with
even short periods of lack of blood flow. Normal
cerebral blood flow (CBF) is around 50–60
ml/100 g/min. The brain may respond to changes
in blood flow by local vasodilatation, opening
19 Medical Therapy of Acute Stroke
collaterals and/or increasing the oxygen and glu-
cose extraction from blood. Neuronal dysfunction
occurs when the CBF decreases to around
20 ml/100 g/min. Permanent damage occurs
when the CBF goes below 10 ml/100 g/min [3].
Fig. 1 Area of high signal intensity seen on diffusion-
weighted imaging suggestive of acute ischemic stroke
Fig. 2 MRI perfusion imaging. (a) CBV (cerebral blood
volume reduced in an area which likely corresponds with
infarcted tissue), (b) CBF (cerebral blood flow which is
reduced in an area larger than ischemic core), and (c) MTT
(mean transit time which demonstrates slow arrival of
415
The majority of strokes are due to a small area
of tissue that is deprived of blood flow. Often, a
single blood vessel becomes blocked. The imme-
diate region surrounding the blood vessel is most
affected – this is called the ischemic core where
cells are damaged and may die. Further away from
the core, the cells may still be receiving a small
amount of oxygen and glucose through diffusion
from collateral vessels. The CBF in these regions
is around 25–50 % of normal. The cellular integ-
rity may be preserved in these regions. These cells
can recover if blood flow is restored to them
quickly. This region is called the penumbra
[3]. Saving this tissue, the penumbra, is the goal
of reperfusion therapy. Perfusion imaging can
help visualize and separate tissue which is poten-
tially salvageable from that which has already
been infarcted (Fig. 2).
Initial Assessment of the Stroke
Patient
The initial assessment of a stroke patient requires
a broad approach as there are many other diagno-
ses which could mimic a stroke as well as stroke
patients generally have multiple comorbidities.
The basics should not be overlooked and the
patient should be assessed for airway, breathing,
contrast in the left hemisphere). The regions outside of
the infarcted area with slow arrival of contrast represent
potentially salvageable tissue amenable to reperfusion
strategies
416
and circulation. Attention should be paid to any
other condition which could easily be reversible –
such as hypoglycemia. A quick, concise history
should be obtained including last known time
normal as this will guide further therapy. A rapid
assessment to determine if a patient is a candidate
for thrombolytic therapy should be completed
including National Institutes of Health Stroke
Scale (NIHSS) and evaluation of inclusion and
exclusion criteria for tPA (tissue plasminogen
activator/brand name: Alteplase) as described
later in this chapter. Point-of-care blood glucose
should be obtained. The NIHSS, which quantifies
the neurological exam, should be performed in the
first 10 min from arrival prior to imaging if possi-
ble. Two IV lines (preferably 18 g) should be
placed in case thrombolytic therapy or if
contrast-enhanced imaging is necessary. Comput-
erized tomography (CT) should be done within
minutes of arrival and interpreted shortly thereaf-
ter. Labs including CBC (complete blood count),
CMP (comprehensive metabolic panel), PT (pro-
thrombin time), PTT (partial thromboplastin
time), INR (international normalized ratio), and
cardiac enzymes (if indicated) should be drawn.
Lytic therapy may be initiated in the imaging suite
to avoid delays. It may be started without waiting
for laboratory test results unless the patient is on
anticoagulation or has history of known bleeding
diathesis. An electrocardiogram (EKG) should be
completed as well. Next, investigations should be
tailored to determine the cause of the stroke in
each individual patient [4].
Summary
• Obtain a quick, concise history including last
known time normal, past medical history, risk
factors for stroke, and medications, and per-
form evaluation of inclusion and exclusion
criteria for tPA.
• NIHSS should be performed.
• IV lines placed.
• Non-contrast head CT should be done and rap-
idly interpreted.
• Labs should be drawn including CBC, CMP,
PT, PTT, and INR and EKG completed.
K.G. DeSousa and A.S. Favate
• Further investigations should be performed for
evaluation of stroke etiology.
Acute Therapy: Ischemic Stroke
The phrase “time is brain” was intended to empha-
size that brain tissue is rapidly lost during a stroke
and rapid evaluation and treatment is necessitated.
During each minute in which a large vessel ische-
mic stroke goes untreated, the patient can lose 1.9
million neurons, 13.8 billion synapses, and 12 km
(7 miles) of axonal fibers [5].
Candidates for acute therapy should be identi-
fied as rapidly as possible. The goal of treatment
during an ischemic stroke should be the restora-
tion of blood flow as quickly and safely as possi-
ble. The approach to acute ischemic stroke
treatment can be divided into vascular and neuro-
nal strategies. Noninterventional vascular strate-
gies focus on recanalization, increasing blood
flow to the penumbra, and avoidance of clot prop-
agation. Neuronal strategies focus on novel treat-
ments such as neuroprotection.
Intravenous Thrombolysis
The first NINDS (National Institute of Neurolog-
ical Disorders and Stroke) trial had shown that IV
tPA had improved patient functional outcome at
3 months if the patient was treated within 3 h of
symptom onset [6]. Later trials including ECASS
3 (European Cooperative Acute Stroke Study)
showed that IV tPA not only was safe but patients
benefitted from receiving the medication up to
4.5 h after stroke onset [7]. However, although
the use of IV tPA within the 3–4.5 h window is
generally accepted, it is currently not FDA (Food
and Drug Administration) approved.
Multiple prior analyses of data of patients who
received tPA have shown that each 15 min reduc-
tion in time to ignition of tPA treatment was asso-
ciated with an increase in the odds of walking
independently at discharge and being discharged
to home rather than an institution. The same
15 min reduction in time to tPA treatment was
associated with a decrease in the odds of death
19 Medical Therapy of Acute Stroke 417

Table 1 Guidelines for the use of IV tissue plasminogen activator (tPA)

Warningsa (relative
Indications Contraindications contraindications)
>18 years of age Hemorrhage on imaging Abnormal blood glucose (<50
mg/dL or >400 mg/dL)
Measurable deficit Subarachnoid hemorrhage on imaging Severe neurologic deficit
NIHSS (NIHSS > 22)
Symptom onset within Serious head or spinal trauma within the last 3 months Major and early infarct signs on
3h CT
No hemorrhage on CT History of intracranial hemorrhage Minor neurologic deficit
No contraindications Uncontrolled hypertension SBP > 185 or diastolic >110 Rapidly improving symptoms
(see next column) mmHg
Seizures at onset of stroke
Active internal bleeding
Intracranial neoplasm, AVM, or aneurysm
Known bleeding diathesis (INR > 1.7, PT > 15, heparin
within the past 48 h w/elevated aPTT
Platelet count <100,000
a
In these certain situations, benefits should be weighed against risks

before discharge and symptomatic hemorrhagic
transformation of infarction. Given these findings,
one should strive to give tPA as rapidly as possible
after identification of a potential candidate rather
than later [8].
The initial NINDS stroke trial included
patients with ischemic stroke who could be treated
with IV tPA within 3 h of symptom onset. Six
hundred twenty-four patients were randomized
to treatment with placebo or IV tPA (0.9 mg/kg
up to 90 mg, 10 % as a bolus followed by the rest
over 60 min). At 3 months a complete or near
recovery occurred in a statistically greater number
of patients assigned to IV tPA than placebo (38 %
vs. 21 %). Three-month mortality was not signif-
icantly different between the two groups. Severe
systemic hemorrhage occurred in less than 1 % of
patients [6].
ECASS 3 trial assigned 821 patients with acute
ischemic stroke to receive IV tPA or placebo.
Additional exclusion criteria were included in
this study. Patients greater than 80 years old,
NIHSS >25, those with a combination of previ-
ous stroke and diabetes, and those on anticoagu-
lants regardless of INR were excluded from
participation in the study. This was done to satisfy
safety concerns from the European regulatory
agency. Depending on local stroke center guide-
lines, the exclusion criteria from ECASS 3 may be
taken into consideration when offering tPA to
patients. The results of ECASS 3 showed that
significantly more patients had a better outcome
with IV tPA than placebo. There was no difference
in mortality between IV tPA and placebo. Symp-
tomatic intracranial hemorrhage was significantly
more likely to occur with IV tPA administration
than with placebo [7].
IV tPA should be offered to patients presenting
within the 4.5 h (3–4.5 h use is recommended by
American Heart Association/American Stroke
Association (AHA/ASA) but not FDA approved
and therefore off-label) window who meet the
inclusion and exclusion criteria. Selection of can-
didates for intravenous thrombolysis requires a
neurologic evaluation, neuroimaging, determina-
tion of eligibility in accordance with the inclusion,
and exclusion criteria (Table 1).
Today’s guidelines for tPA administration
come from the original trials designed to show
the efficacy of tPA in acute stroke patients.
Noncompliance with the guidelines has been
associated with an increase in the risk of hemor-
rhage [9]. However, strict compliance with these
criteria has excluded patients who may have
benefitted from receiving tPA. Therefore, in
extreme emergent situations, a rational individu-
alized decision may be necessary for a certain
patient.
418
Before starting IV tPA, the treating physician
should make sure that treatment is within 4.5 h of
onset of symptoms, there is a measurable
neurologic deficit, inclusion and exclusion criteria
are met, non-contrast head CT or MRI is done
and does not show hemorrhage or large infarct,
blood pressure parameters are met, IV lines are in
place, and accurate body weight has been
determined.
Serum glucose should be checked to ensure
that hypoglycemia is not a cause of the precipitat-
ing neurologic findings. Patients with low platelet
counts or elevated prothrombin times should be
excluded. Uncontrollable hypertension is also
exclusionary.
The tPA dose is calculated at 0.9 mg/kg of
actual body weight with a maximum dose of
90 mg. Ten percent of the dose is given as a
bolus over one minute. The rest is infused over
an hour. Following administration of tPA, close
monitoring of the patient is required with frequent
assessment of vital signs and neurologic status as
per local institution guidelines. A follow-up CT
should be done 24 h post-tPA administration or
sooner if there is any deterioration in
clinical exam.
Tight blood pressure control is essential during
the administration of thrombolytic therapy. The
blood pressure must be less than or equal to
185/110 mmHg before starting the medication.
Aggressive blood pressure lowering can be
achieved using a variety of antihypertensives.
Labetalol or IV nicardipine are routinely used. If
blood pressure cannot be reliably controlled,
thrombolytic therapy should not be administered.
Following administration of tPA, the blood pres-
sure needs to be kept below 180/105 for at least
24 h.
Symptomatic intracranial hemorrhage is one of
the most feared complications of thrombolytic
therapy. If such an event is suspected, discontinu-
ation of tPA is warranted. A stat non-contrast head
CT or MRI should be completed for evaluation.
Administration of medication to reverse the
effects of the thrombolytic therapy should be con-
sidered including cryoprecipitate and/or platelets.
If systemic bleeding occurs, tPA administration
should also be halted, and further workup should
K.G. DeSousa and A.S. Favate
be initiated to evaluate for the source of bleeding
and tailored to the individual circumstance [9].
Antiplatelet Agents
Timely restoration of blood flow in acute ischemic
stroke is the most effective way for saving brain
tissue that has not yet been infarcted. Again, IV
thrombolytic therapy should always be consid-
ered in patients who meet the eligibility criteria.
Early administration of aspirin has been shown
to be beneficial in large randomized controlled
trials [10]. Recently, dual antiplatelet therapy
with aspirin and clopidogrel has been shown to
have better outcomes in a Chinese population with
TIA or minor stroke [11].
The International Stroke Trial (IST) showed
that among 19,435 patients who received
300 mg aspirin, there was a significant reduction
in the number of patients who had recurrent stroke
at 14 days (2.8 % vs. 3.9 %). There was also a
reduction in the combined outcome of nonfatal
stroke or death (11.3 % vs. 12.4 %) [10].
The CHANCE (Clopidogrel in High-risk
Patients with Acute Non-disabling Cerebrovascu-
lar Events) trial randomized Chinese patients to
receive either dual antiplatelet therapy with aspi-
rin and clopidogrel or placebo plus aspirin within
24 h after the onset of TIA or minor stroke. The
study showed that after 90 days there was a sig-
nificant reduction in stroke for the aspirin and
clopidogrel group vs. the aspirin plus placebo
group (8.2 % vs. 11.7 %). The absolute risk reduc-
tion was 3.5 % (hazard ratio 0.68, 95 % CI
0.57–0.81). Hemorrhagic stroke occurrence was
similarly low in both groups (0.3 %) [11].
Anticoagulation
Anticoagulation has consistently not been shown
to be beneficial in many situations of acute ische-
mic stroke and in some studies have been shown
to have worse outcomes. Guidelines by the Amer-
ican Heart Association issued in 2013 have not
recommended anticoagulation to be routinely
used for treatment of acute ischemic stroke [9].
19 Medical Therapy of Acute Stroke
However, certain circumstances may warrant
anticoagulation such as intracardiac thrombus,
mechanical heart valves, TIA (transient ischemic
attack) with a high chance of recurrence or some
arterial dissections, etc. In these situations, ther-
apy should be tailored to the individual patient. In
patients with large territorial infarctions on imag-
ing, anticoagulation should be avoided if possible.
Also, if anticoagulation is required, therapy
should be started without a bolus as there may
be an increased risk of hemorrhagic transforma-
tion of the stroke.
Statin Therapy
Lipid-lowering therapy with a statin has been
shown to reduce the risk of recurrent stroke and
other cardiovascular events [12]. An observa-
tional study of patients who were already on statin
therapy during a stroke and continued statin ther-
apy while inpatient had improved outcomes at
discharge as well as increased survival [13]. Initi-
ation of statin therapy early during hospitalization
has been shown to increase survival [14].
Blood Pressure Management
in Ischemic Stroke
Cerebral blood flow can be controlled by chang-
ing the diameter of vessels which alters their
resistance. Larger diameters result in increased
blood volume and more cerebral blood flow.
Smaller diameters have the opposite effect.
Changes in the cerebral perfusion pressure will
also affect the blood flow. Cerebral blood flow is
under autonomic control and can self-adjust as a
response to changes in many different parameters.
Autoregulation can maintain blood flow within a
mean arterial pressure (MAP) of 60–150 mmHg.
When the MAP is outside this range, cerebral
autoregulation fails as the brain is unable to com-
pensate. Therefore, there is a risk of ischemia at
low and high pressures [15].
During an acute ischemic stroke, blood pres-
sures may be elevated by an acute sympathetic
response, and the patient has underlying chronic
419
hypertension or other stroke-mediated causes.
The elevated blood pressure may be required and
a natural protective response to help under-
perfused tissues [15]. In this case further
neurovascular imaging is required to help guide
treatment. It has been seen in multiple studies that
aggressive acute blood pressure lowering in acute
stroke has been associated with increased neuro-
logic deterioration. Sometimes patients may ben-
efit from artificially raising blood pressure if it is
too low. Caution must be taken as patients may
develop hypertensive encephalopathy from blood
pressures that are too high.
As described earlier, patients who are receiving
thrombolytic therapy should have their blood
pressure treated so that it is less than or equal to
185/110 mmHg. It should be maintained at or
below 180/105 mmHg for at least 24 h after tPA
administration.
For patients presenting with ischemic stroke
who are not candidates for reperfusion therapy,
treatment of blood pressure is not recommended
during the first 24 h unless the pressure elevation
is >220 mmHg or there is evidence of other
end-organ damage such as resulting renal failure,
aortic dissection, hypertensive encephalopathy,
eclampsia/preeclampsia, etc. In most situations,
antihypertensive medication can slowly be
restarted after the first 24 h following a stroke
unless there is specific contraindication. Chroni-
cally hypertensive patients would require a more
gradual lowering of blood pressure because of
improper cerebral blood flow autoregulation. If
treatment is required, the generally accepted
blood pressure lowering is 15 % in the first
24 h [4].
Fluid Management
Most patients with acute ischemic stroke are
euvolemic or hypovolemic. The goal should be
to maintain the patient at a euvolemic state.
Hypovolemia may result in hypoperfusion of
ischemic tissue and thus worsen the infarct.
Hypervolemia can worsen cerebral edema. The
hypovolemic patient should be rapidly fluid resus-
citated if possible (care must be taken in heart
420
failure patients) followed by maintenance fluids.
Euvolemic patients can be maintained based on
body weight and estimated fluid losses. Normal
saline is recommended because it is distributed
evenly into the extracellular spaces. Half-normal
saline can also worsen cerebral edema and should
be avoided if possible [4].
Management of Hyperglycemia
Patients with diabetes have approximately twice
the risk of stroke. Diabetes affects approximately
8 % of the population of the United States. The
prevalence of diabetes in stroke patients is
approximately 15–33 %. Hyperglycemia is com-
mon during acute stroke, and multiple trials have
shown that more than 40 % of stroke patients have
hyperglycemia on admission.
Normal fasting glucose is defined as <100
mg/dL. Impaired fasting glucose is 100–125
mg/dL. A fasting plasma glucose level >126
mg/dL, hemoglobin A1C >6.5 %, and casual
plasma glucose >200 mg/dL with symptoms
meet diagnostic criteria for diabetes [4].
Multiple studies have found worse outcomes
associated with hyperglycemia. In patients who
received IV tPA with hyperglycemia, it has been
shown that there are higher rates of intracerebral
hemorrhage and worse outcomes.
Currently, there are no large-scale trials
looking at aggressive glucose control and stroke
outcomes. Currently, the guidelines state that glu-
cose should be maintained between 140 and
180 mg/dL in all hospitalized patients as per
American Diabetes Association guidelines.
There are multiple protocols for glucose control
and can vary from hospital to hospital. Care must
be taken while managing blood glucose as the
main risk from aggressive blood glucose control
result is hypoglycemia.
Management of Body Temperature
If possible stroke patients should be maintained as
close to normothermic as possible. Hyperthermia
has been associated with poor neurological
K.G. DeSousa and A.S. Favate
outcomes. Around one-third of patients with
stroke may present with concomitant temperature
>37.6  C [16]. The source of the fever should be
sought by the physician. Urinary tract infections,
pneumonia, and sepsis are all sources of fever
which could be treated. Acetaminophen or aspirin
has some effect at bringing down body tempera-
ture. Cooling blankets or cool saline infusions
may also be used [4].
There is some possible evidence that hypother-
mia may be beneficial especially in the setting of
anoxia or cardiac arrest. However, the use of
hypothermia as a method of neuroprotection is
currently being studied and results are not
available.
Neuroprotection
There have been multiple neuroprotective agents
that have shown benefit in animal models. How-
ever, none has been shown to have any benefit in
humans. There are multiple clinical trials that are
ongoing currently. One is currently looking at the
use of high-dose lovastatin given quickly after the
onset of stroke. Currently this study is still enroll-
ing patients and no data is available yet.
Deep Venous Thrombosis (DVT)
Prophylaxis
DVT or pulmonary embolism (PE) is a feared
complication of stroke. A nonfunctional,
nonmobile limb is a setup to form a clot. All
patients with acute stroke should be started on
some sort of DVT prophylaxis and it can be
started soon after a stroke if there is no evidence
of intracranial hemorrhage. DVT prophylaxis
should not be started until after 24 h and negative
imaging for hemorrhage after receiving IV tPA.
Low molecular weight heparin has been shown to
be slightly more effective at lowering the risk of
DVT/PE than unfractionated heparin in ischemic
stroke patients. IV heparin may need to be started
in the event of an acute PE or DVT followed by
long-term anticoagulation [4]. Starting DVT
19 Medical Therapy of Acute Stroke
treatments should be tailored to the patient based
on comorbidities.
Acute Stroke Units
New evidence suggests that patients who have
been hospitalized in a unit specializing in stroke
care consistently have better outcomes regardless
of the type of stroke [17]. These specialized units
have dedicated physicians trained in stroke care as
well; dedicated support staff including nurses and
physical and occupational therapists; and also
quick access to multimodal imaging including
MRI, CT, ultrasound, and diagnostic angiography
as well as trained neurointerventionalists. As per
the current guidelines, patients should be hospi-
talized in stroke units whenever possible. Emer-
gency medical services should also attempt to
bring patients to dedicated stroke centers [4].
The NIH has set recommended response times
for acute ischemic stroke. This is known as the
“golden hour” of evaluation and treatment. Within
10 min the patient should be seen by a physician
and history, lab work, and NIHSS completed. The
stroke team should be notified within 10 min. CT
scanning should be done within 25 min and labs
and CT should be interpreted within 45 min. IV
thrombolytic therapy should be initiated in less
than 60 min from arrival [4].
Acute Therapy: Hemorrhagic Stroke
Although ischemic stroke is the more common
form of stroke, hemorrhagic stroke is an important
cause of morbidity and mortality. Initial treatment
includes both medical and surgical treatment. The
most important goal should be the prevention of
hemorrhage extension and management of intra-
cranial pressure.
For all patients with intracerebral hemorrhage
on anticoagulation or antiplatelet agents, these
agents should be discontinued immediately and
reversal of their effects should be initiated.
Vitamin K, fresh frozen plasma (FFP), or pro-
thrombin complex concentrate (PCC) should be
given to patients on warfarin therapy. Protamine
421
sulfate should be considered in patients on IV
heparin therapy. Platelet transfusion is question-
able in patients with intracranial hemorrhage on
antiplatelet therapy. Platelet transfusion may be
necessary in patients with thrombocytopenia. Cer-
tain factor replacement may be necessary in
patients with severe coagulation factor deficiency.
Hypertension is often a cause of hemorrhagic
stroke and therefore blood pressure control plays a
key role in management. High blood pressure may
cause continued bleeding; however, too low blood
pressure may inhibit cerebral perfusion. The
INTERACT 2 (Intensive Blood Pressure Reduc-
tion in Acute Cerebral Hemorrhage Trial 2) trial
randomized patients to either aggressive blood
pressure lowering with a target systolic blood
pressure (SBP) <140 mmHg or traditional man-
agement with a target of SBP <180 mmHg.
Adverse events were similar in the two groups,
and there was reduced disability in patients receiv-
ing aggressive blood pressure control
[18]. Smaller studies have shown that there is
reduced hematoma expansion in patients who
underwent aggressive blood pressure manage-
ment in the setting of intracerebral hemorrhage.
Management of Intracranial Pressure
The Monroe-Kellie hypothesis recognizes that the
cranial vault is a fixed structure. The volume
inside the cranium has limited compliance. There-
fore, the volume of blood, cerebrospinal fluid
(CSF), and brain tissue should be in a state of
equilibrium. If there is an increase in one of the
components, there must be a decrease in another.
Increased intracranial pressure (ICP) can result
from the hematoma itself but also from the
resulting edema (this can also result from edema
caused by ischemic stroke). As a result, the dam-
aged brain may crush or shift normal tissue and
cause hydrocephalus or, worse, brain herniation.
Therefore, raised ICP should be managed aggres-
sively. Steps to reduce ICP can vary from simple
measures to more invasive and radical
approaches.
Elevation of the head of bed to increase venous
drainage from the head can lower ICP.
422
Hyperventilation to lower the PaCO2 to 25–30 can
produce a large lowering of ICP; however, this is
only a temporizing measure and is only effective
for a few hours. Intravenous mannitol can be used
as well. It is given as a bolus of 1 g/kg followed by
0.25–0.5 g/kg every 6 h. Hypertonic saline to
achieve a serum sodium level of around
150 mEq/L may also be used. Further manage-
ment including surgical decompression or
ventriculostomy placement may be necessary to
reduce ICP [19].
Summary
• Early treatment of stroke starts with emergency
medical services (EMS). Prompt recognition,
assessment, and transport to a qualified stroke
center are recommended.
• An organized protocol of triage and evaluation
should be in place at the stroke center. Rapid
evaluation using imaging, blood work, and a
stroke rating scale such as the NIHSS should
be completed.
• All acute ischemic stroke patients should be
evaluated for possible thrombolytic therapy.
• Blood pressure should be controlled as per
guidelines if the patient is a candidate for
thrombolytic therapy. Otherwise they should
have their blood pressure cautiously lowered.
• Limited evidence is available for the use of
anticoagulation in acute ischemic stroke and
should be avoided except for certain
circumstances.
• Oral administration of antiplatelet agents such
as aspirin is recommended for patients with
acute ischemic stroke within 24 h of onset of
stroke symptoms. Antiplatelet agents should
not be started until at least 24 h after receiving
thrombolytic therapy and follow-up imaging
has not revealed any major hemorrhage.
• Statin therapy is generally safe to be continued
in patients with acute ischemic stroke or may
be administered for secondary stroke
prevention.
• Blood sugar should be managed according to
standard guidelines.
K.G. DeSousa and A.S. Favate
• In settings of intracerebral hemorrhage, steps
should be taken to reverse coagulopathy if
indicated. Aggressive blood pressure control
appears to be safe in these patients.
• ICP management is critical management of
acute stroke patients. There are various
methods that can be used to reduce ICP and
should be tailored to the individual patient.
References
1. Adams HP Jr et al (1993) Classification of subtype of
acute ischemic stroke. Definitions for use in a multi-
center clinical trial. TOAST. Trial of Org 10172 in
acute stroke treatment. Stroke 24(1):35–41
2. Rossi DJ, Oshima T, Attwell D (2000) Glutamate
release in severe brain ischaemia is mainly by reversed
uptake. Nature 403(6767):316–321
3. Markus HS (2004) Cerebral perfusion and stroke.
J Neurol Neurosurg Psychiatry 75(3):353–361
4. Jauch EC et al (2013) Guidelines for the early manage-
ment of patients with acute ischemic stroke: a guideline
for healthcare professionals from the American Heart
Association/American Stroke Association. Stroke
44(3):870–947
5. Saver JL (2006) Time is brain–quantified. Stroke
37(1):263–266
6. (1995) Tissue plasminogen activator for acute ischemic
stroke. The National Institute of Neurological Disor-
ders and Stroke rt-PA Stroke Study Group. N Engl J
Med 333(24):1581–1587
7. Hacke W et al (2008) Thrombolysis with alteplase 3 to
4.5 hours after acute ischemic stroke. N Engl J Med
359(13):1317–1329
8. Hacke W et al (2004) Association of outcome with
early stroke treatment: pooled analysis of ATLANTIS,
ECASS, and NINDS rt-PA stroke trials. Lancet
363(9411):768–774
9. Graham GD (2003) Tissue plasminogen activator for
acute ischemic stroke in clinical practice: a meta-
analysis of safety data. Stroke 34(12):2847–2850
10. (1997) The International Stroke Trial (IST): a
randomised trial of aspirin, subcutaneous heparin,
both, or neither among 19435 patients with acute
ischaemic stroke. International Stroke Trial Collabora-
tive Group. Lancet 349(9065):1569–1581
11. Wang Y et al (2013) Clopidogrel with aspirin in acute
minor stroke or transient ischemic attack. N Engl J Med
369(1):11–19
12. Amarenco P et al (2006) High-dose atorvastatin after
stroke or transient ischemic attack. N Engl J Med
355(6):549–559
13. Blanco M et al (2007) Statin treatment withdrawal in
ischemic stroke: a controlled randomized study. Neu-
rology 69(9):904–910
19 Medical Therapy of Acute Stroke
14. Ni Chroinin D et al (2011) Association between acute
statin therapy, survival, and improved functional out-
come after ischemic stroke: the North Dublin popula-
tion stroke study. Stroke 42(4):1021–1029
15. Qureshi AI (2008) Acute hypertensive response in
patients with stroke: pathophysiology and manage-
ment. Circulation 118(2):176–187
16. Azzimondi G et al (1995) Fever in acute stroke worsens
prognosis. A prospective study. Stroke 26(11):2040–2043
423
17. Candelise L et al (2007) Stroke-unit care for acute
stroke patients: an observational follow-up study. Lan-
cet 369(9558):299–305
18. Anderson CS et al (2013) Rapid blood-pressure lower-
ing in patients with acute intracerebral hemorrhage.
N Engl J Med 368(25):2355–2365
19. Dunn LT (2002) Raised intracranial pressure. J Neurol
Neurosurg Psychiatry 73(Suppl 1):i23–i27
 Endovascular Treatment of Stroke
20
Paolo Machi and Paolo Garofalo

Contents Keywords
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
Ischemic stroke • Thrombectomy •
Synthesis Trial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425 Endovascular treatment • Stent retrievers •
Interventional Management of Stroke III . . . . . . . . . . . . 426 Clot retrievers • Tandem occlusions • Intrave-
MR Rescue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426 nous fibrinolysis • Embolectomy • Intracranial
Stroke Trials of the “New Era” . . . . . . . . . . . . . . . . . . . . . . 427
MR CLEAN Trial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
revascularization
ESCAPE Trial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
EXTEND IA Trial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
Intra-arterial Recanalization Techniques . . . . . . . . . 428 Introduction
Clot (Stent) Retrievers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429 In 2013 results of three trials comparing the effi-
Selection of Patients for Endovascular
cacy of intra-arterial approach (IAA) versus intra-
Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
Tandem Occlusions: Mechanical Recanalization venous fibrinolysis (IVF) for the treatment of the
Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430 acute ischemic stroke were published in the New
Rationale for Endovascular Treatment of Tandem England Journal of Medicine. IVF was at that
Occlusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
time the treatment of reference for acute ischemic
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437 stroke, even if several retrospective clinical series
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437 found that anatomical and clinical outcomes were
better when such treatment was associated with an
endovascular approach. These trials and their neg-
ative results have not influenced the
neurovascular community, above all, because of
several study biases.

Synthesis Trial
P. Machi (*)
Department of Neuroradiology Unit, CHU Gui de This trial [1] compared the 90-day clinical out-
Chauliac, Montpellier University Hospital, Montpellier, come of two groups of patients presenting within
France
e-mail: paolo.machi@gmail.com 4.5 h from the onset of the acute ischemic stroke
treated with IVF or IAA. One hundred and eighty-
P. Garofalo
Medicine School, University of Parma, Parma, Italy one patients assigned to the standard care arm
e-mail: paolo.garofalo89@gmail.com were administered rt -PA 2.75 h (median) after
# Springer Science+Business Media New York 2016 425
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_41
426
the stroke onset. One hundred and eighty-one
patients assigned to the intervention arm received,
by a median of 3.75 h after the stroke onset, intra-
arterial (“in situ”) fibrinolysis or mechanical clot
fragmentation or removal or a combination of
these therapies. The 90-day evaluation showed
better clinical evolution for patients assigned to
the control group. Sixty three out of 181 patients
(34.8 %) of the control group presented without
disability with a modified Rankin scale (mRS)
score 2, compared to 55 out of 181 patients
(30.4 %) of the intervention group. The study
showed that mechanical treatment was not supe-
rior to standard care in the treatment of acute
ischemic stroke. One of the biases of this trial
was that patients randomized to the intervention
arm received different approaches, considered
nowadays not effective, giving to the study a
very low level of reproducibility. In 109 patients
of the intervention arm, the treatment was
performed without the use of a “real”
thrombectomy device. This subgroup of subjects
received “in situ” intra-arterial administration of
rt-PA; in some instance, manual clot fragmenta-
tion was performed by using a micro-wire. Only
56 patients of the intervention group were treated
with a “real” clot removal device, and only
23 patients out of 163 of the intervention group
were treated with a stent retriever. Moreover,
patients of the intervention group did not receive
rt-PA while awaiting for the mechanical treatment.
Another bias of this study was the imaging used
for patient selection. The only radiological inves-
tigation performed before randomization was a
brain non-contrast-enhanced CT scan. This exam-
ination presents limitations for the assessment of
the infarct extension and vessel occlusion
identification.
Interventional Management
of Stroke III
This was a randomized trial [2] comparing out-
come of stroke patients treated within 3 h from
stroke onset by stand-alone IVF or by IVF and
IAA. This trial was stopped prematurely because
of futility after randomization of 656 subjects.
P. Machi and P. Garofalo
434 patients of this cohort received IVF, while
222 IVF plus endovascular approach. The 90-day
outcome, assessed by using the mRS, did not sig-
nificantly differed through the two groups. 40.8 %
of patients of the intervention group presented with
a mRS score 2 similarly to the 38.7 % of the
standard care group. The study showed no signif-
icant difference in terms of clinical outcome com-
paring the interventional and the standard care
groups. Again, imaging criteria selection of this
trial (as for the Synthesis trial) represented an inclu-
sion bias because, at least in the early period, no
brain vessel investigation was performed before
randomization. Furthermore, only 4 out of
434 patients of this cohort received endovascular
treatment performed with a dedicated stent
retriever. In addition, subjects of the interventional
group had to be treated within 7 h from the stroke
onset. In this group late endovascular treatment led,
in some instances, to useless recanalization.
Authors offered as explanation of the neutral
results of their study the fact that patients of the
intervention group were treated with “old-gener-
ation” thrombectomy devices and this was related
to low rate of vessel recanalization.
MR Rescue
This third trial [3] was designed, as well, to eval-
uate the 90-day outcome of stroke patients treated
with IVF or IVF associated with IAA until 8 h
after stroke. In this trial CT and MRI scan were
used to stratify randomization in favorable and
non- favorable imaging patterns. Among the
127 patients randomized, 70 undergo
thrombectomy, while 57 received IVF. Only
58 % of the treated subjects presented a favorable
penumbral pattern. In 67 % of patients of the
intervention group, an occluded arterial recanali-
zation was achieved. At the 3-month follow-up,
the mean mRS score did not differ between
patients of the two groups (3.9 vs. 3.9). Interven-
tional treatment was not superior to IVF either in
subjects with or without a favorable penumbral
pattern. In the analysis of the 90-day outcome, no
interaction was found between pretreatment imag-
ing pattern and treatment assignment.
20 Endovascular Treatment of Stroke
The study showed that a favorable penumbral
pattern on neuroimaging did not identify patients
who would benefit from IAA. Moreover,
thrombectomy was not superior to standard IFV
for the treatment of acute ischemic stroke. Never-
theless, authors suggested that probably “second-
generation” stent retrievers, due to their high effec-
tiveness in arterial reopening, would have a treat-
ment benefit in patients with a ischemic penumbra.
Stroke Trials of the “New Era”
The trial reported above failed in demonstrating the
superiority of the IAA over standard IVF for the
treatment of acute ischemic stroke. However, these
trials presented several biases; subjects randomized
to the intervention group did not receive the same
treatment and only a small number among them
were treated with “second-generation” clot
retrievers. In addition, imaging selection criteria
were not the same in all participating centers. More-
over, given that clinical evolution is related to the
delay of the artery reopening, the delay of treatment
of patients randomized to the intervention group
was in some instances too long (up to 8 h). Based
on results of these trials, other trials have been
designed in order to better evaluate comparison
between standard fibrinolysis and interventional
treatment. In these subsequent trials patients
assigned to the intervention arm were treated within
4.5 h after the stroke onset by experimented opera-
tors with dedicated clot removal devices following
radiological study showing the infarct core along
with the ischemic penumbra zone and the occluded
brain vessel, eventually.
MR CLEAN Trial
MR CLEAN [4] was a phase III, multicenter clin-
ical trial with randomized treatment group assign-
ment, open-label treatment, and blinded end-point
evaluation. This study, together with the other two
reported below, compensating for many aspects the
imbalances that were in the previous trials, con-
firmed the superiority of the interventional treat-
ment over the standard IVF. This study was
427
conducted by randomizing 500 eligible patients
presenting with acute ischemic stroke caused by
an intracranial occlusion of the distal intracranial
carotid artery, middle cerebral artery (M1/M2), or
anterior cerebral artery (A1/A2) and NIHSS score
 of 2. Patients were selected with baseline CT
scan to evaluate the ischemic core by using the
ASPECT score. CT angiography, MR angiography,
or DSA were used to evaluate the baseline vessel
imaging. Two hundred and thirty-three patients
(46.6 %) were assigned to the intervention group
and 267 patients (53.4 %) were assigned to the
control group. Mechanical treatment was
performed in 195 out of 233 patients (83.7 %)
and clot retrievers were used in 190 out of
195 (97 %) patients. A good brain vessel reperfu-
sion was obtained in 115 out of 196 (58.7 %)
patients in the intervention group. This was related
to an increase in functional independence and
reduced mortality at 90-day follow-up. The study
showed, for the first time, that the mechanical
thrombectomy was effective and safe when admin-
istered within 6 h after stroke onset to patients with
acute ischemic stroke caused by a proximal intra-
cranial occlusion of the anterior circulation.
ESCAPE Trial
ESCAPE [5] was a multicenter, prospective ran-
domized, open-label controlled trial with blinded
outcome evaluation, conceived to evaluate the
benefits obtained by rapid endovascular treatment
for acute ischemic stroke patients, who were
selected on the basis of results of CT and CT
angiography. Recruitment could occur up to 12 h
after the stroke onset. CT and CT angiography
were performed to identify subjects with a small
infarct core, defined as an ASPECT score of 6–10,
and an occluded large vessel in the anterior circu-
lation with moderate or good collateral circula-
tion, defined as the filling of 50 % or more of the
middle cerebral artery territory. A total of
316 patients underwent randomization and
divided into two groups. The intervention group
(165) underwent to rapid endovascular treatment
using clot retrievers. Among them, 151 (91.5 %)
underwent to endovascular treatment, and
428
120 (72.7 %) received intravenous rt-PA. Clot
retrievers were used in 130 out of 151 patients
(86.1 %) who underwent to an endovascular pro-
cedure. 150 patients were assigned to the control
arm and received standard of care. Both groups
received intravenous rt-PA within 4.5 h after
stroke symptom onset. Successful reperfusion
with TICI score of 2b was observed in 113 out
of 156 enrolled patients (72.4 %) in the interven-
tion group. Results of this trial showed close cor-
relation with results of MR CLEAN trial that
showed benefits and low complication rates with
endovascular treatment performed predominantly
with retrievable stents. Nevertheless, there are
main factors that distinguish ESCAPE from MR
CLEAN and the other prior trials of endovascular
treatment for stroke: first of all the use of imaging
to exclude subjects with a large infarct core and
poor collateral circulation and then a shorter inter-
val from symptom onset to treatment initiation.
EXTEND IA Trial
The EXTEND IA [6] was a multicenter, prospec-
tive, randomized, open-label, blinded end-point
study designed to evaluate whether patients with
anterior circulation stroke, with evidence of sal-
vageable tissue on perfusion imaging within 4.5 h
after the onset of stroke, will have improved reper-
fusion and early neurologic improvement when
treated by endovascular approach, within 6 h after
stroke onset, employing the Solitaire FR device
after intravenous administration of rt-PA, compared
with rt-PA alone. This study enrolled 70 eligible
patients who had an ischemic stroke due to occlu-
sion of the internal carotid artery or of the first or
second segment of the middle cerebral artery eval-
uated by CT angiography. Thirty five patients
underwent endovascular therapy while the same
number of subjects received standard care.
Endovascular therapy resulted in increased reperfu-
sion rate at 24 h (P < 0.001), and the rate of intra-
cranial recanalization after mechanical approach
was nearly 90 % versus 34 % of the control group
without symptomatic intracerebral hemorrhage, as
compared with the control group. Furthermore,
P. Machi and P. Garofalo
patients of the intervention group showed greater
early neurologic recovery at 72 h in comparison to
the control group. Some essential differences
between EXTEND IA and the previous trials have
been mentioned. They include: the use of CT per-
fusion imaging to select patients with the greatest
potential to benefit from endovascular therapy,
excluding patients with large ischemic cores and
without evidence of clinically significant recover-
able ischemic brain; shorter time from stroke onset
to treatment, 30 min compared with 100 min in the
MR CLEAN trial, which may also have contributed
to the higher proportion of patients with indepen-
dent functional outcome observed; and improved
rates of angiographic revascularization. Under the
light of the results obtained in the trials reported
above, it must be undisputed that early administra-
tion of endovascular thrombectomy without delay
after the initiation of intravenous rt-PA, among
patients with ischemic stroke with a proximal cere-
bral arterial occlusion and salvageable tissue, had
led to a better outcome in comparison to standard
care treatment.
Intra-arterial Recanalization
Techniques
Clot (Stent) Retrievers
Clot retrievers are stent-like that are navigated via
a micro-catheter through the occluded artery and
placed within the thrombus. Such devices present
a double functionality; just after deployment they
act as temporary bypass, restoring the flow
through the thrombus, and in addition, they
allow clot removal by engaging the thrombus
during the retrieving.
Different types of clot retriever tools have been
launched into the market so far. They can be
distinguished, depending on which part of the
thrombus they apply force, in proximal and distal
clot removers [7–9]. In addition, in some
instances other intracranial devices not specific
for clot removal like coronary stents or angio-
plasty balloons have been “off-label” used with
this purpose [9, 10].
20 Endovascular Treatment of Stroke
In the Multi MERCI trial, a study on MERCI
retriever, the first tool approved by the FDA as a
clot removal tool, a recanalization rate of 69.5 %
was reported when mechanical treatment was
associated with intra-arterial injection of rt-PA
[9]. Hemorrhagic complication rate of patients
who received thrombectomy was 9.8 % (16 out
of 131 patients) while perioperative adverse
events rate was 5.5 % (9 out of 131 patients).
The mean number of passes of the device
performed per procedure was 2.9. Comparison
of results obtained in studies in which the tool
used for thrombectomy was a dedicated, second-
generation clot retriever (i.e., Solitaire FR) with
results obtained in the Multi MERCI trial shows
higher target vessel recanalization rate recorded
with second-generation devices, up to 89.2 %
versus 65.5 %. Furthermore, a lower number of
device passes performed per intervention (2.9
vs. 2) along with lower rates of perioperative
complications has been reported for tools of sec-
ond generation. Even if MERCI ® retriever was
the first clot retriever device with FDA approval,
its relatively low efficacy has not encouraged
worldwide use of such tool. A second tool
launched into the market in 2007, with the
approval of the FDA, was the Penumbra system.
The first version of this device was a kit made of a
pump, a micro-catheter, and a mechanical
thrombectomy tool shaped like a little brush.
The micro-catheter had to be navigated up to the
clot and then the clot had to be fragmented with
the “brush” tool while clot fragments were aspi-
rated by a micro-catheter connected to a pump.
Clinical evaluation of this device was done with
the Penumbra Pivotal Stroke Trial [10]. This was a
multicenter registry on results of 125 subjects
treated with the Penumbra System. 81.6 % of
patients of this cohort had target vessels reopening
with a TIMI score of 2–3. Eighteen perioperative
adverse events were recorded in 16 patients (12.8
%). At 3 months of follow-up, 25 % of patients of
this series presented a mRS score of  2. The low
rate of patients with good clinical outcome at the
90-day follow-up seems not related with the
extremely high incidence of target vessel recana-
lization of this series (81.6 %). This was probably
429
due to the fact that the TIMI score, used to eval-
uate the degree of reopening of the target artery,
does not take into consideration the degree of
parenchymal reperfusion. Hence, a high degree
of the TIMI score could not be related to good
clinical recovery. Nowadays, also Penumbra Inc.
produces a stent-based dedicated clot retriever.
Intracranial self-expanding stents (not dedi-
cated stent retrievers) have been, in the past,
used as temporary or permanent endovascular
bypasses for the treatment of ischemic stroke
[10–13]. In this technique, which could be con-
sidered the background for the use of the modern
stent retrievers, a self-expanding stent was placed
through the occluding thrombus, displacing it and
allowing for blood flow restoring. This mecha-
nism facilitates removal of pro-thrombotic fac-
tors. Moreover, it allows interactions between
fibrinolytic drugs and the clot [14].
Since 2010, results of several clinical series
documenting results of Solitaire FR device used
for the endovascular treatment of acute stroke
patients have been reported in literature
[17]. Results of overall studies are concordant in
confirming the efficacy of such device, in terms of
vessel recanalization rate and 3-month clinical
outcome. Nowadays, even under the light of the
results of the trial mentioned above, stent
retrievers are considered as the device of reference
for mechanical endovascular thrombectomy.
From a technical point of view, stent-based clot
retrievers are easily advanced to the target vessel
via a micro-catheter and placed within the clot.
Hence the stent, fully expanded, is retrieved
through the brain vasculature and removed. The
efficacy of this technique is less operator depen-
dent in comparison with other endovascular pro-
cedure (i.e., the treatment of a cerebral aneurysm).
Furthermore, this technique enables reproducible
results throughout an entire interventionist team.
Procedure
Mechanical thrombectomies are performed via a
common femoral artery approach with the patient
under conscious sedation or general anesthesia.
430
A guide catheter is placed in the concerned carotid
or vertebral artery. Hence, a micro-catheter is nav-
igated distal to the “occlusion point” through the
clot over a micro-wire, and the stent is then
deployed within the thrombus. Subsequently, an
angiographic run is carried out in order to verify
its placement. Stent retrievers are usually
maintained in place before the retrieving for few
(4–5) minutes in order to obtain its best expan-
sion. Thereafter, the stent is gently pulled back to
the guiding catheter. During the retrieving, aspi-
ration through the guiding catheter can be
performed both manually and by using a pump
aspirating system. In the cases of vessel tortuosity,
a smaller guiding catheter (a coaxial catheter) can
be navigated within the first one in order to get
more stability and to reach distal intracranial ves-
sel. A coaxial catheter is usually used also in cases
of tandem occlusions; in such instance the first
guide catheter is placed in the common carotid
artery while the second one, with a tapered profile,
is navigated through the proximal occlusion.
Selection of Patients for Endovascular
Treatment
Patients who could benefit of an endovascular
treatment has to present an NIHSS >8. Patient
clinical status on admission, after the treatment,
and at discharge must be assessed using the same
scale. The 90-day clinical outcome has to be eval-
uated by using the modified Rankin scale (mRS).
A score of 2 is considered as a good clinical
outcome. Radiological imaging used for patient
selection, MRI or CT scan, has to evaluate the
presence of a large vessel occlusion, the extension
of the ischemic core, and the presence of the
penumbra zone along with the collateral supply
to the ischemic zone. The ASPECT score, evalu-
ated with DWI or CT perfusion, is currently used
to determine the extension of the ischemic core, in
cases of MCA stroke. A patient presenting with an
ASPECT score of >5 is a good candidate for
mechanical revascularization. Furthermore,
patients with tandem occlusions, involving an
intracranial occlusion associated with a cervical
common carotid occlusion, when presenting with
P. Machi and P. Garofalo
an ASPECT score of >5 are also good candidates
for the combined approach: intravenous fibrinoly-
sis and mechanical recanalization. This dual
approach, in case of tandem occlusion or terminal
“T” carotid occlusion, leads to a higher rate of
vessel recanalization in comparison with stand-
alone fibrinolysis.
Tandem Occlusions: Mechanical
Recanalization Techniques
Intravenous fibrinolysis showed a very low rate of
target vessel recanalization in cases of carotid
bifurcation (carotid “T”) or extra- and intracranial
tandem occlusions [17, 27]. Consequently, more
“invasive” approaches, such as intra-arterial fibri-
nolysis or mechanical thrombectomy, have been
associated with standard intravenous treatment.
The efficacy and the safety of these endovascular
approaches to tandem occlusions stroke have been
confirmed by several clinical series.
When faced with extra- and intracranial tan-
dem occlusions, operators have to evaluate if the
first target of the mechanical treatment should be
the extracranial or the intracranial occlusion.
Authors of the present chapter described a tech-
nique in which the intracranial thrombectomy was
performed as first stage, and then the proximal
cervical occlusion treatment was tapered
depending on the type of underlying etiology
(Fig. 1). In this setting the proximal cervical
carotid thrombosis is firstly partially treated by
angioplasty and manual aspiration in order to
allow the advancement of the guide catheter, or
of the coaxial guiding catheter, beyond the prox-
imal occlusion point. After the positioning of the
guide catheter, the intracranial thrombus is
removed by clot retrieving. In cases in which the
proximal occlusion is due to a dissection, the
guide catheter is advanced over a micro-catheter
through the true lumen of the dissection. Thereaf-
ter, the guide catheter is navigated above the dis-
sected segment and the intracranial clot is
removed as previously described. The theoretical
purpose of this approach is the immediate revas-
cularization of the intracranial vessel in order to
obtain immediate reperfusion of the ischemic
20 Endovascular Treatment of Stroke
Fig. 1 (a–f) DWI shows the extension of the ischemic core
brain zone. Thereafter, the treatment of the extra-
cranial occlusion and the underlying stenosis is
performed, eventually. The stenting of the cervical
carotid is performed only in cases of tight stenosis
or unstable ulcerated plaque. For cases in which
stand-alone angioplasty is sufficient to restore
normal flow within the artery, the treatment of
the cervical stenosis, endovascular or surgical, is
scheduled after the acute phase. Such “staged”
procedure reduces the need for antiplatelet medi-
cations and the risk of bleeding of the ischemic
core in the immediate post-acute phase. In cases of
acute carotid stenting, patients are given intrave-
nous aspirin (250–500 mg) intraoperatively. In
cases of carotid dissection, the stenting of the
dissected segment is performed when collateral
supply to the dissected artery, via the circle of
Willis, is not present or adequate (Figs. 2, 3, 4,
5, and 6).
431
Rationale for Endovascular Treatment
of Tandem Occlusions
Tandem occlusions have been previously treated
by stand-alone intravenous fibrinolysis. Neverthe-
less, efficacy of such approach in comparison with
cases of isolated M1 occlusions seems to be
reduced [18, 19]. Tandem occlusion is an inde-
pendent risk factor for poor clinical outcome for
patients treated exclusively with intravenous fibri-
nolysis [20]. Furthermore, intravenous fibrinoly-
sis presents reduced efficacy also in cases of
carotid bifurcation “T” occlusions [21]. In cases
of tandem occlusion, prompt intracranial recana-
lization is related to improved clinical outcome
regardless of persistence of the proximal cervical
occlusion [22].
Ozdemir et al. [23] published results of a study
involving eight patients presenting with tandem
432
Fig. 2 (a, b) MR
angiography shows the
presence of a proximal
carotid occlusion associated
with M1 intracranial
occlusion (tandem
occlusion)
occlusion strokes. In this series the intracranial
clot was treated by local injection of rt-PA
released through a micro-catheter navigated via
the communicating artery system to the intracra-
nial thrombus. The proximal cervical occlusion of
patients of this series was not treated and the
vessel remained occluded. This technique led to
intracranial recanalization only in two out of eight
patients, in spite of low rate of cerebral revascu-
larization; all patients with a functioning commu-
nicating artery system had favorable outcome at
3 months. Nevertheless, the persistence of the
cervical thrombotic occlusion could be a source
of potential new emboli; this is an independent
risk factor for early re-occlusion after rt-PA
[21]. Lavalée et al. [24] reported on a series of
ten patients presenting with tandem occlusion
strokes due to cervical carotid dissection.
Among them, four out of ten patients received
intravenous rt-PA, while six patients underwent
mechanical clot fragmentation. In all patients
treated by mechanical approach, a complete
recanalization of the involved vessel was
achieved. Authors stated that endovascular
P. Machi and P. Garofalo
treatment seems to be more effective than the
standard intravenous treatment. They did not
specify whether this approach would be effective
also in cases of tandem occlusions due to cervical
carotid atherosclerotic plaques.
Malik et al. [25] reported a series of 27 patients
presenting with tandem occlusions treated by
intra-arterial administration of rt-PA performed
prior to stenting of the cervical carotid. In this
series clot removal/disruption was performed
with different techniques and tools: Merci clot
retrieval device, Penumbra system, or by angio-
plasty and/or stent stenting. Patients of this series
were treated under systemic heparinization. More-
over, patients received a bolus of intravenous
eptifibatide upon stent deployment and, after the
procedure, double antiplatelet regimen with
clopidogrel and aspirin. Complete proximal and
distal recanalization was obtained in 75.3 % of
patients; hemorrhagic complications were
recorded in 10.4 % of the subjects, while the
mortality rate of this series was 24.6 %. A favor-
able clinical outcome was recorded in 41.6 % of
patients. Authors stated that the reopening of the
20 Endovascular Treatment of Stroke 433

Fig. 3 (a, b) DSA shows the occlusion of the cervical was performed with a coaxial catheter navigated through
carotid artery due to an atherosclerotic plaque. (c, d) DSA the proximal stenosis
shows the intracranial M1 occlusion; the angiographic run

cervical carotid allows immediate intracranial jails the thrombus, reducing the risk of distal
revascularization and a better positioning of the embolization. By the way the authors state that
guiding catheter. At the same time, the stent this approach could require prolonged time for
deployed at the level of the proximal stenosis stent placement and that the dual antiplatelet
434
Fig. 4 (a, b) DSA run performed with a micro-catheter
navigated within the M1 occlusion. (c, d) DSA run
performed after deployment of the stent retriever. (e, f)
therapy could result in hemorrhagic transforma-
tion especially in patients with a large stroke
extension or ineffective intracranial clot removal.
The authors suggest that a tapered stand-alone
cervical stenosis angioplasty could reduce the
risk of hemorrhagic complications.
Srinivasan et al. [26] reported a series of seven
patients presenting with tandem occlusion
strokes. In five out of seven cases, the proximal
occlusion was due to thrombosis developed on an
P. Machi and P. Garofalo
DSA run performed after retrieving of the stent retriever:
complete reopening of M1, presence of an embolic occlu-
sion of the anterior cerebral artery
atherosclerotic plaque. In the other two cases the
proximal occlusion was due to a dissection and to
the consequences of an endoartherectomy. Six
patients presented a concomitant M1 occlusion.
Complete recanalization was obtained in all cases
of M1 occlusion by “in situ” rt-PA injection
performed through a micro-catheter navigated
within the cervical occlusion. Management of
the proximal occlusion varied depending on the
type of underlying cause. Patients with cervical
20 Endovascular Treatment of Stroke 435

Fig. 5 (a, b) Stand-alone angioplasty of the proximal cervical carotid stenosis. (c, d) Final DSA control shows reopening
of the right anterior cerebral artery
436 P. Machi and P. Garofalo

Fig. 6 (a–d) Stenting of the cervical stenosis performed 15 days after the acute phase initial treatment

atherosclerotic plaque received angioplasty and
stenting. In the case of cervical dissection, the
artery was treated by deploying two covered
stents along the dissected segment. Four out of
six patients of this cohort showed a good clinical
outcome at 1-month follow-up examination. The
authors claim that in tandem occlusion stroke, the
first target to be treated should be the intracranial
occlusion because it is the cause of the neurolog-
ical symptoms. They agree with such aptitude
because in their opinion, the first purpose of the
endovascular approach to tandem occlusions
should be the prompt and complete recanalization
of the intracranial vessel [27–30]. The first
20 Endovascular Treatment of Stroke
approach to the cervical carotid is performed in
order to navigate the guiding catheter above the
proximal occlusion. Thereafter, the intracranial
occlusion is treated by using clot retriever device.
Finally, the carotid cervical occlusion/stenosis is
treated by medical management, stand-alone
angioplasty, or angioplasty and stenting. Acute
stenting is performed in cases of dissection, with-
out collateral supply, or unstable, occlusive ath-
erosclerotic plaque. In such instance, patient is
given intravenous aspirin (250–500 mg) just
upon stent placement. Stand-alone angioplasty
seems to be the best approach in cases of
nonocclusive thrombus. The stenting of the cervi-
cal carotid or the surgical treatment is in this case
performed a few days (10–15) after the acute
phase. This staged approach may reduce the risk
of hemorrhagic transformation and leads to a bet-
ter antiplatelet preparation of the patient.
Summary
In patients presenting with acute ischemic stroke
due to a large vessel occlusion, an early recanaliza-
tion of the occluded artery is related to better clin-
ical outcome. From the beginning of 2015 several
trials demonstrated improvements, in terms of ana-
tomical and clinical results, when stroke patients
were treated by standard intravenous therapy asso-
ciated with intra-arterial mechanical therapy.
Before then, recommended therapy for acute ische-
mic stroke was intravenous administration of rt-PA
within 4.5 h from symptom onset. Results of the
early randomized, controlled trials appeared in lit-
erature since 2013, comparing mechanical therapy
associated with intravenous rt-PA (standard care)
versus stand-alone intravenous rt-PA, failed to
demonstrate the superiority of the mechanical
approach over standard care.
In the present chapter they firstly report results
of randomized trials as unfavorable to mechanical
approach and discuss their limitation and biases.
Then they report about recent trials appeared in
2015 demonstrating the superiority of this
approach. Moreover, they describe clot retrieval
techniques and the endovascular treatment of tan-
dem occlusions.
437
References
1. Ciccone A, Valvassori L, Nichelatti M et al (2013)
Endovascular treatment for acute ischemic stroke. N
Engl J Med 368:904–913
2. Broderick J, Palesch Y, Demchuk A et al (2013)
Endovascular therapy after intravenous t-PA versus
t-PA alone for stroke. N Engl J Med 368:893–903
3. Kidwell C, Jahan R, Gornbein J et al (2013) A trial of
imaging selection and endovascular treatment for
ischemic stroke. N Engl J Med 368:914–923
4. Berkhemer OA, Fransen PSS, Beumer D (2015) A
randomized trial of intraarterial treatment for acute
ischemic stroke (MR CLEAN). N Engl J Med
372:11–20
5. Goyal M, Demchuk AM, Menon M (2015) Random-
ized assessment of rapid endovascular treatment of
ischemic stroke. N Engl J Med
6. Campbell BCV, Mitchell TJ, Kleining HM (2015)
Endovascular therapy for ischemic stroke with perfu-
sion imaging selection. N Engl J Med
7. Higashida RT, Furlan AJ, Roberts H et al (2003) Trial
design and reporting standards for intra-arterial cere-
bral thrombolysis for acute ischemic stroke. Stroke 34:
e109–e137
8. The IMS II Trial Investigators (2007) The Interven-
tional Management of Stroke (IMS) II study. Stroke
38:2127–2135
9. Smith WS, Sung G, Saver J et al (2008) Mechanical
thrombectomy for acute ischemic stroke: final results
of the Multi MERCI trial. Stroke 39:1205–1212
10. Penumbra PivotalStroke Trial Investigators (2009) The
Penumbra Pivotal Stroke trial: safety and effectiveness
of a new generation of mechanical devices for clot
removal in intracranial large vessel occlusive disease.
Stroke 40:2761–2768
11. Mazighi M, Serfaty JM, Labreuche J et al (2009) Com-
parison of intravenous alteplase with a combined intra-
venous –endovascular approach in patients with stroke
and confirmed arterial occlusion (RECANALISE
study): a prospective cohort study. Lancet Neurol
8:802–809
12. Gralla J, Schroth G, Remonda L et al (2006) Mechan-
ical thrombectomy for acute ischemic stroke:
thrombus-device interaction, efficiency, and complica-
tions in vivo. Stroke 37:3019–3024
13. Hussain MS, Kelly ME, Moskowitz SI et al (2009)
Mechanical thrombectomy for acute stroke with the
alligator retrieval device. Stroke 40:3784–3788
14. Levy EI, Ecker RD, Horowitz MB et al (2006) Stent-
assisted intracranial recanalization for acute stroke:
early results. Neurosurgery 58:458–463
15. Kelly ME, Furlan AJ, Fiorella D et al (2008) Recana-
lization of an acute middle cerebral artery occlusion
using a self-expanding, reconstrainable, intracranial
microstent as a temporary endovascular bypass. Stroke
39:1770–1773
16. Machi P, Costalat V, Lobotesis K et al (2012) Solitaire
FR thrombectomysystem. immediateresults in
438
56 consecutive acute ischemic stroke patients. J
Neurointerv Surg 4:62–66
17. Costalat V, Machi P, Lobotesis K et al (2011) Rescue,
combined, and stand-alone thrombectomy in the man-
agement of large vessel occlusion stroke using the
Solitaire device: a prospective 50-patient single-center
study: timing, safety, and efficacy. Stroke
42:1929–1935
18. Kim K, Garami Z, Mikulik R et al (2005) Early recan-
alization rates and clinical outcomes in patients with
tandem internal carotid artery/middle cerebral artery
occlusion and isolated middle cerebral artery occlu-
sion. Stroke 36:869–871
19. Linfante I, Llinas RH, Selim M et al (2002) Clinical
and vascular outcome in internal carotid artery versus
middle cerebral artery occlusions after intravenous tis-
sue plasminogen activator. Stroke 33:2066–2071
20. The National Institute of Neurological Disorders and
Stroke rtPA Stroke Study Group (1995) Tissue plas-
minogen activator for acute ischemic stroke. N Engl J
Med 333:1581–1587
21. Christou I, Feldeberg RA, Demchuk AM et al (2002)
Intravenous tissue plasminogen activator and improve-
ment in acute ischemic stroke patients with internal
carotid artery occlusion. J Neuroimaging 12:119–123
22. Mourand I, Brunel H, Vendrell JF et al (2010)
Endovascular stent-assisted thrombolysis in acute
occlusive carotid artery dissection. Neuroradiology
52:135–140
23. Ozdemir O, Bussiere M, Leung A et al (2008) Intra-
arterial thrombolysis of occluded middle cerebral
artery by use of collateral pathways in patients with
P. Machi and P. Garofalo
tandem cervical carotid artery/ middle cerebral artery
occlusion. AJNR Am J Neuroradiol 29:1596–1600
24. Lavallée PC, Mickaël M, Saint-Maurice JP et al (2007)
Stent-assisted endovascular thrombolysis versus intra-
venous thrombolysis in internal carotid artery dissec-
tion with tandem cerebral artery occlusion. Stroke
38:2270–2274
25. Malik AM, Nirav AV, Ridwan L et al (2011)
Endovascular treatment of extracranial/intracranial
anterior circulation occlusion. Stroke 42:1653–1657
26. Srinivasan A, Goyal M, Stys P, Sharma M, Lum C
(2006) Microcatheter navigation and thrombolysis in
acute symptomatic cervical internal carotid occlusion.
AJNR Am J Neuroradiol 27:774–779
27. Rubiera M, Ribo M, Delgado-Mederos R et al (2006)
Tandem internal carotid artery/middle cerebral
artery occlusion: an independent predictor of poor out-
come after systemic thrombolysis. Stroke
37:2301–2305
28. Zaidat OO, Suarez JI, Sunshine JL et al (2005) Throm-
bolytic therapy of acute ischemic stroke: correlation of
angiographic recanalization with clinical outcome.
AJNR Am J Neuroradiol 26:880–884
29. Kimura K, Iguchi Y, Shibazaki K et al (2008) Large
ischemic lesions on diffusion-weighted imaging done
before intravenous tissue plasminogen activator throm-
bolysis predicts a poor outcome in patients with acute
stroke. Stroke 39:2388–2391
30. Malgorzata M, Michels P, Jensen AM et al (2008)
Endovascular treatment in proximal and intracranial
carotid occlusion 9 hours after symptom onset. Neuro-
radiology 50:599–604
 Sickle Cell Disease and Stroke
21
Akifumi Fujita, Chie Asai, Yu-Ming Chang, Nadja Kadom,
Martin H. Steinberg, Naoko Saito, and Osamu Sakai

Contents Blood Transfusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440
Etiology, Classification, and Pathophysiology . . . . . . 441
Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442
Clinical Manifestations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442
Painful Episodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442
Sickle Cell Acute Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442
Infarcts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
Cognitive Deficits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
Hydroxyurea (HU) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
Hematopoietic Cell Transplantation . . . . . . . . . . . . . . . . . 444
Neurovascular Complications Concept and
Imaging Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444
Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444
Moyamoya Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449
Extracranial Vasculopathy . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
Posterior Reversible Encephalopathy
Syndrome (PRES) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 454
Other Sickle Cell Disease Imaging
Manifestations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 454
Craniofacial Bone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 454
Temporal Bone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456

A. Fujita • C. Asai • Y.-M. Chang • N. Kadom Advanced Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457

Departments of Radiology, Boston University, Boston Diffusion Tensor Imaging (DTI) . . . . . . . . . . . . . . . . . . . . . 457
Medical Center, Boston, MA, USA Perfusion Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459
e-mail: akifuji@jichi.ac.jp; chie_asai1116@yahoo.co.jp; MR Spectroscopy (MRS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 460
ychangneuro@gmail.com; nadja.kadom@bmc.org Quantitative MRI (qMRI) . . . . . . . . . . . . . . . . . . . . . . . . . . . 460
Nuclear Medicine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461
M.H. Steinberg
Departments of Medicine, Pediatrics, and Pathology and Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461
Laboratory Medicine, Boston University, Boston Medical Prevention of Primary Stroke . . . . . . . . . . . . . . . . . . . . . . . . 461
Center, Boston, MA, USA Prevention of Secondary Stroke . . . . . . . . . . . . . . . . . . . . . 462
e-mail: mhsteinb@bu.edu
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463
N. Saito
Department of Diagnostic Radiology, Saitama Medical References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463
University International Medical Center, Saitama, Japan
e-mail: naoko.saito.ns@gmail.com
O. Sakai (*)
Departments of Radiology, Boston University, Boston
Medical Center, Boston, MA, USA
Department of Otolaryngology – Head and Neck Surgery,
Boston University, Boston Medical Center, Boston, MA,
USA
Department of Radiation Oncology, Boston University,
Boston Medical Center, Boston, MA, USA
e-mail: osamu.sakai@bmc.org

# Springer Science+Business Media New York 2016 439

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_11
440 A. Fujita et al.

• Vasculopathy • Vaso-occlusion • Hemolytic

Abstract
anemia • Stroke • Ischemic stroke • Hemor-
Sickle cell disease (SCD) is a hereditary disor-
rhagic stroke • Silent cerebral infarct (SCI) •
der that is so named for the presence of char-
Cerebral fat embolism • Moyamoya syndrome
acteristic malformed sickle-shaped red blood
• Posterior reversible encephalopathy syn-
cells (RBCs). These abnormal RBCs are short-
drome (PRES) • Bone infarction • Osteomye-
lived and can also cause vaso-occlusion
litis • Labyrinthine hemorrhage (LH) •
throughout the body, which may lead to ische-
Labyrinthitis ossificans (LO) • Arterial spin
mia and infarcts in multiple organ systems.
labeling (ASL) • Positron emission tomogra-
Neurovascular complications of SCD include
phy (PET) • Single-photon emission computed
cerebral vasculopathy, such as stroke and
tomography (SPECT) • Transcranial Doppler
neurocognitive deficits, which are significant
(TCD)
causes of morbidity and mortality in both chil-
dren and adults. In current clinical practice,
various conventional imaging tools, such as
Introduction
computed tomography (CT), magnetic reso-
nance imaging (MRI), and catheter angiogram,
Sickle cell disease (SCD) is an inherited hemato-
still play an important role in diagnosis and
logic disorder characterized by the formation of
follow-up of neurovascular complications of
misshapen red blood cells (RBCs) when sickle
SCD. Ultrasounds, including transcranial
hemoglobin becomes deoxygenated and polymer-
Doppler (TCD) ultrasound, are useful in
izes. SCD causes a broad spectrum of clinical
screening for intra- and extracranial
manifestations in various organ systems and is
vasculopathy. Advanced imaging techniques,
associated with significant morbidity and mortal-
such as diffusion tensor imaging (DTI), perfu-
ity in children and adults. Imaging plays a crucial
sion imaging, MR spectroscopy, quantitative
role in the detection and description of the extent
MRI (qMRI), and nuclear medicine such as
of involvement of SCD in the brain, head and
single-photon emission computed tomography
neck, and spine.
(SPECT) and positron emission tomography
Neurovascular complications, such as cerebro-
(PET), are emerging as diagnostic tools in the
vascular ischemia/infarcts, hemorrhage, various
evaluation of SCD.
intra- and extracranial vasculopathy, and cogni-
In this chapter, a brief overview of the eti-
tive impairment, are among the most significant
ology, classification, pathophysiology, epide-
complications in SCD. SCD also affects the osse-
miology, clinical manifestation, and treatment
ous structures of the head and neck, including the
of SCD is provided. The complications of SCD
inner ears. Neuroimaging plays an important role
seen across the spectrum of conventional to
in diagnosing neurological complications of SCD
advanced imaging modalities are also illus-
as well as in monitoring the effects of medical and
trated and discussed. The focus is on
surgical treatments. Currently, neuroimaging of
neurovascular complications and its manage-
SCD largely relies on various conventional imag-
ment, but there will also be a discussion of
ing modalities such as computed tomography
posterior reversible encephalopathy syndrome,
(CT), magnetic resonance imaging (MRI), Dopp-
craniofacial osseous complications, and inner
ler ultrasound, and catheter angiogram. Recently,
ear complications, which may be encountered
quantitative MRI (qMRI) and functional MRI
in the daily neuroradiology practice.
techniques, such as diffusion tensor imaging
(DTI), perfusion imaging, and MR spectroscopy,
Keywords have been introduced to assess and monitor
Sickle cell disease (SCD) • Quantitative MRI patients with SCD. In this chapter, the pathophys-
(qMRI) • Diffusion tensor imaging (DTI) • iology, epidemiology, clinical features, and char-
Perfusion imaging • MR spectroscopy (MRS) acteristic neuroimaging manifestations of SCD
21 Sickle Cell Disease and Stroke
with conventional as well as advanced imaging
techniques are reviewed. Current therapeutic
approaches and management of SCD and the
potential role of imaging in management will
also be discussed.
Etiology, Classification,
and Pathophysiology
The term SCD refers to a variety of β-globin
(HBB) genotypes that all result in abnormally
shaped RBCs. The mutation underlying SCD is
an A-to-T transversion in the codon for amino
acid position 6 in the β-globin gene. Because of
this mutation, a valine residue replaces the normal
glutamic acid residue (glu6val) and HbS β-globin
chains are substituted for normal β-globin chains
[1, 2]. Each hemoglobin molecule contains two
β-globin and two α-globin subunits. Hemoglobin
with two normal β-globin chains is referred to as
hemoglobin A (HbA) and is the major hemoglo-
bin of adults. Fetal hemoglobin (HbF), a molecule
of two γ-globin and two α-globin chains, is most
prevalent in the fetus and newborns up to about
three months of age. Other variant hemoglobin
that can be present as compound heterozygotes
with HbS are hemoglobin C (HbC), hemoglobin E
(HbE), and hemoglobin D (HbD) [1, 2]. SCD is a
recessive trait, which means that patients must
carry two abnormal alleles of HBB in order for
the disease to fully manifest. Homozygosity for
the HbS gene or sickle cell anemia is the most
common genotype of SCD [1, 2]. HbSC disease is
the second most common genotype. SCD can also
be caused by compound heterozygosity for HbS
and one of the many forms of β-thalassemia.
Deoxygenation causes HbS molecules to poly-
merize, causing RBCs to assume various abnor-
mal shapes, some of which resemble a sickle. The
change in shape causes RBC membrane damage
and is associated with a decrease in RBC elastic-
ity, rendering them unable to pass through small
capillaries. Hemolytic anemia or a shortened
lifespan of the RBC is also a feature of this cellular
damage [1, 2]. Sickled cells can become irrevers-
ibly fixed in the “sickle” shape because of perma-
nent membrane damage and fail to resume normal
441
shape after oxygen tension is restored. These sick-
led cells and other RBCs can be sequestered in the
spleen and other reticuloendothelial cell-
containing organs and also lyse in the circulation,
leading to anemia.
Patients who carry one abnormal HBB allele
and a normal β-globin gene are said to have sickle
cell trait (HbAS) and have 60 % HbA and 40 %
HbS. HbAS is considered a benign entity with
only rare complications. In carriers of HbAS,
because each RBC contains 40 % HbS, polymer-
ization of HbS occurs only under special condi-
tions, like in the hypoxic, acidotic, and
hyperosmolar conditions of the renal medulla
and in the spleen of some people who are hypoxic.
One prospective study in patients with HbAS
showed that even otherwise asymptomatic
patients had a high percentage of vasculopathy,
such as arterial tortuosity, but the pathophysio-
logic consequences remain unclear [3]. Patients
with HbAS may develop hyposthenuria and occa-
sional hematuria and have a higher incidence of
thromboembolic disease.
The two major pathophysiological processes in
SCD are vaso-occlusion and hemolytic anemia
that is both intra- and extravascular [1, 2, 4].
3Vaso-occlusion can occur at both the micro-
and macrovascular levels. At the microvascular
level, the abnormal adherence of sickled RBCs
to the capillary and venule endothelium can lead
to intravascular stasis of blood flow. Stasis attracts
leukocytes and fosters platelet adherence and
degranulation, as well as fibrin deposition. This
microvascular occlusion can lead to tissue ische-
mia and infarct. Inflammation can enhance the
expression of adhesion molecules, further increas-
ing the tendency of sickled erythrocytes to adhere
to the vascular endothelium, worsening vaso-
occlusion [1, 2, 4]. At the macrovascular level,
abnormal adhesion of sickled RBCs and white
blood cells (WBCs) to vessel sites of high flow
turbulence may damage the vascular wall,
resulting in intimal hyperplasia of large vessels,
stenosis, and progressive occlusion of the large
cerebral arteries [5]. Chronic hemolytic anemia
from the accelerated breakdown of sickled RBCs
may also result in macrovascular occlusion from
intimal hyperplasia secondary to an inflammatory
442
cascade triggered by increased free hemoglobin
levels [1, 2, 4]. In addition, sickled RBCs may
adhere to damaged vessel walls, which can result
in secondary thrombosis and distal branch emboli
[5–7]. In the past, small vessel occlusion by intra-
vascular sickling and slugging of RBCs was con-
sidered to be the primary cause of strokes in SCD.
It is now understood, however, that occlusion of
small vessels accounts for about 25 % of cerebral
infarct in SCD while macrovascular complica-
tions account for 75 % [8].
Epidemiology
SCD is most common in people from sub-Saharan
Africa, South and Central America, the Caribbean
islands, the Mediterranean countries, India, Saudi
Arabia, and their descendants worldwide [1, 2]. It
is now estimated that 200,000 children affected
with SCD are born every year, approximately
2,000 children in the USA. The estimated US
prevalence ranges from 70,000 to 140,000
[4]. About 8 % of African Americans have
HbAS. In some regions of Africa, more than a
quarter of the populations are carriers; very high
gene frequencies are also found in some parts of
Saudi Arabia and Greece, in tribal populations of
India, and in Brazil [1, 2, 4]. HbAS confers some
protection from Plasmodium falciparum malaria;
the underlying protective mechanism is still sub-
ject of scientific investigation [9].
Mortality in patients with SCD has signifi-
cantly improved in high-resource countries. The
universal newborn screening for hemoglobinopa-
thies has become standard practice in such coun-
tries, enabling early diagnosis and patient
management [1, 4]. In 2000, the universal admin-
istration of the pneumococcal vaccine has further
decreased the mortality and morbidity of child-
hood invasive pneumococcal disease, including in
children with SCD. As a result of these programs,
SCD-related death fell by 42 % from 1999
through 2002 [10]. It is estimated that 94 % of
children with SCD now live to adulthood, and
more than 98 % of children with SCD in high-
resource countries are living into adulthood
[11]. This has caused a mortality shift toward
A. Fujita et al.
older ages and SCD is considered a chronic con-
dition requiring comprehensive, lifelong manage-
ment [4]. However, in low-resource countries,
more than 50 % of children younger than five
years of age die due to complication of SCD [4].
Clinical Manifestations
Painful Episodes
Pain, both acute and chronic, is a hallmark of
SCD. Acute painful episodes are the most com-
mon vaso-occlusive events with inflammatory
and ischemic consequences in patients with SCD
of all ages, although they are more common in
teenagers and young adults [1, 2, 4]. In young
children, one of the first manifestations of SCD
is the so-called hand-foot syndrome (i.e., sickle
cell dactylitis), a painful swelling of the hands and
feet due to inflammation of the metacarpal and
metatarsal periosteum. Painful episodes may
involve any bones, muscles, mesentery, and
other organs. Although acute vaso-occlusive
pain is generally self-limiting and does not
result in permanent organ damage, increased fre-
quency of pain is associated with early death in
patients with SCD who are older than 20 years
[12]. The cause of chronic pain is poorly under-
stood, but it can be debilitating and may result
from leg ulcers, avascular necrosis of bone, and
neuropathic complications [4]. An important con-
sideration for the neuroradiologist is that the skull
and maxillofacial bones may be affected by acute
pain. Typical symptoms are headaches or local-
ized pain and swelling. Frequently, these are
caused by bone infarcts with subperiosteal
hemorrhage [13].
Sickle Cell Acute Events
Besides the painful episodes described above,
patients with SCD may also experience splenic
and hepatic sequestration crisis, aplastic crisis,
acute chest syndrome (ACS), and stroke. In addi-
tion, SCD patients are more susceptible to infec-
tion and its complications, such as sepsis,
21 Sickle Cell Disease and Stroke
pulmonary embolism, pulmonary hypertension,
cardiomyopathy, and hepatic disease [1, 2, 4].
Infection with B19 parvovirus can suppress the
production of RBCs in patients with SCD for 2 to
3 days, causing an acute, life-threatening aplastic
anemia with pallor, tachycardia, and fatigue.
Severe anemia can also result acutely from
sequestration of blood in the spleen or liver
[2]. ACS is the second most common cause of
hospitalization among patients with SCD and a
leading cause of mortality [14]. ACS is an acute
lung injury that causes development of a new
alveolar pulmonary infiltrate involving at least
one lung segment. ACS is caused by a combina-
tion of pulmonary infection, fat embolism, and
intravascular pulmonary sequestration of sickled
erythrocytes, resulting in lung injury and infarct.
Recurrent ACS can lead to chronic respiratory
insufficiency. Occasionally, ACS culminates
with multi-organ failure [2].
Infarcts
Infarcts are the consequences of vaso-occlusive
crises. Infarcts may involve the brain, liver,
spleen, kidney, and bones. Infections often pre-
cede vaso-occlusive episodes in children,
suggesting that fever, dehydration, and acidosis
may be contributing factors [6]. Bone infarction
can occur anywhere in the skeleton. It results
directly from the sickling of RBCs in the bone
marrow, which causes stasis of blood and seques-
tration of cells [13, 15].
Stroke is one of the most severe complications
of SCD because it may result in permanent neu-
rological deficits or death. The incidence of stroke
varies depending on the SCD genotype. The prev-
alence of cerebral vascular accident (CVA) was
11 % in patients under 20 years of age with homo-
zygous HBSS [16]. Ischemic strokes are more
common in the children, adolescent, and old
adult SCD patients, while the incidence of hem-
orrhagic strokes peaks between the ages of 20–29
years [16]. Silent cerebral infarcts (SCIs) occur
more commonly than overt strokes, with a preva-
lence of 21.8 % in children with HbSS ages 6–19
years [17]. SCIs are defined as a lesion on MRI,
443
without corresponding neurological deficits last-
ing longer than 24 h. SCIs are clinically signifi-
cant because they indicate a higher likelihood of
subsequent overt stroke [18]. Although SCD is
recognized as a hypercoagulable state, the inci-
dence of venous infarction due to thrombosis is
unknown [19].
Cognitive Deficits
As more SCD patients survive to adulthood, the
neurocognitive dysfunction caused by SCD is
becoming an emerging problem in young adults
[20]. Studies correlating MRI evidence of brain
damage have come to controversial conclusions,
with some stating that neurocognitive deficits in
SCD patients are associated with focal brain
injury and silent strokes in children while others
state that these deficits are associated more with
anemia and age. A recent study suggests that
volume reduction of the basal ganglia and thala-
mus may correlate with cognitive deficits in adult
with SCD [21]. These reports suggest that the
cause of the neurocognitive impairment is multi-
factorial and that further correlating studies using
functional imaging for cerebral blood flow and
hypoxia may be needed [20].
Treatment
Hydroxyurea (HU)
Hydroxyurea (HU) is a drug that promotes the
production of HbF, the normal hemoglobin
which is more prevalent in newborns. HU is cur-
rently the only established preventive pharmaco-
logic treatment for both children and adults with
SCD. The presence of HbF inhibits HbS polymer-
ization and higher levels of HbF are associated
with reduced mortality in SCD patients [2]. In a
randomized controlled trial, HU was shown to
decrease the frequency of painful episodes and
acute chest syndrome and the need for blood
transfusion and hospital admissions in SCD
patients [22].
444
Blood Transfusion
Chronic blood transfusions have been demon-
strated to reduce the risk of both primary and
secondary strokes and SCIs [23] and prevent
repeated ACS [2, 4]. There is currently no specific
treatment for acute stroke in SCD. However,
blood transfusion within 48 hours after acute
stroke may decrease the incidence of subsequent
strokes [24]. The aim of transfusion therapy is to
reduce HbS to below 30 %, which effectively
prevents overt stroke and SCIs [25]. However, it
can result in morbidity related to iron overload
and iron deposition in multiple organs, including
the liver, heart, and endocrine systems, and often
necessitates iron chelation therapy [2, 4].
Hematopoietic Cell Transplantation
Hematopoietic cell transplantation is the only
curative treatment for SCD thus far. However,
there is a lack of suitable allogeneic marrow
donors (usually HLA-matched siblings) for
about 90 % of eligible patients [2]. There is a
5–10 % mortality rate associated with this treat-
ment, mostly due to graft failure, lack of response
to first-line immunosuppressive therapy, graft-
versus-host disease (GVHD), relapse after the
initial successful treatment, and secondary clonal
and malignant disease [26]. There are especially
poor outcomes in patients over the age of 16 years
who are subjected to myeloablative conditioning
regimens, making the use of non-myeloablative
regimens standard in older patients [2].
Neurovascular Complications Concept
and Imaging Features
Stroke
Stroke or CVA is defined as a sudden-onset neu-
rological deficit that persists for more than
24 hours. SCD is one of the most common causes
of stroke in children and young adults. The Coop-
erative Study of Sickle Cell Disease (CSSCD)
classified stroke in SCD patients as ischemic,
A. Fujita et al.
hemorrhagic, or transient ischemic attack
[16]. The special forms of ischemic stroke in
SCD patients are the so-called silent cerebral
infarcts (SCIs) and stroke caused by cerebral fat
embolism [27].
Imaging approaches for stroke in SCD patients
are the same as for non-SCD patients. The pur-
pose of acute stroke imaging in SCD is to deter-
mine the extent of brain damage, to differentiate
ischemic stroke from hemorrhagic stroke, and to
help identify and delineate underlying
vasculopathy.
CT of the brain without contrast is usually the
first imaging modality used in the acute setting,
which allows for quick assessment of hemor-
rhagic versus ischemic stroke, for detection of
other causes to explain new neurological deficits,
and for evaluation of the extent of damage and
complications. Other benefits of CT include its
easier accessibility and 24/7 availability at most
institutions, as well as short exam times [28]. Dis-
advantages of CT include ionizing radiation expo-
sure, low sensitivity for detection of small
infarcts, and low sensitivity for stroke detection
in the first several hours [29].
MRI with diffusion-weighted imaging (DWI)
is very useful in detecting acute infarcts as well as
determining the exact location and extent of ische-
mic lesions [30, 31]. Apparent diffusion coeffi-
cient (ADC) maps in conjunction with DWI help
to pinpoint regions of early ischemia and infarct
[32]. T2-weighted (T2-WI) imaging and fluid-
attenuated inversion recovery (FLAIR) imaging
are used for detecting T2 high signal intensity
lesions [33], such as acute stroke and silent
infarcts. Gradient echo T2*-weighted (T2*WI)
images and susceptibility-weighted images
(SWI) are used to detect hemorrhagic lesions
[34]. While MRI offers the benefits of increased
sensitivity for pathology compared to CT and
avoidance of radiation exposure, it requires con-
siderable patient cooperation and in many pediat-
ric patients, may require sedation. The other
disadvantages of MRI are limited access for
acutely ill patients at many institutions and rela-
tively long duration of imaging studies.
Vascular imaging studies in the head and neck
may be considered in all patients with SCD,
21 Sickle Cell Disease and Stroke
Fig. 1 A 38-year-old woman with SCD, presenting with
altered mental status. (a) Axial FLAIR image shows mul-
tiple high signal intensity lesions of the bilateral
periventricular white matter (arrows). (b, c)
Corresponding axial DWI (b = 1,000) (b) with ADC
irrespective of whether they had previous epi-
sodes of stroke [30, 35]. CTA and MRA can be
used to detect intracranial arterial occlusion or
stenosis and aneurysms. CTA exposes patients to
high levels of ionizing radiations and contrast
media; therefore, MRA is preferred in children
[30]. MR venography can be useful when venous
thrombosis is suspected. Conventional catheter
angiogram is an invasive procedure but yields
higher anatomical detail of vascular anatomy and
pathology than other imaging modalities, includ-
ing information on hemodynamic function. Cath-
eter angiogram is usually indicated prior to
surgery or endovascular interventions [31]. A
careful risk–benefit evaluation should be
performed prior to subjecting an SCD patient to
catheter angiogram. If anesthesia is necessary,
special care should be taken in the perioperative
care of patients with SCD [36].
Ischemic Stroke
The majority of stroke in SCD is caused by large
vessel disease, which may involve the proximal
cerebral arteries and the neck arteries [8, 27].
Macrovascular stroke in SCD has imaging
features of cortical involvement and is more com-
monly unilateral [8]. DWI with ADC map is
essential for diagnosing acute ischemic infarct
and for differentiating it from chronic ischemic
changes. Both acute and chronic changes can have
445
map (c) shows multiple foci of restricted diffusion in the
left periventricular white matter consistent with acute
infarcts (arrows). Right periventricular lesions suggest
chronic ischemia or prior infarcts
high intensity on T2-WI images. With FLAIR
imaging, abnormal T2 high signal may be more
easily detected than conventional T2-WI imaging,
particularly in the periventricular and subcortical
regions (Figs.1, 2, 3, 4, and 5).
Arterial stenosis or occlusion is often seen in
conjunction with infarcts of the cerebral gray mat-
ter (the cortex, basal ganglia, and thalamus)
(Fig. 2). However, white matter infarcts can be
seen in the setting of arterial stenosis or occlusion
as well, particularly in border (watershed) zones
[8] (Fig. 3). The anterior circulation is more often
involved than the posterior circulation. The radi-
ologist should screen for loss of vascular flow
voids on T2-WI images, which may indicate
vasculopathy (Fig. 2e). FLAIR imaging can
show high signal in cortical vessels in the setting
of altered flow velocities, such as slow distal flow
of a stenotic arterial branch or leptomeningeal
collaterals [37] (Fig. 6). In these cases, the radiol-
ogist may decide to add brain MRA or recom-
mend other angiographic imaging such as CTA or
catheter angiogram for further evaluations regard-
ing diagnosis, treatment, and management.
Hemorrhagic Stroke
Hemorrhagic stroke in SCD is relatively more
common in adults than in children. Although it
is rare, primary hemorrhagic stroke, such as sub-
arachnoid and parenchymal hemorrhage, is
446
Fig. 2 A 6-year-old girl presenting with right-sided weak-
ness. (a, b) Axial FLAIR images show multiple high-signal-
intensity lesions at the left frontal border zone (watershed)
areas (arrows). (c, d) Corresponding axial DWI (b = 1,000)
shows restricted diffusion in the left MCA territory and
Fig. 3 A 16-year-old woman presenting with headache
and syncope. (a) Axial FLAIR image demonstrates
encephalomalacia (asterisks) with surrounding gliosis
(arrows) in the right frontoparietal ACA-MCA border
A. Fujita et al.
border zone larger than that shown on FLAIR images con-
sistent with acute infarct (arrows). (e) Axial T2-WI image
shows loss of flow voids in the left proximal MCA (arrow).
(f) MRA shows obvious stenosis of left MCA (arrow) and
decrease of flow-related signal distally
zone (watershed) distribution indicating old ischemic
changes. (b) MRA shows decreased caliber and flow-
related signal with wall irregularity of the visualized right
extracranial ICA (arrows)
21 Sickle Cell Disease and Stroke
Fig. 4 A 22-year-old woman presenting with seizure.
Axial T2-WI image shows bilateral frontal high signal
intensity, indicating chronic ischemic changes and
encephalomalacia (asterisks), predominantly in
ACA-MCA border zone (watershed) distribution
frequently fatal in SCD patients. Aneurysms are
the most common identified cause of subarach-
noid hemorrhage in adult patients with SCD [27],
although their exact prevalence is unknown.
Compared with patients without SCD, aneurysms
in SCD patients are often multiple, have an
increased propensity for the posterior cerebral
circulation, and may be prone to rupture at smaller
sizes [38]. Moyamoya syndrome is a common
cause of parenchymal hemorrhage in SCD in chil-
dren [8], which will be discussed later in more
detail.
Non-contrast CT is the first imaging modality
of choice in the diagnosis of acute hemorrhagic
stroke (Figs. 7 and 8). In the current clinical prac-
tice, MRI may occasionally be obtained prior to
CT when there is high suspicion for acute stroke
based on clinical symptoms. Radiologists should
be familiar with the MR imaging features of acute
hemorrhage that are the same as non-SCD
patients, to avoid delayed diagnosis of hemor-
rhagic strokes. The accuracy of MRI in detecting
447
Fig. 5 A 14-year-old girl presenting with past episode of
right hemiplegia. Axial FLAIR image shows high signal
(arrows) and prominent sulci (asterisks), indicating vol-
ume loss and gliosis with cortical involvement in the left
cerebral hemisphere
acute parenchymal hemorrhage is similar to CT,
especially when gradient echo sequences, such as
T2*-WI imaging and SWI, are used [39]. Catheter
angiogram using three-dimensional (3D) digital
subtraction angiography is still the gold standard
for the diagnosis and preoperative delineation of
intracranial aneurysms. Noninvasive techniques
such as MRA and CTA are more commonly
used, and recently their accuracy is considered
satisfactory [40, 41]. CTA is often the first-line
modality to search for underlying aneurysms in
the setting of acute subarachnoid hemorrhage,
while MRA is commonly used in the non-acute
setting (Figs. 9 and 10).
Silent Cerebral Infarct (SCI)
SCI is the most common neurovascular complica-
tion in both pediatric and adult patients with SCD.
SCIs have been seen in children under the age of
6 years [42]. They may represent a different path-
ogenesis from either ischemic or hemorrhagic
448 A. Fujita et al.

Fig. 6 A 10-year-old girl

presenting with left facial
droop, drooling, and hand
tingling. (a) (b) Axial
FLAIR images show
multiple high signal
intensity vessels in the
Sylvian fissure and
frontoparietal sulci
(arrows), indicating slow
flow or thrombosis. (c)
MRA shows loss of flow-
related signal in the right
MCA distribution
(asterisks) and bilateral
proximal ACAs (arrows)

strokes as they involve small vessels in watershed
distributions instead of the larger cerebral vessels
[43]. Measurement of arterial velocities with
transcranial Doppler (TCD) ultrasound are not
predictive of SCI risk, unlike with overt ischemic
stroke, which supports a different pathophysiol-
ogy between overt stroke and SCI [44]. Several
studies showed that SCIs are more frequent in
patients with arterial stenosis and intracranial arte-
rial tortuosity, suggesting that large vessel disease
may play a role in the pathophysiology of these
lesions [42, 45, 46]. Children with arterial stenosis
are at higher risk of brain parenchymal injury as
they have more SCIs [47] (Fig. 10).
The presence of SCI is considered a predictor of
future strokes [17, 48, 49]. SCIs have been shown
to progress in both number and size over time [49].
Associated neurocognitive abnormalities also may
be progressive [50]. Small foci of acute SCI may be
found in asymptomatic children with SCD [42].
SCIs may continue to progress in spite of initiation
of chronic transfusion therapy after a stroke [51];
however, a recent study showed that in patients
without a stroke and not at high risk of a stroke,
chronic transfusions reduced the occurrence of new
SCIs by 56 % over a median observation period of
3 years [23].
SCIs are defined as T2 high signal intensity
lesions of at least 3 mm in the greatest linear
dimension in children and 5 mm in adults, visible
on at least two planes of T2-WI images [20, 52]
(Figs. 11 and 12). Because of their small size,
detection of SCI depends largely on the MR imag-
ing technique, such as the use of FLAIR, slice
thickness, multiplanar acquisitions, and the mag-
netic field strength [53]. With further advances in
21 Sickle Cell Disease and Stroke
Fig. 7 A 20-year-old man presenting with altered mental
status. Axial non-contrast CT shows a high density area in
the right frontal lobe white matter with surrounding low
density indicating edema (arrow), compatible with acute
hemorrhage
imaging and more utilization of 3.0-T or higher
magnet field units, it is likely that more cases of
SCI will be detected in individuals with SCD [49].
Cerebral Fat Embolism
Fat embolism syndrome is a very rare but poten-
tially lethal complication of SCD that is not
widely recognized [54]. In non-SCD patients, fat
embolism syndrome is usually a complication of
long bone fractures. The pathogenesis of fat
embolism syndrome in SCD remains controver-
sial. It is thought to be caused by bone marrow
infarcts and necrosis, with subsequent emboliza-
tion through osseous venous channels into the
venous circulation. Neurologic dysfunction in
the setting of fat embolism syndrome may result
directly from vessel occlusion by fat emboli, from
disruption of the blood–brain barrier by toxic free
fatty acids, or both [54, 55].
Typical MR imaging findings of cerebral fat
embolism include tiny foci of restricted diffusion
and corresponding T2 high signal intensities
representing microinfarcts. Lesions can involve
449
Fig. 8 A 14-year-old boy presenting with severe head-
ache. Axial non-contrast CT shows high density in the right
Sylvian fissure consistent with acute subarachnoid hemor-
rhage (arrows)
the cerebral white matter, basal ganglia, thalamus,
brain stem, and cerebellum. This MR imaging
pattern, sometimes called a “starfield” pattern, is
not very specific and can also be seen in cardio-
genic or septic emboli, vasculitis, and tiny hem-
orrhagic metastases [54]. T2*-WI images and
SWI frequently demonstrate microhemorrhages
in cerebral fat embolism [55].
Moyamoya Syndrome
Vascular pathology in SCD includes arterial tor-
tuosity, stenosis, occlusion, and aneurysm forma-
tion. A special form of cerebrovascular
manifestation in SCD patients is “moyamoya syn-
drome,” a progressive stenosis of major intracra-
nial arteries with formation of numerous collateral
circulations, typically via lenticulostriate and
thalamoperforating arteries (Fig. 13). Moyamoya
syndrome typically involves the distal ICA (inter-
nal carotid artery) and proximal ACA (anterior
450 A. Fujita et al.

Fig. 9 A 34-year-old man

presenting with headache.
(a) MRA shows an ovoid
focus of flow-related signal
arising from the junction of
the cavernous–clinoid
segments of the left ICA
extending inferiorly,
consistent with an
aneurysm (arrow). (b)
Catheter angiogram of the
left ICA lateral view
confirmed the aneurysm
(arrow)

Fig. 10 A 22-year-old man presenting with headache and cerebral infarcts. (b) MRA shows a 3 mm aneurysm at the
vertigo. (a) Axial FLAIR image shows multiple high signal medial portion of the left IC-ophthalmic segment (arrow).
foci (arrows) in the left frontal white matter without abnor- Tortuous extracranial right internal carotid artery and bilat-
mal diffusion restriction (not shown) consistent with silent eral vertebral arteries are also noted (open arrows)

cerebral artery) and MCA, with relative sparing of
the posterior circulation. The name moyamoya
syndrome is derived from moyamoya disease, a
rare, presumed genetic, progressive cerebrovascu-
lar disorder. The term “moyamoya” means “puff
of smoke” in Japanese and refers to the appear-
ance of collateral vessels on cerebral angiogram
[56]. The pathophysiology of moyamoya syn-
drome in SCD is poorly understood. It is hypoth-
esized that sickle-shaped RBCs occlude the vasa
vasorum of the carotid arteries [31], leading to
intimal hyperplasia [57]. The prevalence of
moyamoya syndrome in SCD patients has been
reported for 35 % of conventional catheter angio-
gram studies [7] and in 20–40 % of MRA studies
[8, 58]. It was shown that 75 % of patients with
moyamoya syndrome demonstrated radiographic
progression of arteriopathy and 65 % of patients
showed clinical progression during follow-up
[59] (Fig. 14).
Moyamoya syndrome can be detected on
MRA as progressive large vessel occlusions and
major collateral flow patterns, the so-called
moyamoya vessels. On T2-WI images,
21 Sickle Cell Disease and Stroke 451

moyamoya syndrome presents with loss of flow
voids in the circle of Willis (Fig. 15) and on time-
of-flight MRA as a loss of flow-related signal
(Fig. 16). Collaterals typically present as multiple
flow voids on T2-WI and T1-WI images in the
thalamoperforate and lenticulostriate territories.
High signal intensity vessels within the cortical
sulci on FLAIR images can indicate a decreased
flow of cortical arteries [37] (Fig. 6).
Incidentally discovered asymptomatic
moyamoya syndrome in children has the potential
to progress, both radiographically and clinically.
Monitoring and screening with imaging in high-
risk patients facilitates early referral to
neurosurgery and allows for timely revasculariza-
tion therapy [59]. Patients with moyamoya syn-
drome may have a particularly poor prognosis and
may benefit from revascularization of ischemic
brain parenchyma by establishing collaterals
from the external carotid artery (ECA) branches
to the ICA territories [60]. Superficial temporal
artery (STA) to middle cerebral artery (MCA)
bypass is a direct revascularization procedure,
which is technically difficult. Indirect revascular-
ization procedures such as encephaloduroarterio-
synangiosis (EDAS) and pial synangiosis are
becoming popular for the surgical treatment of
moyamoya syndrome [61]. Pial synangiosis is a
technique in which the dura and the arachnoid are
opened and the STA adventitia is sutured directly
to the pia [62], and it has a relatively low risk of
complications [63]. Pial synangiosis typically
results in an increase in collaterals from the STA
or middle meningeal artery (MMA) to the brain
[62]. MRA or CTA can be used to monitor the
patency of synangiosis between STA branches
and pial arterial branches after cerebral revascu-
larization (Fig. 14).

Extracranial Vasculopathy

Extracranial ICA stenosis is rare, but it has been

Fig. 11 A 6-year-old girl presenting with SCD for screen-
ing MRI. Coronal FLAIR image shows high signal foci in
the right frontal subcortical white matter (arrow), compat-
ible with silent cerebral infarct
reported that there is an association between the
presence of extracranial internal carotid
arteriopathy and stroke risk in SCD
[64–66]. Doppler ultrasound screening studies

Fig. 12 A 42-year-old
woman presenting with left
facial numbness. (a) Axial
FLAIR image shows
multiple high signal
intensity lesions in the right
frontoparietal white matter
(arrows). (b) DWI
(b = 1,000) shows
corresponding restricted
diffusion in the postcentral
gyrus consistent with acute
infarct (arrow)
452
Fig. 13 A 24-year-old man presenting with multiple epi-
sodes of ischemic infarct. (a) Axial FLAIR image shows
bilateral periventricular high signal intensity with cystic
foci compatible with old ischemic change (arrows). (b)
MRA shows occlusion of the right ICA terminus (asterisk)
and severe stenosis of the left ACA (arrow) and distal
MCA (open arrow). (c) Catheter angiogram of the right
found stenosis or occlusion of the cervical ICA in
3–6 % of the children, and the majority of these
also had strokes [65]. One MRA study reported
that patients with an acute neurologic event, TCD
abnormalities, and previous history of stroke also
showed cervical ICA stenosis and/or occlusion.
The most commonly involved artery is the prox-
imal ICA, within 1 cm from its origin from the
CCA [66] (Fig. 17). In addition, a recent study
showed that the high velocity of extracranial ICA
A. Fujita et al.
ICA shows occlusion of the supraclinoid ICA (arrow).
Posterior communicating artery (PComm) is markedly
prominent (asterisk) and diffuse moyamoya-like collateral
vessels are noted (open arrows). (d) Subsequent image of
the ICA catheter angiogram shows more apparent
“moyamoya” (puff-of-smoke) appearance of the collateral
vessels (open arrows)
detected with submandibular Doppler ultrasound
is highly predictive of intracranial arterial stenosis
seen on MRA [67]. The differential diagnosis of
extracranial ICA stenosis and/or occlusion
includes atherosclerosis, arteritis, and
fibromuscular dysplasia. Age, location, and lack
of systemic inflammation or hypercoagulability
help differentiate SCD-related extracranial inter-
nal carotid arteriopathy from other
conditions [68].
21 Sickle Cell Disease and Stroke 453

Fig. 14 A 15-year-old man

presenting with recurrent
headache. (a, b) In 2005,
Axial FLAIR image (a)
shows normal brain and
MRA (b) shows possible
stenosis of the right
proximal ACA and bilateral
proximal MCA (asterisk).
(c, d) In 2006, axial FLAIR
image (c) still shows normal
brain, but MRA (d) shows
obvious occlusion of the
right terminal ICA
(asterisk). The patient
received bilateral pial
synangiosis for cerebral
revascularization in 2007.
(e, f) In 2013, axial FLAIR
image (e) shows bilateral
frontal old infarcts with
encephalomalacia and
gliosis (arrows). MRA (f)
shows bilateral occlusion of
the terminal ICAs
(asterisks) and bilateral
prominent superficial
temporal arteries
(arrowheads) after
revascularization. Serial
FLAIR images and MRA
images over 8 years
demonstrate progression of
arterial stenosis, bilateral
frontal lobe infarcts, and
atrophy in spite of
revascularization

Arterial tortuosity is considered to be an tortuosity may serve as a predictor of chronic

adaptive response to chronic anemia in patients
with SCD. Chronic anemia is associated with
increased cardiac output and high blood flow
velocity due to the increased need for blood sup-
ply, which can lead to arterial tortuosity and can
occur in all cervical arteries [68]. Cerebrovascular
brain hypoxia [53]. On imaging, tortuosity can
be diagnosed when the following features are
present: dilatation or ectasia of an artery segment,
abnormal increase in length of an arterial segment,
and obvious bowing of an artery [53, 69]
(Figs. 17, 18, and 19).
454
Fig. 15 A 22-year-old woman presenting with seizure.
Axial T2-WI image shows loss of normal flow voids of
the bilateral MCA and ACA (arrows). In contrast, promi-
nent posterior circulation flow voids are noted
(arrowheads)
Fig. 16 A 38-year-old woman with SCD presenting with
altered mental status. MRA shows multiple stenoses and
occlusions of the right ICA terminus, bilateral ACAs,
bilateral PCAs (posterior cerebral artery), and left MCA
(arrows)
A. Fujita et al.
Posterior Reversible Encephalopathy
Syndrome (PRES)
There have been several case reports of posterior
reversible encephalopathy syndrome (PRES), also
known as reversible posterior leukoence-
phalopathy syndrome (RPLS), in patients with
SCD [70–74]. The vast majority of the patients
are in the pediatric population [71]. It is believed
that there are two major pathomechanisms in
PRES: a cytotoxic event after an increase in blood
pressure that causes vasoconstriction and cerebral
ischemia or a vasogenic etiology where hyperten-
sion in combination with endothelial dysfunction
and failure of cerebral autoregulation causes vaso-
dilation [70]. Studies have identified PRES as a
common neurological complication of SCD, and
this entity needs to be considered when evaluating
a child with acute neurological deterioration, espe-
cially if the patient is hypertensive, has history of
ACS, and has received systemic steroid therapy or
blood transfusion [72–74]. A diagnosis of PRES
may predict a high risk for morbidity and mortality
[73]. On MR imaging, PRES usually presents as
symmetrical areas of high signal intensities on
T2-WI and FLAIR images without associated
restricted diffusion, suggesting vasogenic edema,
which is the same as non-SCD patients [73]
(Fig. 20).
Other Sickle Cell Disease Imaging
Manifestations
There are several osseous complications related to
SCD. These include chronic bone marrow
changes in response to long-standing anemia and
acute bone infarcts as well as osteomyelitis. Oste-
oporosis and iron deposition in the bone marrow
due to repeated transfusions can also be seen [13,
15, 68, 75, 76].
Craniofacial Bone
Bone Marrow Hyperplasia
As a result of chronic anemia in SCD patients, the
red marrow in bony erythrocyte production never
21 Sickle Cell Disease and Stroke
Fig. 17 A 30-year-old woman with SCD, evaluation for
revascularization. (a) Catheter angiogram of the right CCA
(lateral view) shows complete occlusion of the extracranial
ICA near its origin (arrow). (b) Catheter angiogram of the
Fig. 18 A 34-year-old woman with SCD, presenting for
annual MRI follow-up. MRA shows no significant stenosis
or occlusion of the intracranial arteries; however, dilatation
and elongation of the extracranial left ICA are seen
(arrows)
undergoes the conversion to fatty yellow marrow
[68, 75]. Furthermore, as the anemia becomes
more severe and long-standing, the bone marrow
spaces expand secondary to increased RBC pro-
duction [68, 75]. This marrow expansion is most
frequently noted clinically in the skull and facial
455
left CCA (lateral view) shows stenosis of the ICA at the
ophthalmic segment (arrow). (c) Catheter angiogram of the
left subclavian artery shows tortuosity of the left VA
(arrows)
bones. The non-conversion of the bone marrow is
demonstrated by diffuse low signal intensity on
T1-WI images relative to the signal in the
intervertebral disks [15] (Figs. 21 and 22).
Bone Infarct
Bone infarcts are more common than osteomyeli-
tis among children with SCD. Ischemia causes
pain, even before infarct occurs. For SCD patients
presenting with headache or localizing facial pain,
acute craniofacial bone infarct should be a major
diagnostic consideration [13, 15]. The orbital
walls, mandible, cranial vault, and skull base are
commonly affected by bone infarcts in SCD
patients [13, 15] (Fig. 23). In the spine, it is
hypothesized that bone infarcts contribute to the
characteristic “letter H” deformity of vertebral
bodies that is typical of patients with SCD [77]
(Fig. 22).
MR imaging is more sensitive than CT in
detecting bone infarcts. MR imaging findings of
acute bone infarcts include increased signal on
T2-WI images, particularly with fat-suppression
techniques such as STIR, and associated high
456
Fig. 19 A 20-year-old man with SCD. MRA shows tor-
tuous left extracranial ICA (arrows) and aneurysmal dila-
tation of the extracranial segment of the left VA (open
arrow)
signal intensity on DWI with corresponding low
signal intensity on ADC map (Fig. 24). There may
also be subperiosteal hemorrhage. The high signal
intensity of the affected bones on fat-suppressed
T1-WI images, which may represent the seques-
tered RBCs in the infarcted marrow [13, 78], can
be very helpful in the differentiation of bone
infarcts from osteomyelitis (Fig. 25).
Osteomyelitis
Osteomyelitis is less common than bone infarcts
and typically affects long bones. However, osteo-
myelitis may also occur in craniofacial bones,
particularly the mandible, because of its relatively
poor blood supply [13, 68] (Fig. 26). Differentia-
tion between infarct and osteomyelitis in early
presentation with imaging alone can be quite chal-
lenging. The typical MR imaging findings,
including bone marrow edema, heterogeneous
bone marrow enhancement, and soft tissue swell-
ing, can be seen in osteomyelitis as well as bone
A. Fujita et al.
Fig. 20 A 5-year-old girl with SCD, presenting with
headache and hypertensive status. Axial FLAIR image
shows high signal intensity lesions in the bilateral occipital
lobes (arrows). Given the distribution, clinical course, and
documented hypertension, the findings are compatible
with PRES
infarcts. In the presence of fluid collections, MR
susceptibility artifact is suggestive of blood prod-
ucts from subperiosteal hematoma, which is more
commonly seen with bone infarcts. Infectious
fluid collections may exhibit restricted
diffusion [13].
Temporal Bone
Sensorineural hearing loss (SNHL) is a well-
recognized inner ear complication of SCD [68,
79]. The most accepted pathogenesis of SNHL is
the recurrent vaso-occlusion of the labyrinthine
blood vessels, either in the distribution of the
anteroinferior cerebellar artery or a branch of the
basilar artery, which can result in labyrinthine
hemorrhage (LH) and labyrinthitis ossificans
(LO) [79]. In the past, LH was considered rela-
tively rare in patients of SCD with sudden-onset
SNHL and vertigo, but it was recently shown that
abnormalities can be detected by MR imaging in
21 Sickle Cell Disease and Stroke
Fig. 21 An 18-year-old man with SCD. Sagittal T1-WI
image shows the diffuse expansion of the marrow space of
the skull (arrows)
Fig. 22 A 10-year-old boy with SCD, presenting with
back pain. Sagittal T1-WI image of the lumbar spine
shows low signal in the vertebral bodies (asterisks) relative
to the intervertebral disk, compatible with an activated
bone marrow triggered by chronic anemia. Note also the
“H”-shaped deformity of vertebral bodies, compatible with
end-plate infarcts and subsequent height loss
approximately one-third of SCD patients with
inner ear symptoms, with males preferentially
affected [79].
High signal intensity on non-contrast high-reso-
lution T1-WI images is a characteristic finding for
LH. Contrast-enhanced T1-WI images can exclude
457
Fig. 23 A 6-year-old boy presenting with temporoman-
dibular joint pain and headache. Axial fat-suppressed
T2-WI image demonstrates high signal intensity in the
left mandibular ramus (arrow), compatible with bone
infarct
tumors such as schwannomas and hemangiomas
that would show intense and localized enhance-
ment, and fat-suppressed MR imaging or high-
resolution temporal bone CT may be used to
exclude lipoma and fat-containing tumors (Fig. 27).
The characteristic imaging finding of LO is the
high density of the membranous labyrinth on
CT. On high-resolution T2-WI images, the normal
T2 high signal from the fluid in the membranous
labyrinth is lost. LO is the end stage of
labyrinthitis, characterized pathologically by pro-
liferation of fibroblasts and finally osteoblasts. In
patients with SCD presenting with inner ear com-
plaints, dedicated temporal bone imaging should
be performed, preferably by MR imaging
(Fig. 28).
Advanced Imaging
Diffusion Tensor Imaging (DTI)
Diffusion tensor imaging (DTI), which measures
the directional tendencies of Brownian water mol-
ecule motion in biologic tissues, has a potential to
458
Fig. 24 A 15-year-old boy with SCD, presenting with
headache. (a) Coronal T2-WI image shows high signal
intensity in the skull consistent with bone marrow edema
(arrows), as well as mixed abnormal signal intensity in the
Fig. 25 A 17-year-old man presenting with headache.
Sagittal T1-WI image shows high signal intensity of the
bone marrow in the parietal bone consistent with bone
infarct with hemorrhage (arrow) (Adapted with permission
from Lippincott Williams and Wilkins/Wolters Kluwer
Health: Journal of Computer Assisted Tomography [13],
copyright (2013))
detect and quantify microstructural brain changes
earlier than conventional MRI [80]. This
advanced MRI technique can show the orientation
and integrity of white matter fibers in vivo. The
apparent diffusion coefficient (ADC), which is
comparable to mean diffusivity as a measure of
the degree of restriction to water diffusion, and
A. Fujita et al.
subgaleal fluid collection consistent with hemorrhage
(open arrow). (b) Axial DWI (b = 1,000) shows an area
of increased signal in the skull, compatible with bone
infarct (arrows)
Fig. 26 A 25-year-old man presenting with left jaw swell-
ing and pain. Coronal contrast-enhanced CT shows periph-
erally enhancing low density fluid collections medial and
lateral to the left mandibular ramus (arrows), consistent
with subperiosteal abscess formation
fractional anisotropy (FA), a measure of the pre-
ponderant directionality of water diffusion, are
two of the most frequently used DTI metrics for
measuring microstructural tissue damage in
patients with brain disease [80]. DTI directional-
ity information is used for fiber tractography,
which is the only way to obtain 3D fiber architec-
ture of white matter tracts in vivo [80].
Using region of interest (ROI) based analysis
in a study comparing SCD patients with controls,
21 Sickle Cell Disease and Stroke
FA values in the prefrontal segment of the corpus
callosum (which covers the first sixth of the cor-
pus callosum), upper frontal white matter area,
centrum semiovale, anterior periventricular
white matter area, posterior periventricular white
matter area, and brain stem were significantly
lower in SCD patients than controls. ADC values
were significantly higher in SCD patients com-
pared to controls in the prefrontal segment of the
Fig. 27 A 14-year-old man presenting with sudden onset
of right-sided hearing loss. Axial T1-WI image shows high
signal intensity in the right cochlea (arrow) compared to
left, suggesting labyrinthine hemorrhage
Fig. 28 A 12-year-old boy presenting with hearing loss.
(a) Axial high-resolution T2-WI image (DRIVE sequence)
shows signal loss of the right cochlea (asterisk). (b) Axial
459
corpus callosum and sensory segment of the cor-
pus callosum (which covers the posterior
one-third but not including the posterior
one-fourth), caudate nucleus, thalamus, and pons
[81]. The increase in ADC along the fibers may
indicate extracellular water content increase sec-
ondary to axonal loss and fiber density reduction,
both of which could be attributed to chronic ische-
mia. One study found a significant reduction in
DTI fiber counts in the corpus callosum of SCD
patients compared to controls. A significant
reduction in DTI fiber counts was also observed
in the corticospinal tracts bilaterally. These results
can be explained by axonal damage due to
vasculopathy [81].
Perfusion Imaging
Perfusion imaging can be achieved with contrast-
enhanced CT, contrast-enhanced MRI,
non-contrast MRI with arterial spin labeling
(ASL) technique, and positron emission tomogra-
phy (PET) imaging [82]. Perfusion imaging does
not affect treatment decisions in SCD patients.
However, there is value in studying brain perfu-
sion in SCD to better understand the pathophysi-
ology of the disease. Perfusion imaging allows the
calculation of relative cerebral blood flow (CBF)
and blood volume in the brain. Because of the
very high radiation dose compared to
CT shows ossification of the cochlea corresponding to
signal loss on MRI (a), consistent with labyrinthitis
ossificans (arrow)
460
Fig. 29 A 10-year-old boy with SCD presenting with
acute visual symptoms. (a) Axial DWI (b = 1,000)
shows mildly increased signal in the bilateral parietal
lobes. (b) Arterial spin labeling perfusion image shows
decreased blood flow in the bilateral parieto-occipital
conventional head CT, CT perfusion and PET
imaging should be cautiously used in children
[82]. Perfusion MRI techniques commonly used
include dynamic susceptibility contrast (DSC)
MRI and ASL. Compared with ASL, DSC MRI
perfusion provides higher spatial resolution,
requires shorter scanning time, and can measure
other hemodynamic parameters such as cerebral
blood volume (CBV) and mean transit time
(MTT) simultaneously [83]. ASL using either con-
tinuous or pulsed ASL sequence is a noninvasive
MR perfusion imaging technique that uses arterial
blood water protons as an endogenous tracer of
perfusion [84–86] (Fig. 29). Perfusion imaging
studies can show changes of cerebral blood flow
immediately after the event [87]. ASL studies of
whole brain CBF in SCD patient compared to
controls have been controversial [84–86]. ASL
has been used to correlate CBF with both full-
scale and performance IQ [85]. Using ASL, studies
have identified right-left CBF asymmetries that
could be early indicators of subclinical pathologi-
cal changes in microvasculature and/or hemody-
namics in SCD patients. The described perfusion
asymmetries, however, could not always be asso-
ciated with the presence of ischemic lesions
[84–86].
A. Fujita et al.
lobes (arrows). (c) Axial DWI (b = 1,000) 2 days after
the initial MRI shows discrete areas of diffusion restriction
in the bilateral parietal lobes (arrows) and in the left frontal
periventricular watershed distribution (open arrows)
MR Spectroscopy (MRS)
MR spectroscopy (MRS) is an advanced MR
imaging method for quantitative and qualitative
assessment of the metabolic state of brain tissue.
Reports of MRS in SCD are limited. In 1992, one
study showed that normally appearing brain tis-
sues on conventional MRI did not show abnormal
MRS findings [88], meaning there was no addi-
tional increase in sensitivity in disease detection
with MRS. The same group reported a reversible
temporary decrease of N-acetylaspartate (NAA),
which is considered a marker of neuronal tissue, at
the time of a stroke episode [89]. Another study
showed that NAA was increased in the basal
ganglia and generally in the brain of SCD patients
[90]. These results suggest that brain NAA
appears not to be a reliable marker of viable neu-
rons in SCD patients at this moment.
Quantitative MRI (qMRI)
Remarkable advances in MR imaging have led to
the development of a variety of quantitative MR
imaging (qMRI) techniques, including diffusion-
weighted imaging (DWI), diffusion tensor
21 Sickle Cell Disease and Stroke
imaging (DTI), relaxometry, magnetization trans-
fer imaging, perfusion imaging, MR spectros-
copy, and volumetry [91].
qMRI studies of T1 spin–lattice relaxation
times have shown reduced gray matter relaxivity
in pediatric patients with SCD compared to
healthy controls [92, 93], especially in the caudate
nucleus and in cortical gray matter. In addition,
volumetric deficits of gray matter in patients with
SCD were reported in the absence of volumetric
deficits in the white matter [94]. These observa-
tions support the idea that gray matter is selec-
tively vulnerable to injury in pediatric SCD
patients. In the future, the use of qMRI tech-
niques, such as T1 values and volumetry, may
serve as early markers of brain damage in young
SCD patients [92–94].
qMRI studies examined with mixed TSE pulse
sequence showed T1 lengthening, T2 and secular-
T2 shortening, and an increase in bone marrow
volume in patients with SCD [75, 76]. The T1
lengthening supports the fact that there is less
yellow marrow volume and more red marrow
volume in patients with SCD, which suggests the
failure of red-to-yellow marrow conversion. T2
and secular-T2 shortening also supports the fail-
ure of red-to-yellow marrow conversion.
Increased iron deposition as a result of hemolysis
in subjects with severe disease and repeated blood
transfusions can account for the shortening of T2
and secular-T2 times. By using the specific
changes seen in T1, T2, and secular-T2 relaxation
times, qMRI can be a useful tool in the assessment
and monitoring of disease severity in patients with
SCD [75, 76]. In addition, qMRI has been used to
evaluate the extracranial structures, such as sali-
vary glands [95] and lacrimal glands [96].
Nuclear Medicine
Brain perfusion single-photon emission computed
tomography (SPECT) is a functional neuroimag-
ing technique that allows for noninvasive evalua-
tion of the physiological and pathophysiological
status of the brain. SPECT imaging may demon-
strate cerebral perfusion deficits in SCD patients
461
with normal brain MR images. One study found
that neurologically intact adult SCD patients
exhibit brain perfusion decreases in areas such as
the basal ganglia, thalamus, watershed areas of the
anterior circulation, and cerebellar and occipital
cortex [97]. These findings may contribute to the
understanding of the mechanisms underlying the
evolution of cerebral vasculopathy in SCD
patients. Calculation of cerebrovascular reserve
with acetazolamide (Diamox)-induced cerebro-
vascular vasodilatation in conjunction with
SPECT imaging has a role in identifying SCD
patients who are at a high risk for stroke [98].
Since the late 1990s, perfusion and metabolic
imaging techniques, such as positron emission
tomography (PET), have been increasingly used
to investigate the microvascular blood flow in
patients with SCD. One study showed that PET
may detect abnormalities missed by structural
MRI in patients with SCD and a history of stroke
[99]. Another study found that metabolic compro-
mise in SCD patients exceeded the corresponding
anatomic abnormality demonstrated on
MRI [100].
Management
Prevention of Primary Stroke
Early screening of SCD patients with TCD ultra-
sound is the gold standard for assessing stroke risk
[25]. In the absence of early TCD screening,
stroke occurs in 7.4 % and 11 % of patients with
SCD by 14 and 20 years of age, respectively, [16,
101] and by age 30 and 45 years, 15 % and 24 %
will experience overt stroke, respectively [16,
102]. In a newborn cohort, an early TCD screen-
ing and transfusion program employed in patients
at risk decreased the overt stroke risk to 1.9 % by
18 years of age [103]. It has been recommended to
obtain serial TCD and brain MRI/MRA studies to
monitor the effectiveness of transfusion in chil-
dren with abnormal TCD velocities [104]
(Fig. 30).
The time-averaged mean of the maximal veloc-
ity (TAMMV) in large intracranial arteries is a
462
Fig. 30 An 8-year-old boy
with SCD presenting with
annual follow-up with
transcranial Doppler
screening. Peak systolic
velocity (PSV) in the left
MCA was 258.7 cm/s
(<200, normal; 200–249,
conditional; 250
abnormal). Follow-up
MRA showed left MCA
stenosis (not shown)
powerful, noninvasive tool that can be utilized to
classify an individual patient’s risk of stroke
[25]. According to the TCD criteria for SCD,
TAMMV < 170 cm/s in any intracranial artery is
considered normal, and TAMMV of 170–199
cm/s is conditional. TAMMV > 200 cm/s is con-
sidered abnormal and indicates the need for trans-
fusion. TAMMV > 200 cm/s is associated with a
40 % risk of stroke within the next 3 years
[105]. The Stroke Prevention Trial in Sickle Cell
Anemia (STOP) documented that chronic transfu-
sion therapy in patients with a mean blood flow
velocity of >200 cm/s with TCD prevented the
initial occurrence of stroke, decreasing the annual
incidence of stroke from 10 % to <1 % [25]. TCD
is repeated every 12 months after a normal scan
and every 3 months with a conditional scan.
Although TAMMV is the gold standard TCD
measurement in the setting of SCD, in clinical
practice, peak systolic velocity (PSV) measure-
ments are more commonly used in vascular ultra-
sound practice and have been shown to be highly
correlated with TAMMV in the prediction of
future stroke [106] (Fig. 30). However, a recent
abstract has demonstrated that PSV alone may
overestimate the risk of stroke in children, dem-
onstrating the importance of TAMMV as the
A. Fujita et al.
primary initial measurement for stroke-risk
stratification [107].
Prevention of Secondary Stroke
SCIs that may be found in approximately 20–37 %
of children with SCD are associated with increased
risk of subsequent overt stroke [42, 44, 53]. Trans-
fusion therapy has been shown to reduce overt
strokes in patients with SCIs. Despite transfusion
treatment, there remains a 20 % secondary stroke
rate [108] and a 25 % likelihood of additional
SCIs [51]. Hydroxyurea (HU) has also been
shown to reduce TCD velocities and has shown
some benefit in reducing the risk of secondary
stroke [109]. A randomized clinical trial, Stroke
With Transfusions Changing to Hydroxyurea
(SWiTCH), evaluated HU with phlebotomy as an
alternative to the ongoing transfusion therapy,
but the study was closed early after an interim
analysis revealed equivalent liver iron between
the two groups, failing to meet the primary end
point [110]. MRI can monitor the progression and
extent of the ischemic change, and MRA can mon-
itor the stenosis and/or occlusion of the intra- and
extracranial arteries (Fig. 14).
21 Sickle Cell Disease and Stroke
Summary
Sickle cell disease (SCD) is one of the most
important hereditary hematologic disorders
which may result in severe neurovascular compli-
cations due to vaso-occlusion. Radiological eval-
uation is extremely important in daily practice,
such as TCD in stroke-risk screening with veloc-
ity measurements of the cerebral arteries, CT in
detecting hemorrhagic strokes, MRI in diagnosing
acute infarct, white matter changes and cerebral
vasculopathy, and catheter angiogram in evaluat-
ing precise vascular features and in treatment
planning. Radiologic imaging features may also
support the current understanding of the patho-
physiology underlying neurovascular complica-
tions of SCD. Advanced MR imaging and
nuclear imaging have the potential to provide
additional findings that may improve the manage-
ment of the neurovascular complications of SCD,
but further investigations are still necessary.
Finally, radiologists and clinicians should be
familiar with the pathophysiology and imaging
findings of the neurovascular complications of
SCD as well as various complications in the cra-
niofacial region and spine, which may be crucial
for timely diagnosis and appropriate treatment.
References
1. Rees DC, Williams TN, Gladwin MT (2010) Sickle-
cell disease. Lancet 376:2018–2031
2. Steinberg MH (2008) Sickle cell anemia, the first
molecular disease: overview of molecular etiology,
pathophysiology, and therapeutic approaches. Sci
World J 8:1295–1324
3. Steen RG, Hankins GM, Xiong X, Wang WC, Beil K,
Langston JW, Helton KJ (2003) Prospective brain
imaging evaluation of children with sickle cell trait:
initial observations. Radiology 228:208–215
4. Kanter J, Kruse-Jarres R (2013) Management of
sickle cell disease from childhood through adulthood.
Blood Rev 27:279–287
5. Connes P, Verlhac S, Bernaudin F (2013) Advances in
understanding the pathogenesis of cerebrovascular
vasculopathy in sickle cell anaemia. Br J Haematol
161:484–498
6. Lonergan GJ, Cline DB, Abbondanzo SL (2001)
Sickle cell anemia. Radiographics 21:971–994
463
7. Moran CJ, Siegel MJ, DeBaun MR (1998) Sickle cell
disease: imaging of cerebrovascular complications.
Radiology 206:311–321
8. Moritani T, Numaguchi Y, Lemer NB, Rozans MK,
Robinson AE, Hiwatashi A, Westesson PL (2004)
Sickle cell cerebrovascular disease: usual and unusual
findings on MR imaging and MR angiography. Clin
Imaging 28:173–186
9. Glushakova S, Balaban A, McQueen PG, Coutinho R,
Miller JL, Nossal R, Fairhurst RM, Zimmerberg J
(2014) Hemoglobinopathic erythrocytes affect the
intraerythrocytic multiplication of Plasmodium
falciparum in vitro. J Infect Dis 210:1100–1109
10. Halasa NB, Shankar SM, Talbot TR, Arbogast PG,
Mitchel EF, Wang WC, Schaffner W, Craig AS, Grif-
fin MR (2007) Incidence of invasive pneumococcal
disease among individuals with sickle cell disease
before and after the introduction of the pneumococcal
conjugate vaccine. Clin Infect Dis 44:1428–1433
11. Quinn CT, Rogers ZR, McCavit TL, Buchanan GR
(2010) Improved survival of children and adolescents
with sickle cell disease. Blood 115:3447–3452
12. Platt OS, Brambilla DJ, Rosse WF, Milner PF,
Castro O, Steinberg MH, Klug PP (1994) Mortality
in sickle cell disease. Life expectancy and risk factors
for early death. N Engl J Med 330:1639–1644
13. Watanabe M, Saito N, Nadgir RN, Liao JH, Flower
EN, Steinberg MH, Sakai O (2013) Craniofacial bone
infarcts in sickle cell disease: clinical and radiological
manifestations. J Comput Assist Tomogr 37:91–97
14. Gladwin MT, Vichinsky E (2008) Pulmonary compli-
cations of sickle cell disease. N Engl J Med
359:2254–2265
15. Ejindu VC, Hine AL, Mashayekhi M, Shorvon PJ,
Misra RR (2007) Musculoskeletal manifestations of
sickle cell disease. Radiographics 27:1005–1021
16. Ohene-Frempong K, Weiner SJ, Sleeper LA, Miller
ST, Embury S, Moohr JW, Wethers DL, Pegelow CH,
Gill FM (1998) Cerebrovascular accidents in sickle
cell disease: rates and risk factors. Blood 91:288–294
17. Pegelow CH, Macklin EA, Moser FG, Wang WC,
Bello JA, Miller ST, Vichinsky EP, DeBaun MR,
Guarini L, Zimmerman RA, Younkin DP, Gallagher
DM, Kinney TR (2002) Longitudinal changes in
brain magnetic resonance imaging findings in chil-
dren with sickle cell disease. Blood 99:3014–3018
18. Miller ST, Macklin EA, Pegelow CH, Kinney TR,
Sleeper LA, Bello JA, DeWitt LD, Gallagher DM,
Guarini L, Moser FG, Ohene-Frempong K,
Sanchez N, Vichinsky EP, Wang WC, Wethers DL,
Younkin DP, Zimmerman RA, DeBaun MR, Cooper-
ative Study of Sickle Cell D (2001) Silent infarction
as a risk factor for overt stroke in children with sickle
cell anemia: a report from the Cooperative Study of
Sickle Cell Disease. J Pediatr 139:385–390
19. Sidani CA, Ballourah W, El Dassouki M,
Muwakkit S, Dabbous I, Dahoui H, Al-Kutoubi A,
Abboud MR (2008) Venous sinus thrombosis leading
464
to stroke in a patient with sickle cell disease on
hydroxyurea and high hemoglobin levels: treatment
with thrombolysis. Am J Hematol 83:818–820
20. Vichinsky EP, Neumayr LD, Gold JI, Weiner MW,
Rule RR, Truran D, Kasten J, Eggleston B, Kesler K,
McMahon L, Orringer EP, Harrington T, Kalinyak K,
De Castro LM, Kutlar A, Rutherford CJ, Johnson C,
Bessman JD, Jordan LB, Armstrong FD, Neuropsy-
chological Dysfunction, Neuroimaging Adult Sickle
Cell Anemia Study Group (2010) Neuropsychologi-
cal dysfunction and neuroimaging abnormalities in
neurologically intact adults with sickle cell anemia.
JAMA 303:1823–1831
21. Mackin RS, Insel P, Truran D, Vichinsky EP,
Neumayr LD, Armstrong FD, Gold JI, Kesler K,
Brewer J, Weiner MW, Neuropsychological Dysfunc-
tion, Neuroimaging Adult Sickle Cell Anemia Study
Group (2014) Neuroimaging abnormalities in adults
with sickle cell anemia: associations with cognition.
Neurology 82:835–841
22. Charache S, Terrin ML, Moore RD, Dover GJ, Barton
FB, Eckert SV, McMahon RP, Bonds DR (1995)
Effect of hydroxyurea on the frequency of painful
crises in sickle cell anemia. Investigators of the Mul-
ticenter Study of Hydroxyurea in Sickle Cell Anemia.
N Engl J Med 332:1317–1322
23. DeBaun MR, Gordon M, McKinstry RC, Noetzel MJ,
White DA, Sarnaik SA, Meier ER, Howard TH,
Majumdar S, Inusa BP, Telfer PT, Kirby-Allen M,
McCavit TL, Kamdem A, Airewele G, Woods GM,
Berman B, Panepinto JA, Fuh BR, Kwiatkowski JL,
King AA, Fixler JM, Rhodes MM, Thompson AA,
Heiny ME, Redding-Lallinger RC, Kirkham FJ,
Dixon N, Gonzalez CE, Kalinyak KA, Quinn CT,
Strouse JJ, Miller JP, Lehmann H, Kraut MA, Ball
WS Jr, Hirtz D, Casella JF (2014) Controlled trial of
transfusions for silent cerebral infarcts in sickle cell
anemia. N Engl J Med 371:699–710
24. Hulbert ML, Scothorn DJ, Panepinto JA, Scott JP,
Buchanan GR, Sarnaik S, Fallon R, Chu JY,
Wang W, Casella JF, Resar L, Berman B,
Adamkiewicz T, Hsu LL, Smith-Whitley K,
Mahoney D, Woods G, Watanabe M, DeBaun MR
(2006) Exchange blood transfusion compared with
simple transfusion for first overt stroke is associated
with a lower risk of subsequent stroke: a retrospective
cohort study of 137 children with sickle cell anemia.
J Pediatr 149:710–712
25. Adams RJ, McKie VC, Hsu L, Files B, Vichinsky E,
Pegelow C, Abboud M, Gallagher D, Kutlar A,
Nichols FT, Bonds DR, Brambilla D (1998) Preven-
tion of a first stroke by transfusions in children with
sickle cell anemia and abnormal results on
transcranial Doppler ultrasonography. N Engl J Med
339:5–11
26. Peinemann F, Bartel C, Grouven U (2013) First-line
allogeneic hematopoietic stem cell transplantation of
HLA-matched sibling donors compared with first-line
ciclosporin and/or antithymocyte or antilymphocyte
A. Fujita et al.
globulin for acquired severe aplastic anemia.
Cochrane Database Syst Rev 7, CD006407
27. Webb J, Kwiatkowski JL (2013) Stroke in patients
with sickle cell disease. Expert Rev Hematol
6:301–316
28. Arkuszewski M, Melhem ER, Krejza J (2010)
Neuroimaging in assessment of risk of stroke in chil-
dren with sickle cell disease. Adv Med Sci
55:115–129
29. Smajlovic D, Sinanovic O (2004) Sensitivity of the
neuroimaging techniques in ischemic stroke. Med
Arh 58:282–284
30. Paonessa A, Limbucci N, Tozzi E, Splendiani A,
Gallucci M (2010) Radiological strategy in acute
stroke in children. Eur J Radiol 74:77–85
31. Roach ES, Golomb MR, Adams R, Biller J, Daniels S,
Deveber G, Ferriero D, Jones BV, Kirkham FJ, Scott
RM, Smith ER, American Heart Association Stroke
Council, Council on Cardiovascular Disease in the
Young (2008) Management of stroke in infants and
children: a scientific statement from a Special Writing
Group of the American Heart Association Stroke
Council and the Council on Cardiovascular Disease
in the Young. Stroke 39:2644–2691
32. Gonzalez RG, Schaefer PW, Buonanno FS,
Schwamm LH, Budzik RF, Rordorf G, Wang B,
Sorensen AG, Koroshetz WJ (1999) Diffusion-
weighted MR imaging: diagnostic accuracy in
patients imaged within 6 hours of stroke symptom
onset. Radiology 210:155–162
33. Rydberg JN, Hammond CA, Grimm RC, Erickson
BJ, Jack CR Jr, Huston J 3rd, Riederer SJ (1994)
Initial clinical experience in MR imaging of the
brain with a fast fluid-attenuated inversion-recovery
pulse sequence. Radiology 193:173–180
34. Cheng AL, Batool S, McCreary CR, Lauzon ML,
Frayne R, Goyal M, Smith EE (2013) Susceptibility-
weighted imaging is more reliable than T2*-weighted
gradient-recalled echo MRI for detecting
microbleeds. Stroke 44:2782–2786
35. Strouse JJ, Lanzkron S, Urrutia V (2011) The epide-
miology, evaluation and treatment of stroke in adults
with sickle cell disease. Expert Rev Hematol
4:597–606
36. Frietsch T, Ewen I, Waschke KF (2001) Anaesthetic
care for sickle cell disease. Eur J Anaesthesiol
18:137–150
37. Maeda M, Tsuchida C (1999) “Ivy sign” on fluid-
attenuated inversion-recovery images in childhood
moyamoya disease. AJNR Am J Neuroradiol
20:1836–1838
38. Brandao RA, de Carvalho GT, Reis BL, Bahia E, de
Souza AA (2009) Intracranial aneurysms in sickle cell
patients: report of 2 cases and review of the literature.
Surg Neurol 72:296–299 (discussion 299)
39. Santhosh K, Kesavadas C, Thomas B, Gupta AK,
Thamburaj K, Kapilamoorthy TR (2009) Susceptibil-
ity weighted imaging: a new tool in magnetic reso-
nance imaging of stroke. Clin Radiol 64:74–83
21 Sickle Cell Disease and Stroke
40. Lu L, Zhang LJ, Poon CS, Wu SY, Zhou CS, Luo S,
Wang M, Lu GM (2012) Digital subtraction CT angi-
ography for detection of intracranial aneurysms: com-
parison with three-dimensional digital subtraction
angiography. Radiology 262:605–612
41. Sailer AM, Wagemans BA, Nelemans PJ, de Graaf R,
van Zwam WH (2014) Diagnosing intracranial aneu-
rysms with MR angiography. systematic review and
meta-analysis. Stroke 45:119–126
42. Kwiatkowski JL, Zimmerman RA, Pollock AN,
Seto W, Smith-Whitley K, Shults J, Blackwood-
Chirchir A, Ohene-Frempong K (2009) Silent infarcts
in young children with sickle cell disease. Br J
Haematol 146:300–305
43. Pegelow CH, Wang W, Granger S, Hsu LL,
Vichinsky E, Moser FG, Bello J, Zimmerman RA,
Adams RJ, Brambilla D, Trial S (2001) Silent infarcts
in children with sickle cell anemia and abnormal
cerebral artery velocity. Arch Neurol 58:2017–2021
44. Dowling MM, Quinn CT, Rogers ZR, Buchanan GR
(2010) Acute silent cerebral infarction in children
with sickle cell anemia. Pediatr Blood Cancer
54:461–464
45. Kinney TR, Sleeper LA, Wang WC, Zimmerman RA,
Pegelow CH, Ohene-Frempong K, Wethers DL, Bello
JA, Vichinsky EP, Moser FG, Gallagher DM, DeBaun
MR, Platt OS, Miller ST (1999) Silent cerebral
infarcts in sickle cell anemia: a risk factor analysis.
The Cooperative Study of Sickle Cell Disease. Pedi-
atrics 103:640–645
46. Wang WC, Gallagher DM, Pegelow CH, Wright EC,
Vichinsky EP, Abboud MR, Moser FG, Adams RJ
(2000) Multicenter comparison of magnetic reso-
nance imaging and transcranial Doppler ultrasonog-
raphy in the evaluation of the central nervous system
in children with sickle cell disease. J Pediatr Hematol
Oncol 22:335–339
47. Arkuszewski M, Krejza J, Chen R, Ichord R,
Kwiatkowski JL, Bilello M, Zimmerman R, Ohene-
Frempong K, Melhem ER (2014) Sickle cell anemia:
Intracranial stenosis and silent cerebral infarcts in
children with low risk of stroke. Adv Med Sci
59:108–113
48. Schatz J, White DA, Moinuddin A, Armstrong M,
DeBaun MR (2002) Lesion burden and cognitive
morbidity in children with sickle cell disease.
J Child Neurol 17:891–895
49. DeBaun MR, Armstrong FD, McKinstry RC, Ware
RE, Vichinsky E, Kirkham FJ (2012) Silent cerebral
infarcts: a review on a prevalent and progressive
cause of neurologic injury in sickle cell anemia.
Blood 119:4587–4596
50. Steen RG, Miles MA, Helton KJ, Strawn S, Wang W,
Xiong X, Mulhern RK (2003) Cognitive impairment
in children with hemoglobin SS sickle cell disease:
relationship to MR imaging findings and hematocrit.
AJNR Am J Neuroradiol 24:382–389
51. Hulbert ML, McKinstry RC, Lacey JL, Moran CJ,
Panepinto JA, Thompson AA, Sarnaik SA,
465
Woods GM, Casella JF, Inusa B, Howard J,
Kirkham FJ, Anie KA, Mullin JE, Ichord R,
Noetzel M, Yan Y, Rodeghier M, Debaun MR
(2011) Silent cerebral infarcts occur despite regular
blood transfusion therapy after first strokes in children
with sickle cell disease. Blood 117:772–779
52. Casella JF, King AA, Barton B, White DA, Noetzel
MJ, Ichord RN, Terrill C, Hirtz D, McKinstry RC,
Strouse JJ, Howard TH, Coates TD, Minniti CP,
Campbell AD, Vendt BA, Lehmann H, Debaun MR
(2010) Design of the silent cerebral infarct transfusion
(SIT) trial. Pediatr Hematol Oncol 27:69–89
53. Steen RG, Emudianughe T, Hankins GM, Wynn LW,
Wang WC, Xiong X, Helton KJ (2003) Brain imaging
findings in pediatric patients with sickle cell disease.
Radiology 228:216–225
54. Gibbs WN, Opatowsky MJ, Burton EC (2012) AIRP
best cases in radiologic-pathologic correlation: cere-
bral fat embolism syndrome in sickle cell beta-
thalassemia. Radiographics 32:1301–1306
55. Mossa-Basha M, Izbudak I, Gurda GT, Aygun N
(2012) Cerebral fat embolism syndrome in sickle
cell anaemia/beta-thalassemia: importance of
susceptibility-weighted MRI. Clin Radiol
67:1023–1026
56. Currie S, Raghavan A, Batty R, Connolly DJ,
Griffiths PD (2011) Childhood moyamoya disease
and moyamoya syndrome: a pictorial review. Pediatr
Neurol 44:401–413
57. Amlie-Lefond C, Bernard TJ, Sebire G, Friedman
NR, Heyer GL, Lerner NB, DeVeber G, Fullerton
HJ, International Pediatric Stroke Study Group
(2009) Predictors of cerebral arteriopathy in children
with arterial ischemic stroke: results of the Interna-
tional Pediatric Stroke Study. Circulation
119:1417–1423
58. Dobson SR, Holden KR, Nietert PJ, Cure JK, Laver
JH, Disco D, Abboud MR (2002) Moyamoya syn-
drome in childhood sickle cell disease: a predictive
factor for recurrent cerebrovascular events. Blood
99:3144–3150
59. Lin N, Baird L, Koss M, Kopecky KE, Gone E,
Ullrich NJ, Scott RM, Smith ER (2011) Discovery
of asymptomatic moyamoya arteriopathy in pediatric
syndromic populations: radiographic and clinical pro-
gression. Neurosurg Focus 31:E6
60. Arias EJ, Derdeyn CP, Dacey RG Jr, Zipfel GJ (2014)
Advances and surgical considerations in the treatment
of moyamoya disease. Neurosurgery 74:S116–S125
61. Adelson PD, Scott RM (1995) Pial synangiosis for
moyamoya syndrome in children. Pediatr Neurosurg
23:26–33
62. Robertson RL, Burrows PE, Barnes PD, Robson CD,
Poussaint TY, Scott RM (1997) Angiographic
changes after pial synangiosis in childhood
moyamoya disease. AJNR Am J Neuroradiol
18:837–845
63. Kennedy BC, McDowell MM, Yang PH, Wilson CM,
Li S, Hankinson TC, Feldstein NA, Anderson RC
466
(2014) Pial synangiosis for moyamoya syndrome in
children with sickle cell anemia: a comprehensive
review of reported cases. Neurosurg Focus 36:E12
64. Calviere L, Viguier A, Guidolin B, Tall P, Larrue V
(2007) Cervical artery stenoses in sickle cell disease.
Eur Neurol 58:120–121
65. Deane CR, Goss D, Bartram J, Pohl KR, Height SE,
Sibtain N, Jarosz J, Thein SL, Rees DC (2010) Extra-
cranial internal carotid arterial disease in children
with sickle cell anemia. Haematologica
95:1287–1292
66. Telfer PT, Evanson J, Butler P, Hemmaway C,
Abdulla C, Gadong N, Whitmarsh S, Kaya B,
Kirkham FJ (2011) Cervical carotid artery disease in
sickle cell anemia: clinical and radiological features.
Blood 118:6192–6199
67. Verlhac S, Balandra S, Cussenot I, Kasbi F, Vasile M,
Kheniche A, Elmaleh-Berges M, Ithier G,
Benkerrou M, Bernaudin F, Sebag G (2014) Extracra-
nial carotid arteriopathy in stroke-free children with
sickle cell anemia: detection by submandibular Dopp-
ler sonography. Pediatr Radiol 44:587–596
68. Saito N, Nadgir RN, Flower EN, Sakai O (2010)
Clinical and radiologic manifestations of sickle cell
disease in the head and neck. Radiographics
30:1021–1034
69. Steen RG, Reddick WE, Glass JO, Wang WC (1998)
Evidence of cranial artery ectasia in sickle cell disease
patients with ectasia of the basilar artery. J Stroke
Cerebrovasc Dis 7:330–338
70. Frye RE (2009) Reversible posterior leukoence-
phalopathy syndrome in sickle-cell anemia. Pediatr
Neurol 40:298–301
71. Geevasinga N, Cole C, Herkes GK, Barnett Y, Lin J,
Needham M (2014) Sickle cell disease and posterior
reversible leukoencephalopathy. J Clin Neurosci
21:1329–1332
72. Henderson JN, Noetzel MJ, McKinstry RC, White
DA, Armstrong M, DeBaun MR (2003) Reversible
posterior leukoencephalopathy syndrome and silent
cerebral infarcts are associated with severe acute
chest syndrome in children with sickle cell disease.
Blood 101:415–419
73. Khademian Z, Speller-Brown B, Nouraie SM,
Minniti CP (2009) Reversible posterior leuko-
encephalopathy in children with sickle cell disease.
Pediatr Blood Cancer 52:373–375
74. Kolovou V, Zampakis P, Ginopoulou A,
Varvarigou A, Kaleyias J (2013) Reversible posterior
leukoencephalopathy syndrome after blood transfu-
sion in a pediatric patient with sickle cell disease.
Pediatr Neurol 49:213–217
75. Elias EJ, Liao JH, Jara H, Watanabe M, Nadgir RN,
Sakai Y, Erbay K, Saito N, Ozonoff A, Steinberg MH,
Sakai O (2013) Quantitative MRI analysis of cranio-
facial bone marrow in patients with sickle cell disease.
AJNR Am J Neuroradiol 34:622–627
76. Liao JH, Jara H, Nadgir R, Elias E, Nowrouzi N,
Saito N, Steinberg MH, Sakai O (2013) qMRI
A. Fujita et al.
relaxometry of mandibular bone marrow: a
monomodal distribution in sickle cell disease.
J Magn Reson Imaging 37:1182–1188
77. Marlow TJ, Brunson CY, Jackson S, Schabel SI
(1998) “Tower vertebra”: a new observation in sickle
cell disease. Skeletal Radiol 27:195–198
78. Jain R, Sawhney S, Rizvi SG (2008) Acute bone
crises in sickle cell disease: the T1 fat-saturated
sequence in differentiation of acute bone infarcts
from acute osteomyelitis. Clin Radiol 63:59–70
79. Saito N, Watanabe M, Liao J, Flower EN, Nadgir RN,
Steinberg MH, Sakai O (2011) Clinical and radiologic
findings of inner ear involvement in sickle cell dis-
ease. AJNR AmJ Neuroradiol 32:2160–2164
80. Le Bihan D, Mangin JF, Poupon C, Clark CA,
Pappata S, Molko N, Chabriat H (2001) Diffusion
tensor imaging: concepts and applications. J Magn
Reson Imaging 13:534–546
81. Balci A, Karazincir S, Beyoglu Y, Cingiz C,
Davran R, Gali E, Okuyucu E, Egilmez E (2012)
Quantitative brain diffusion-tensor MRI findings in
patients with sickle cell disease. AJR Am J
Roentgenol 198:1167–1174
82. Behpour AM, Shah PS, Mikulis DJ, Kassner A (2013)
Cerebral blood flow abnormalities in children with
sickle cell disease: a systematic review. Pediatr
Neurol 48:188–199
83. Kirkham FJ, Calamante F, Bynevelt M, Gadian DG,
Evans JP, Cox TC, Connelly A (2001) Perfusion
magnetic resonance abnormalities in patients with
sickle cell disease. Ann Neurol 49:477–485
84. Oguz KK, Golay X, Pizzini FB, Freer CA, Winrow N,
Ichord R, Casella JF, van Zijl PC, Melhem ER (2003)
Sickle cell disease: continuous arterial spin-labeling
perfusion MR imaging in children. Radiology
227:567–574
85. Strouse JJ, Cox CS, Melhem ER, Lu H, Kraut MA,
Razumovsky A, Yohay K, van Zijl PC, Casella JF
(2006) Inverse correlation between cerebral blood
flow measured by continuous arterial spin-labeling
(CASL) MRI and neurocognitive function in children
with sickle cell anemia (SCA). Blood 108:379–381
86. van den Tweel XW, Nederveen AJ, Majoie CB, van
der Lee JH, Wagener-Schimmel L, van Walderveen
MA, Poll The BT, Nederkoorn PJ, Heijboer H,
Fijnvandraat K (2009) Cerebral blood flow measure-
ment in children with sickle cell disease using contin-
uous arterial spin labeling at 3.0-Tesla MRI. Stroke
40:795–800
87. Sanelli PC, Sykes JB, Ford AL, Lee JM, Vo KD,
Hallam DK (2013) Imaging and treatment of patients
with acute stroke: an evidence-based review. AJNR
Am J Neuroradiol 35:1045–1051
88. Wang Z, Bogdan AR, Zimmerman RA, Gusnard DA,
Leigh JS, Ohene-Frempong K (1992) Investigation of
stroke in sickle cell disease by 1H nuclear magnetic
resonance spectroscopy. Neuroradiology 35:57–65
89. Wang Z, Zimmerman RA, Sauter R (1996) Proton
MR spectroscopy of the brain: clinically useful
21 Sickle Cell Disease and Stroke
information obtained in assessing CNS diseases in
children. AJR Am J Roentgenol 167:191–199
90. Steen RG, Ogg RJ (2005) Abnormally high levels of
brain N-acetylaspartate in children with sickle cell
disease. AJNR Am J Neuroradiol 26:463–468
91. Watanabe M, Liao JH, Jara H, Sakai O (2013) Multi-
spectral quantitative MR imaging of the human brain:
lifetime age-related effects. Radiographics
33:1305–1319
92. Steen RG, Langston JW, Ogg RJ, Xiong X, Ye Z,
Wang WC (1999) Diffuse T1 reduction in gray matter
of sickle cell disease patients: evidence of selective
vulnerability to damage? Magn Reson Imaging
17:503–515
93. Steen RG, Schroeder J (2003) Age-related changes in
the pediatric brain: proton T1 in healthy children and
in children with sickle cell disease. Magn Reson
Imaging 21:9–15
94. Steen RG, Emudianughe T, Hunte M, Glass J, Wu S,
Xiong X, Reddick WE (2005) Brain volume in pedi-
atric patients with sickle cell disease: evidence of
volumetric growth delay? AJNR Am J Neuroradiol
26:455–462
95. Liao J, Saito N, Ozonoff A, Jara H, Steinberg M,
Sakai O (2012) Quantitative MRI analysis of salivary
glands in sickle cell disease. Dentomaxillofac Radiol
41:630–636
96. Buch K, Watanabe M, Elias EJ, Liao JH, Jara H,
Nadgir RN, Saito N, Steinberg MH, Sakai O (2014)
Quantitative magnetic resonance imaging analysis of
the Lacrimal gland in sickle cell disease. J Comput
Assist Tomogr 38:674–680
97. Deus-Silva L, Bonilha L, Damasceno BP, Costa AL,
Yasuda CL, Costa FF, Santos AO, Etchebehere EC,
Oquendo-Nogueira R, Fockink R, de Freitas CF,
Camargo EE, Li LM, Cendes F, Saad ST (2013) Brain
perfusion impairment in neurologically asymptomatic
adult patients with sickle-cell disease shown by Voxel-
based analysis of SPECT images. Front Neurol 4:207
98. Kedar A, Drane WE, Shaeffer D, Nicole M, Adams C
(2006) Measurement of cerebrovascular flow reserve
in pediatric patients with sickle cell disease. Pediatr
Blood Cancer 46:234–238
99. Powars DR, Conti PS, Wong WY, Groncy P, Hyman C,
Smith E, Ewing N, Keenan RN, Zee CS, Harold Y, Hiti
AL, Teng EL, Chan LS (1999) Cerebral vasculopathy
in sickle cell anemia: diagnostic contribution of posi-
tron emission tomography. Blood 93:71–79
100. Reed W, Jagust W, Al-Mateen M, Vichinsky E (1999)
Role of positron emission tomography in determining
the extent of CNS ischemia in patients with sickle cell
disease. Am J Hematol 60:268–272
467
101. Quinn CT, Rogers ZR, Buchanan GR (2004) Survival
of children with sickle cell disease. Blood
103:4023–4027
102. Switzer JA, Hess DC, Nichols FT, Adams RJ (2006)
Pathophysiology and treatment of stroke in sickle-cell
disease: present and future. Lancet Neurol 5:501–512
103. Bernaudin F, Verlhac S, Arnaud C, Kamdem A,
Chevret S, Hau I, Coic L, Leveille E,
Lemarchand E, Lesprit E, Abadie I, Medejel N,
Madhi F, Lemerle S, Biscardi S, Bardakdjian J,
Galacteros F, Torres M, Kuentz M, Ferry C,
Socie G, Reinert P, Delacourt C (2011) Impact of
early transcranial Doppler screening and intensive
therapy on cerebral vasculopathy outcome in a new-
born sickle cell anemia cohort. Blood
117:1130–1140, quiz 436
104. Sheehan VA, Hansbury EN, Smeltzer MP, Fortner G,
McCarville MB, Aygun B (2013) Transcranial Dopp-
ler velocity and brain MRI/MRA changes in children
with sickle cell anemia on chronic transfusions to
prevent primary stroke. Pediatr Blood Cancer
60:1499–1502
105. Adams RJ, Nichols FT, Figueroa R, McKie V, Lott T
(1992) Transcranial Doppler correlation with cerebral
angiography in sickle cell disease. Stroke
23:1073–1077
106. Jones A, Granger S, Brambilla D, Gallagher D,
Vichinsky E, Woods G, Berman B, Roach S,
Nichols F, Adams RJ (2005) Can peak systolic veloc-
ities be used for prediction of stroke in sickle cell
anemia? Pediatr Radiol 35:66–72
107. Ishola T, Quinn CT (2013) Transcranial Doppler peak
systolic velocities overestimate the risk of stroke in
sickle cell anemia. Blood 122:2240
108. Scothorn DJ, Price C, Schwartz D, Terrill C,
Buchanan GR, Shurney W, Sarniak I, Fallon R, Chu
JY, Pegelow CH, Wang W, Casella JF, Resar LS,
Berman B, Adamkiewicz T, Hsu LL, Ohene-
Frempong K, Smith-Whitley K, Mahoney D, Scott
JP, Woods GM, Watanabe M, Debaun MR (2002)
Risk of recurrent stroke in children with sickle cell
disease receiving blood transfusion therapy for at
least five years after initial stroke. J Pediatr
140:348–354
109. Madden NA, Jones GL, Kalpatthi R, Woods G (2014)
Practice patterns of stroke screening and hydroxyurea
use in children with sickle cell disease: a survey of
health care providers. J Pediatr Hematol Oncol 36:
e382–e386
110. Ware RE, Helms RW, Investigators SW (2012) Stroke
with transfusions changing to Hydroxyurea
(SWiTCH). Blood 119:3925–3932
 Hypoxic–Ischemic Encephalopathy
(Preterm, Term, and Adult) 22
Mauricio Castillo and Francisco Chiang

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470 Hypoxic–ischemic injury (HII) to the brain is
usually a devastating event and an important
Physiopathology of HII and Factors
Influencing the Patterns of Injury . . . . . . . . . . . . . . . . . 470
cause of morbidity and mortality in the United
States and elsewhere in the world. Neuroimag-
HII in the Preterm Neonate . . . . . . . . . . . . . . . . . . . . . . . . 472 ing plays a pivotal role in diagnosis, treatment,
Severe Asphyxia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 472
Mild-to-Moderate Asphyxia . . . . . . . . . . . . . . . . . . . . . . . . . 473 and long-term prognosis determination for
these patients. The correct diagnosis made on
HII in the Term Neonate . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
Severe Asphyxia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
the basis of different imaging modalities
Partial Asphyxia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 482 requires knowledge of the different manifesta-
tions of this type of injury. Some of the factors
HII in Postnatal Infants and Young Children . . . . 483
Severe Asphyxia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483 that contribute to the different findings are
Mild-to-Moderate Asphyxia . . . . . . . . . . . . . . . . . . . . . . . . . 484 brain maturity, duration and severity of the
HII in Older Children and Adults . . . . . . . . . . . . . . . . 485 insult, underlying cause, and associated
disorders.
Delayed White Matter Injury . . . . . . . . . . . . . . . . . . . . . 486
Severe HII will result in preferentially deep
Role of Proton MR Spectroscopy in gray matter damage in preterm and term
the Evaluation of HII . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488 infants, with peri-rolandic involvement more
Imaging Choices in HII . . . . . . . . . . . . . . . . . . . . . . . . . . . . 490 frequently observed in the latter age group. In
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 492 these patients, a less profound insult will result
in germinal matrix hemorrhages or
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 492
periventricular leukomalacia (PVL) in preterm
neonates and parasagittal watershed infarcts in
term neonates. In the postnatal period, severe
insults produce diffuse gray matter injury, with
relative sparing of the peri-rolandic cortex and
posterior circulation structures. In older chil-
dren and adults, profound insults produce
injury in the deep gray matter nuclei, cortices,
M. Castillo (*) • F. Chiang hippocampi, and cerebellum.
Division of Neuroradiology, Department of Radiology,
The use of advanced MRI techniques such
University of North Carolina, Chapel Hill, NC, USA
e-mail: mauricio_castillo@med.unc.edu; as DWI and MR spectroscopy is useful in
franciscochiang@gmail.com

# Springer Science+Business Media New York 2016 469

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_28
470
making the diagnosis especially in the acute
setting where conventional imaging might be
less sensitive.
Keywords
HII • Hypoxic–ischemic injury • Brain ische-
mia • Global brain hypoxia
Introduction
Hypoxic–ischemic injury (HII) to the brain is
usually a devastating event, still considered the
third leading cause of death in the United States
with approximately half a million new victims per
year. Death will occur in nearly one third of these
patients, while another third will end up suffering
severe neurologic deficits with important func-
tional impairment [1]. Even more, perinatal
asphyxia is considered an important risk factor
for cognitive and behavioral difficulties in surviv-
ing children [2].
This serious and life-threatening condition is
most often caused by insults as cardiac arrest,
asphyxia, poisoning (drug overdose or carbon
monoxide intoxication), and head trauma. Sup-
portive care is the most used primary treatment
strategy, which in majority of cases fails to pre-
vent the brain injury that quickly follows the
primary insult. Other treatment techniques are
being studied, including promising new targeted
neuroprotective approaches such as hypothermia
and excitatory amino acid antagonists. However,
most of these new treatments have a limited win-
dow of time to be effective even as little as 6 h post
insult [3]. For this reason, times spent in the trans-
portation of patients to the hospital and early
detection of this injury are extremely important
to the final outcomes of these patients [4]. As
treatment protocols seize being based on time
and become based on tissue status, neuroimaging
plays an important role with ultrasonography
(US), computed tomography (CT), and magnetic
resonance (MR) imaging permitting early inter-
vention. Also, imaging in the subacute stages is
useful in predicting long-term outcomes of these
patients by providing precise information about
the severity and extent of injury [5–8].
M. Castillo and F. Chiang
It is necessary to emphasize, however, that
findings in HII are variable. They depend on
many different factors such as age (brain matu-
rity), duration, severity, and exact type of insult
and also the modality and timing of the imaging
studies. Because of this difficult imaging scenario
and the necessity of making a fast and accurate
diagnosis, the radiologist requires a thorough
knowledge of the imaging patterns of HII and be
able to identify subtle findings by looking for the
specific areas most likely to be injured when HII is
suspected clinically.
In this chapter, first the pathophysiologic fea-
tures of HII and how they influence the imaging
patterns are discussed. Then, the specific imaging
manifestations in term (>36 weeks gestational
age) and preterm (<36 weeks gestational age)
neonates, postnatal infants, children, and adults
are reviewed.
Physiopathology of HII and Factors
Influencing the Patterns of Injury
Earlier in this chapter, some of the most important
causes of HII were mentioned. Independent from
the underlying causes, the main pathophysiologic
processes that lead to HII are remarkably similar.
The main two processes that cause HII are ische-
mia and hypoxemia. Depending on the type of
injury, the duration of injury, and the patient’s
age (brain maturity), these two processes will
play different roles influencing the pathologic
and imaging findings.
In general, infants and small children are more
inclined to suffer asphyxia, due to perinatal
asphyxia or being traumatic in nature (pool acci-
dents, etc.). In asphyxia, the chain of events that
lead to brain HII starts with hypoxemia. If the
hypoxemia is maintained long enough and if the
redistribution of oxygenated blood fails to com-
pensate for it, cardiac hypoxia and decreased car-
diac output occur. After this, brain ischemia
ensues due to decreased blood flow. For this rea-
son, the type of brain injury caused by asphyxia is
ischemia superimposed on hypoxia. Even more,
there is evidence that ischemia superimposed on
hypoxia is required for the HII to occur as acute
22 Hypoxic–Ischemic Encephalopathy (Preterm, Term, and Adult)
hypoxia alone is unlikely to cause damage to the
neonatal brain unless it is prolonged. This reflects
the resistance of the immature brain to hypoxia
when compared to the adult brain as well as better
compensatory mechanisms in babies [9]. In con-
trast to this, in adults, direct blood flow reductions
as in cardiac arrest or cerebrovascular disease,
which lead to primary ischemia with secondary
hypoxia, are common underlying mechanisms.
Another important aspect of HII physiopathol-
ogy is that not all brain structures are affected to
the same degree in a global hypoxic–ischemic
event. It is known that there are certain tissues
more likely to be injured or injured earlier because
of their different sensitivity to hypoxia. This con-
cept is known as selective vulnerability. HII
causes damage through a process called
excitotoxicity [9]. This process refers to cellular
death due to the excessive stimuli of certain amino
acid receptors that normally mediate physiologic
excitatory effects of dicarboxylic acid glutamate,
a ubiquitous neurotransmitter in the brain. This
stimuli end up triggering a complex cascade of
biochemical events that finally produces selective
death of certain populations of neurons. This pro-
cess appears to be even more important in devel-
oping brains than in adults.
Brain ischemia, on the other hand, produces a
change from oxidative phosphorylation to
anaerobic metabolism which is inefficient. Rapid
depletion of adenosine triphosphate (ATP) occurs,
and lactate accumulates in cells with loss of
normal functions in their membrane. The failure
to produce ATP and the subsequent depolarization
of the neuronal cell membranes cause excessive
release of glutamate from nerve terminals, which
in combination with the reduced activity of glial
pumps that normally keep synaptic glutamate
levels low produces a rapid increase in glutamate
concentration. When this process occurs,
neurons and other cells with the appropriate
receptors die due to a complex neurotoxic cas-
cade. In immature brains, glutamate and reduced
membrane potential contribute to the opening of
N-methyl-D-aspartate (NMDA) receptors which
cause a massive influx of calcium (Ca2+) into
neurons. Direct effects of Ca2+ include generation
of oxygen free radicals such as nitric oxide and
471
peroxynitrite. Also, damage to the mitochondria
may diminish their capacity of handling multiple
oxidation reactions and results in further loss of
ATP enabling the production of more free radicals
and perpetuating the cycle of membrane depolar-
ization and NMDA receptor channel opening.
Finally, delayed depletion of energy and elevated
brain lactate levels are associated with neuronal
death and resultant neurodegeneration after
asphyxia.
Finally, from all of these biochemical and path-
ophysiological data mentioned before, there are
some important conclusions to be kept in mind
when analyzing the imaging patterns of HII.
First, some areas of the brain are more susceptible
to ischemic injury than others primarily
because of their concentrations of glutamate and
other excitatory amino acid receptors (primarily
located in the gray matter), a concept known as
selective vulnerability. Second, there are some
areas of the brain with more energy demand
than others. In these, energy will be depleted
faster and therefore the injury will ensue earlier.
Third, due to delayed neuron death (apoptosis),
not all of the injury is evident until days after the
initial insult.
It is important to realize that in any given
patient, the sites of the brain that are most vulner-
able and are first affected by HII will be deter-
mined by the degree of maturity of the brain,
which depends on the age of the patient. This is
one of the reasons why HII manifestations in the
perinatal period (up to 1 month of age) differ from
those seen in older infants and in adults. Even
between term and preterm infants, there are
important differences in patterns of injury. It is
relevant to be cognizant of the degree of brain
maturity to correctly interpret the studies for
suspected HII. Another important factor determin-
ing the distribution of injuries is the severity and
duration of the HII. Severe episodes of
hypoxia–ischemia produce different injury pat-
terns when compared to lesser ones. In insults of
short duration, usually there is no brain injury
especially in the immature brain due to its
enhanced resistance to hypoxia. In the pediatric
population, the time of hypoxia–ischemia
required to produce brain injury has been
472
estimated to be approximately 15 min and in
adults 3–4 min [10]. With all these factors that
influence the patterns of injury in mind, we can
now review the different manifestations of HII in
imaging studies in preterm, term, and older
children and adults.
HII in the Preterm Neonate
Preterm neonates are susceptible to suffer HII, and
this type of injury is more common in this group
than in term neonates. The consequences of HII
can be devastating, as at least 5 % of infants born
before 32 weeks gestational age and approxi-
mately 19 % of preterm neonates less than
28 weeks develop cerebral palsy. If one takes all
cases of cerebral palsy, approximately 50 % of
them occur in preterm neonates [11]. This situa-
tion can be explained mainly by two reasons. First,
in this group, there is increased frequency in events
leading to HII, such as respiratory distress syn-
drome, pneumothorax, patent ductus arteriosus,
pregnancy infection, etc. Second, preterm infants
have less autoregulation and compensatory capac-
ity when compared to term neonates. Also, in this
age group, making a diagnosis is more difficult
because the clinical examination is limited, signs
of HII may be lacking, or if present, they can be
mistaken for developmental immaturity. Further-
more, preterm infants not uncommonly have tran-
sient neurological abnormalities and decreased
muscle tone as part of their immature central ner-
vous system, and most of these patients do not
develop long-term deficits. For this is the reason,
imaging is the key to arrive at the correct diagnosis.
Imaging manifestations of HII in preterm
babies differ from those seen in term infants,
mainly because of their degree of immaturity. In
terms of the imaging findings, HII can be divided
in two groups: those with severe and those with
mild-to-moderate asphyxia. One must realize that
this is an arbitrary and somewhat simplistic divi-
sion because in any given patient, there is a con-
tinuous range of insults that are variable in terms
of duration and severity, and therefore it is not
surprising to have different imaging features in the
same patient.
M. Castillo and F. Chiang
Severe Asphyxia
The injury pattern of preterm babies following a
severe but brief hypoxic event resembles the pat-
tern seen in term infants (see below) but with
some differences [12]. These patients often show
lesions in the basal ganglia, thalami, brainstem
structures, cerebellum, and corticospinal tracts, as
well as decreased cerebral hemispheric white mat-
ter. Although basal ganglia injury is frequent, it is
less severe than the involvement of the thalami in
this age group and especially among those less than
32 weeks of age. When involved, the basal ganglia
tend to atrophy without scarring. Overall, the thal-
ami, anterior vermis, and dorsal brainstem are the
most commonly involved structures when pro-
found asphyxia happens [3, 13]. The peri-rolandic
cortex is usually spared in these patients. Germinal
matrix hemorrhages and periventricular white mat-
ter injuries may also be seen in this group.
The reason for the differences in the extent of
injury between the putamen and thalamus that is
seen in these patients is likely related to the degree
of myelination of these structures. The thalamus,
the pallidothalamic fibers of the posterior internal
capsule, and globus pallidus myelinate at about
24–25 weeks of gestation, whereas the corpus
striatum (caudate nucleus and putamen) does not
myelinate until 35–36 weeks. Myelinated struc-
tures are more metabolically active, as shown in
F-18 FDG positron emission tomography studies
[14], when compared to less myelinated areas, as
the putamen and caudate, and therefore they are
more susceptible to hypoxic injury [15].
Usually, the first imaging study performed in
preterm neonates is a cranial transfontanelle
US. As this is a noninvasive, bedside examina-
tion, it can be easily used in the intensive care unit
as a screening tool to determine significant brain
injury. The findings in US suggesting HII include
increased echogenicity in the thalami at 2–3 days
following the insult; nevertheless, US may appear
normal, especially in the first 2 days. MRI is
generally the next study performed in preterm
neonates with suspected HII and is complimen-
tary to US and necessary to accurately assess the
extent of the injury and predict outcomes espe-
cially cognitive delay [16].
22 Hypoxic–Ischemic Encephalopathy (Preterm, Term, and Adult)
Fig. 1 Severe asphyxia in a preterm neonate. DWI image
shows hyperintensity in the thalami
Conventional MRI sequences can be normal or
show only subtle abnormalities within the first day
after injury. DWI is the most sensitive sequence in
the first 24 h showing abnormalities in the thalami
(Fig. 1). Then, after approximately 2 days post
injury, T2 hyperintensity can be seen in the thal-
ami and basal ganglia. By the third day, T1 short-
ening ensues in the injured areas. As in term
infants, the DWI abnormalities peak at about
3–5 days and then tend to pseudo-normalize
[3, 10, 17]. Thereafter, T2 hyperintensity appears
at approximately 7 days and T1 shortening per-
sists into the chronic stage (Fig. 2).
Mild-to-Moderate Asphyxia
The most characteristic pattern of injury in mild-
to-severe asphyxia in preterm babies is deter-
mined by direct injury and hemorrhage of the
germinal matrix. The germinal matrix is a highly
cellular region that lines the walls of the lateral
ventricles in fetal life and from where neurons and
glia arise and migrate. It is most active between
the second half of the first trimester and the first
half of the second trimester. Thereafter, it starts to
473
Fig. 2 T1WI in a preterm neonate who suffered severe
asphyxia shows hyperintensity in the thalami and posterior
lentiform nuclei. The high signal in T1 can persist into the
chronic stage
involute by the first half of the third trimester so
that by the 34th week, the germinal zones have
almost completely involute. For this reason, ger-
minal matrix hemorrhages are infrequent before
this age [3]. A very important anatomic landmark
to recognize is the caudothalamic notch. The
caudothalamic notch is a groove where the last
remnant of the germinal matrix to involute – the
ganglionic eminence – is located. The
caudothalamic notch is a groove located between
the caudate head and the thalamus and is where
most of germinal matrix hemorrhages originate.
The pathogenesis of germinal matrix hemorrhage
is related to the relative higher vascularization of
this region and the properties of the vascular bed
in it. The capillaries in this region are fragile,
mainly because they are lined only by simple
endothelium and lack the muscular or collagenous
layers that are present in the larger blood vessels.
Second, cells that compose their endothelium
have high concentration of mitochondria
reflecting their high oxidative metabolic require-
ments. This makes them susceptible to hypoxic
conditions. It is believed that first, this fragile
endothelium suffers a loss of integrity due to
474
hypoxia and then with restoration of the normal
circulation by resuscitation, bleeding ensues. This
hemorrhage can be localized in the caudothalamic
notch or extend to the ventricles. The prevalence
of intraventricular hemorrhage in preterm neo-
nates weighting less than 2 Kg has been estimated
at approximately 25 % and most of them are
related to hemorrhages of the germinal matrix.
Also, it is known that most hemorrhages happen
within the first 24 h of life and that infants who are
very premature and with a very low birth weight
are at higher risk for developing intraventricular
hemorrhage [18, 19].
Germinal matrix hemorrhages are divided in
four grades reflecting their locations and degree of
dilatation of the ventricles (Table 1).
Grades I–III are hemorrhages that arise from
the germinal matrix and have variable extension
to the lateral ventricles. Grade IV hemorrhages are
not germinal matrix hemorrhages but are
Table 1 Germinal matrix hemorrhage (GMH) –
periventricular hemorrhage (IVH) grading
Grade I Subependymal GMH (mostly in the
caudothalamic groove)
Grade II GMH and IVH with or without mild
ventriculomegaly
Grade III GMH and IVH with ventriculomegaly
Grade IV Above + periventricular parenchymal
hemorrhagic infarction (not true GMH)
Fig. 3 US image in a preterm patient. Coronal (a) and
sagittal (b) images demonstrate bilateral areas of
subependymal echogenicity, right greater than left. The
M. Castillo and F. Chiang
parenchymal periventricular hemorrhagic
infarcts, probably venous in origin, with extension
to the ventricular system. There is a correlation
between higher hemorrhage grade and higher
perinatal mortality rates as well as a higher prev-
alence of long-term neurological sequelae [10].
Germinal matrix and intraventricular hemor-
rhages can be adequately evaluated with cranial
US (Figs. 3, 4, and 5), keeping in mind that
sometimes the findings can be subtle and difficult
to visualize. US of the posterior fossa by a poste-
rior fontanelle approach, as a complement to the
classic anterior fontanelle examination, can help
to better visualize the posterior supra- and
infratentorial structures. This can help to diagnose
subtle intraventricular hemorrhages when the ven-
tricles are not dilated and cerebellar hemorrhages
that are believed to be underdiagnosed. These last
types of hemorrhages are clinically silent but are
not uncommon and are recognized in 10–20 % of
autopsies. In fact, cerebellar hemorrhages are
believed to be no different in origin than
caudothalamic notch hemorrhages also arising
from germinal matrix remnants within the exter-
nal granule cell layer of the cerebellum and in the
subependymal layer of the roof of the fourth ven-
tricle [3, 20, 21]. Cerebellar hemorrhages are seen
as lentiform or crescent-shaped hyperechoic
lesions located posterior and peripheral in the
cerebellar hemispheres. MRI is usually the next
sagittal image confirms the location in the caudothalamic
groove. Choroid plexus is large and thought not to be
related to the hemorrhage. GMH grade I
22 Hypoxic–Ischemic Encephalopathy (Preterm, Term, and Adult) 475

Fig. 4 US coronal (a) and sagittal (b) images in a preterm neonate demonstrate extension of the left side hemorrhage into
the lateral ventricles, right greater than left. The ventricles are not enlarged. GMH grade II

Fig. 5 US coronal (a) and sagittal (b) images in a preterm neonate demonstrate bilateral intraventricular hemorrhage with
enlargement of the lateral ventricles. GMH grade III

study used most of the times to detect concomitant
injuries such as white matter injury of prematurity
or deep gray matter injury.
Another common manifestation that can be
seen in mild-to-moderate asphyxia in preterm
babies is periventricular leukomalacia (PVL),
also known as white matter injury of prematurity.
This type of injury also appears to be inversely
related to gestational age at birth. Its pathogenesis
is believed to be related to selective vulnerability
of oligodendrocyte precursors and to perturba-
tions in cerebral blood flow in a context of
anatomic and physiological immaturity of the
blood vessels in preterm patients [22, 23]. These
oligodendrocyte precursors are late precursors
known as preoligoendrocytes, and the white mat-
ter in the period prior to myelination is populated
with them. This is the period of higher risk for
PVL, as these late precursor cells are believed to
be even more susceptible to hypoxia than the
earlier precursor cells or the mature oligodendro-
cytes [24]. This explanation is supported by the
fact that the prevalence of PVL declines after
32 weeks, the same time that the population of
476
Fig. 6 US coronal (a) and sagittal (b) images in a preterm neonate demonstrate the third stage of PVL with the
development of bilateral periventricular cysts
these late precursors in the periventricular white
matter maturates into oligodendrocytes. Also,
damage to a particular subpopulation of vulnera-
ble neurons plays a role in the development of
PVL. These are called subplate neurons and they
contribute to cortical development and in particu-
lar to the formation of thalamocortical connec-
tions. They form a transient cell population that
peaks at approximately 24 weeks (the onset of the
developmental window of vulnerability) and later
undergoes apoptosis [25]. The subplate which can
reach up to four times the width of the cortical
plate has been shown by MRI to be affected by
hypoxic injury.
PVL is most commonly seen in the peritrigonal
region and adjacent to the foramina of Monro [33,
37]. It can have a cavitary or non-cavitary presen-
tation, this last type being more frequent. The
most commonly encountered neurological
sequelae are motor and visual impairments
because of the direct injury to the corticospinal
tracts and geniculocalcarine tracts that pass
through affected regions in the periventricular
white matter [3, 26]. Spastic diplegia is also a
common motor sequelae of PVL, in which the
degree of motor impairment is greater in the
lower extremities, and occurs more frequently in
preterm infants with PVL than in term infants
[3, 11]. At a histological level, PVL evolves first
with necrosis and cavitation that thereafter pro-
gress to porencephalic cysts. Later, these cysts
M. Castillo and F. Chiang
collapse resulting in gliosis and loss of white
matter volume that is seen in imaging studies [3,
26].
By US, there are four stages of PVL that some-
what correlate with its histological characteristics.
First, there is congestion in the periventricular
white matter, which in US is seen as increased
echogenicity that usually adopts an elongated and
globular morphology, sometimes referred as
“flares.” This increased echogenicity usually is
seen in the first 48 h. In the second stage, there is
a relative return to normal which occurs mostly by
the 2–4 weeks. In the third stage, the development
of cysts is evident in US at approximately 3–6
weeks (Fig. 6). Finally, in the fourth and last stage,
there is resolution of the cysts, with evidence of
volume loss with enlargement of the lateral ven-
tricles. This last stage happens at approximately
6 months of age [27].
US is usually used as the first examination in
evaluating suspected HII cases. Nevertheless, it
lacks the sensitivity and positive predictive value
and, as mentioned before, the study can be normal
in patients that develop PVL. Conversely, in other
cases, US shows increased echogenicity in the
periventricular areas of normal neonates. The
presence of prolonged hyperechogenicity of the
periventricular white matter has a fairly low sen-
sitivity and positive predictive value for the detec-
tion of PVL [28]. Serial US examinations improve
substantially the detection of transient cystic
22 Hypoxic–Ischemic Encephalopathy (Preterm, Term, and Adult) 477

Fig. 7 (a, b) T2WI of a preterm neonate who suffered dark fluid levels can be seen inside the lateral ventricles
mild-to-moderate asphyxia shows T2 hyperintensity in the compatible with intraventricular hemorrhage. Small left
periventricular white matter in the setting of PVL. Also, side periventricular cyst is present

lesions and can be better than MRI studies for this

purpose. This is important because it has prognos-
tic value as most of patients with cystic changes
present neurological sequelae [29]. For these rea-
sons, the primary roles of US are to detect germi-
nal matrix hemorrhages in the immediate
postnatal period and detect cystic changes later
in perinatal life [3].
MRI allows better visualization of the
periventricular white matter lesions and is a useful
complement to cranial US especially among
patients without cystic lesions. It also allows
better depiction of hemorrhages and/or white
matter volume loss which also has prognostic
value [29]. In MR images, early injury to the
white matter appears as foci of T1 hyperintensity
in the larger areas of T2 hyperintensity. These T1 Fig. 8 T2WI of a preterm neonate who suffered moderate
asphyxia demonstrates characteristic hyperintensity in the
hyperintense foci must be distinguished periventricular white matter, more pronounced at the level
from hemorrhages, and they do not produce T2 of the peritrigonal regions compatible with PVL. Also,
shortening. These T1 abnormalities may represent note the “wavy” appearance of the ventricular walls
focal areas of mineralization, while in the white
matter reactive gliosis develops [30]. These CT is usually avoided in neonates because of
changes are usually evident at the third–fourth the exposure to ionizing radiation. It does not
days post injury, and then they give way to a mild provide much more information than the US and
T2 shortening of the white matter at days 6–7. The MRI, but it could be important and helpful in
high T2 signal is most evident in the peritrigonal confirming PVL end-stage injuries later in life.
regions (Figs. 7 and 8). In the last stage of PVL, MRI and CT show a
478
characteristic loss of volume of the periventricular
white matter and centra semiovale with secondary
enlargement of the lateral ventricles in particular
their trigones. MRI and CT also show the charac-
teristic irregular outline and wavy appearance of
the outer wall of the lateral ventricles. MRI better
shows the loss of volume in the corpus callosum,
particularly in the posterior aspect of the body and
splenium [31].
HII in the Term Neonate
As mentioned before, HII is also considered an
important cause of death, neurodevelopmental dis-
orders, and disability in term neonates, although its
incidence and prevalence has declined over the last
decade and is now estimated to be between 2 and
4 per 1,000 live term births [3, 32, 33]. Risk factors
for HII can be divided in antepartum factors and
intrapartum factors. Antepartum risk factors
include maternal hypotension, infertility treatment,
multiple gestation, prenatal infection, gestation
41 weeks, and thyroid disease. Among the most
important intrapartum factors are forceps delivery,
breech extraction, umbilical cord prolapse,
abruptio placentae, tight nuchal cord, maternal
fever, prolonged membrane rupture, abnormal
cardiotocography, shoulder dystocia, and thick
meconium. The most popular hypothesis is that
most of the HII cases are attributable only to
antepartum risk factors; however, there are new
reports that point to the intrapartum factors as nec-
essary to develop this condition. In approximately
10 % of HII cases, there are postnatal complica-
tions such as sepsis, shock, and/or severe respira-
tory distress [33–35].
The clinical manifestations at birth of HII in
term infants include nonspecific signs and symp-
toms that evolve over a period of days. Data
suggest that the infants at risk for severe HII can
be reliably identified by a group of clinical man-
ifestations that include evidence of intrapartum
distress (e.g., fetal heart rate abnormality), severe
functional depression (low 5 min Apgar score),
need for resuscitation in the delivery room, severe
fetal acidemia, abnormal early neurologic exami-
nation, and abnormal electroencephalogram.
M. Castillo and F. Chiang
Table 2 HII in preterm neonates
Severe Injury in the deep gray matter, mostly
the thalami but also the basal ganglia,
dorsal brainstem, cerebellum, and
corticospinal tracts as well as a
diminished volume of the cerebral
hemispheric white matter
Mild to Germinal matrix hemorrhage
moderate IVH
PVL
These patients, in the first hours after a severe
insult, may present with depressed consciousness,
periodic breathing or apnea, or bradycardia. In
cases where severe injury to cortical regions has
ensued, hypotonia and seizures may occur. In
patients that survive, severe HII typically develops
including quadriparesis, choreoathetosis, severe
seizure, and/or mental retardation. In cases of mod-
erate HII, spastic diplegia or quadriplegia almost
always develops and is usually referred to as cere-
bral palsy. On the other hand, in mild cases of HII,
term infants may develop mild developmental
delay or recover completely.
The imaging patterns can be subdivided
depending on the severity of the hypoxic injury
into severe and partial asphyxia (Table 2).
In term neonates, transfontanelle US is the first
imaging study to be obtained when HII is
suspected. Although some abnormalities can be
detected by US, it has a low sensitivity and there-
fore a negative study should not be used as a
definite evidence of an absence of hypoxic injury.
If there is strong clinical suspicion of HII and US is
negative, MR imaging should be obtained to eval-
uate the presence and severity of the injury. It is
important to remember that, as mentioned before,
the biochemical and histological features of HII
that influence the imaging findings vary with time
so that a study performed only hours after the event
will be different from one done several days later.
Severe Asphyxia
In term neonates, severe asphyxia results mainly
in a central pattern of injury that usually involves
22 Hypoxic–Ischemic Encephalopathy (Preterm, Term, and Adult)
Fig. 9 (a, b) DWI in a term neonate with severe asphyxia demonstrates diffusion restriction in the ventrolateral thalami
and peri-rolandic cortex
the deep gray matter including the putamina, ven-
trolateral thalami, hippocampi, dorsal brainstem,
and lateral geniculate nuclei. Occasionally, the
peri-rolandic cortex is also involved. The expla-
nation for this pattern of injury is, as we men-
tioned before, the active state of myelination of
these areas and the high concentration of NMDA
receptors which makes them more susceptible to
neonatal HII [6, 36]. The rest of the cortex is
usually spared or shows mild abnormalities since
it is generally less metabolically active. However,
if the injury is prolonged, the remaining cortex
will be injured and portrays a worse
prognosis [10].
Transfontanelle US, although it is the most
commonly used technique and usually the first
one in cases of suspected HII, is less sensitive
(about 50 % in the first week of life) and specific
compared with CT and MRI and carries less
interobserver agreement [3, 37, 38]. Its sensitivity
increases when it is performed after 7 days. Early
US findings include a generalized increase in
cerebral echogenicity and diffuse cerebral edema
with obliteration of the cerebrospinal fluid (CSF)
containing spaces. Subtle increased echogenicity
in the basal ganglia, thalami, and brainstem can be
seen in the first week but are more apparent after
7 days [38, 39]. The presence of thalamic
479
hyperechogenicity generally suggests a severe
insult and poor outcome [40]. At a later stage,
the imaging pattern reflects the loss of volume
including prominence of the ventricles and extra
axial CSF-containing spaces, likely due to atro-
phy. Doppler US during the initial US examina-
tion may be useful and improves sensitivity and
specificity by showing diminished resistive
indexes (<60) in the anterior and middle cerebral
arteries. These lower resistive indexes have been
also associated with poorer clinical outcome, even
in absence of other US abnormalities [41].
MRI is probably the most accurate modality to
assess neonatal HII especially when performed
with diffusion-weighted imaging (DWI) in the
first 24 h, when DWI is most sensitive to detect
injuries which may still not be visible in conven-
tional T1- and T2-weighted images. DWI shows
high signal (with corresponding low ADC values)
in the ventrolateral thalami and basal ganglia (par-
ticularly the posterior putamina), peri-rolandic
regions, and along the corticospinal tracts
(Fig. 9). It is important to highlight the fact that
even with the high sensitivity of this technique,
the findings in DWI in the first 24 h usually
underestimate the ultimate extent of the injury,
and although rare, some reports of normal find-
ings in the first 24 h have been reported [17, 42].
480 M. Castillo and F. Chiang

Fig. 10 (a, b) T1WI showing hyperintensity in the posterior lentiform nuclei, ventrolateral thalami, and pericentral
cortex in a preterm neonate who suffered profound asphyxia

It is believed that the reason for this delay in

showing the full extent of the injury may be
based in the important role of apoptosis in HII,
and as explained before, the time that takes for
ATP to be depleted which precedes the death of
neurons and the resultant neurodegeneration at a
macroscopic level. Abnormalities on DWI peak at
3–5 days. By the end of the first week, the
hyperintensity in injured areas in DWI tends to
decrease, phenomenon known as “pseudo-nor-
malization” [4, 6, 10, 20]. It is important to realize
that although the images seemingly improve, this
does not imply that there is a real reversal or
improvement of the underlying injury, just reso-
lution of signal abnormalities on DWI. Because of
the rare possibility of a false-negative DWI study
Fig. 11 T1WI. A different patient showing T1
when performed in the first days, it is
hyperintensity in the posterior lentiform nuclei and ventro-
recommended to repeat the examination at 2–4 lateral thalami
days when the signal abnormality is expected to
be greatest or perform an evaluation with proton
MR spectroscopy (MRS). nuclei and ventrolateral thalami (Figs. 10 and 11).
It is well known that conventional MRI Sometimes, signal intensity changes may also be
sequences with T1- and T2-weighted images seen in the dorsal brainstem and basal ganglia
obtained within the first day are frequently normal [36]. The T2 hyperintensity usually develops
and therefore are less useful than DWI to diagnose later than the T1 shortening and usually by the
acute HII. By the second day, conventional second week, affects the thalami and posterior
sequences, especially T1-weighted images, start putamina. As mentioned before, cortical abnor-
to show hyperintensity in the posterior lentiform malities can also occur.
22 Hypoxic–Ischemic Encephalopathy (Preterm, Term, and Adult) 481

Fig. 12 (a, b) T2WI in a term neonate with chronic changes from HII shows hyperintensity in the corticospinal tracts,
putamina, and ventrolateral thalami. Also, some loss of volume is noted

The cause of the abnormalities seen with the

MRI conventional sequences in the basal ganglia
and thalami remains incompletely understood,
with possible explanations including hemorrhage,
transient calcium deposition, lipid release from
myelin breakdown, free fatty acids, and even
paramagnetic effects from free oxygen radicals.
In infants, cortical abnormalities are likely due to
laminar necrosis [36, 43]. A possible explanation
for the delay in appearance of the T2
hyperintensity changes is the high water content
of the white matter, so subtle abnormalities are
obscured and difficult to identify at first. The T1
shortening in posterior putamina, thalami, and
peri-rolandic cortex can persist for several
months. Because of all these reasons, DWI is
very useful in the first days, especially in the first Fig. 13 NECT in acute severe HII shows hypodensity of
24 h when conventional MR images are likely to the basal ganglia. CT is not usually the imaging modality of
be normal. On the other hand, conventional MRI choice in these patients
sequences are useful at the end of the first week
when the DWI images pseudo-normalize. Later,
in the chronic phase, the imaging findings reflect ganglia lesions includes HII and inborn errors of
atrophy of injured structures and T2 metabolism. The latter ones are suspected if there
hyperintensity especially in the ventrolateral thal- is no history of HII and if other imaging features
ami, posterior putamina, and corticospinal tracts outside of the typical spectrum of HII, like local-
(Figs. 12 and 13) [36]. The major imaging differ- ized white matter and cortical abnormalities, atro-
ential diagnosis in newborns with bilateral basal phy, or heterotopias, are seen.
482
Fig. 14 (a, b) DWI in a term neonate who suffered partial asphyxia demonstrates hyperintensity in watershed areas with
the corresponding low signal in the ADC map, compatible with watershed infarcts
Partial Asphyxia
When partial asphyxia occurs, the pattern of
injury changes and as mentioned before, the neo-
natal brain is more resistant to hypoxia than the
adult brain. For this reason, in mild or moderate
insults of short duration, there may be little or no
injury [10]. Compensatory mechanisms that take
place when hypoxia is established have been well
studied in animal models. With prolonged fetal
hypoxia, blood shunting to vital brain structures
occurs, including the brainstem, basal ganglia,
hippocampi, and cerebellum. Thus, less metabol-
ically active regions of the brain receive less blood
and are more susceptible to injury, specifically the
cortex and white matter. This is the reason why in
mild-to-moderate HII, the brainstem, cerebellum,
and deep gray matter are generally spared. When
the autoregulatory mechanisms are exceeded, the
result is injury to the watershed zones which
become relatively hypoperfused. The clinical
manifestations of this process generally are sei-
zures, hypotension, and possibly proximal
extremity weakness and/or spasticity [10].
US diagnosis of this type of injury is difficult as
this technique has low sensitivity for examination
of the cortical and subcortical areas that are close
to the calvarium. For this reason, MRI is the
M. Castillo and F. Chiang
modality of choice when studying term infants
with suspected partial asphyxia. Regarding the
MRI sequences that are more useful, again DWI
are the most sensitive and the first to show abnor-
malities in the first 24 h. These abnormalities
include hyperintensity with corresponding low
ADC values (diffusion restriction) in the water-
shed territories (Fig. 14). Interpretation of DWI is
sometimes difficult because of the high content of
water of the brain at this age which makes the
hyperintensity of HII subtle. To facilitate the
correct diagnosis, it is important to interpret
DWI with the corresponding ADC map
which will show areas of low signal confirming
true diffusion restriction [17]. T1 and T2 images
may be normal in the first 24 h, but by the
second day, they show T2 hyperintensity in the
cortex and subcortical areas related to
cortical swelling and loss of differentiation
between the gray and white matter contrast.
These findings are more evident in watershed
zones but occasionally can be appreciated in the
hemispheres [3]. Deep gray matter structures will
be most likely spared in these patients. In the
chronic stage, there are signs of atrophy with
loss of volume of the white matter and cortical
thinning predominantly in the parasagittal water-
shed zones (Table 3).
22 Hypoxic–Ischemic Encephalopathy (Preterm, Term, and Adult)
Table 3 HII in term neonates
Severe Injury of the deep gray nuclei
(putamina, ventrolateral thalami),
hippocampi, dorsal brainstem,
and lateral geniculate nuclei
Occasionally, the peri-rolandic
cortex
Partial or less Cortical watershed zones
severe asphyxia
HII in Postnatal Infants and Young
Children
In this group of patients, the most common causes
of HII are accidents such as drowning or choking
and non-accidental trauma. There are differences
between the pattern of injury seen in neonates and
that in infants and young children. These differ-
ences occur mostly because of disparities in the
brain maturation that advances rapidly during the
perinatal period. It has been estimated that the
myelination process is completed by about
2 years of age, and around this period, the pattern
of injury starts to resemble the one encountered in
adults.
Severe Asphyxia
Severe episodes of asphyxia in infants between
1 and 2 years of age result in injuries to the
caudate nuclei, putamina, lateral geniculate
nuclei, hippocampi, and cerebral cortex. The ante-
rior frontal and parieto-occipital cortex will be
most affected with relative sparing of the peri-
rolandic cortex and thalami [36].
In patients who experience asphyxia after the
immediate perinatal period, but before 1 year of
age, findings are often a mixture between those of
neonatal asphyxia and later infantile asphyxia and
result in the involvement of the basal ganglia in
particular the posterior putamina and lateral thal-
ami and also in involvement of the dorsal mid-
brain and cortex. The reasons for these differences
between neonates and older children are not
entirely understood, but it has been suggested
that the relative sparing of the thalami in older
483
Fig. 15 NECT of severe diffuse anoxic injury shows the
“reversal sign.” Note the subtle decreased density of the
cortical gray matter and basal ganglia relative to the denser
white matter
infants may be due to redistribution of blood flow
from the anterior circulation to the posterior cir-
culation after asphyxia [36]. Also, some differ-
ences may be due to physiologic and
biochemical changes that occur with maturation;
this results in changes in regional energy require-
ments and thus also in changes in the regional
susceptibility to hypoxia. By the time the anterior
fontanelle closes (at approximately 4 months), US
cannot be used anymore and CT becomes the
study of choice. CT examinations that are done
too early, before 24 h, can show only subtle
hypodensity in the deep gray matter structures or
be negative. Later, CT demonstrates diffuse basal
ganglia abnormalities along with diffuse cortical
hypoattenuation with loss of gray-white matter
differentiation and sulcal and cisternal efface-
ment, all of which are consequences of cerebral
edema. The peri-rolandic cortex may be relatively
spared [13, 44, 45]. At 4–6 days, hemorrhagic
infarcts may be evident in the basal ganglia. In
some patients, the “reversal sign” can be seen
(Fig. 15). The reversal sign refers to a reversal in
the normal CT attenuation patterns between gray
and white matter probably due to the congestion
of deep medullary veins secondary to obstruction
of venous outflow by cerebral edema and
484
Fig. 16 NECT demonstrating the “white cerebellum”
sign in a child with severe HII. Note the diffuse
hypodensity of the cerebral parenchyma that makes the
cerebellum look denser
subsequent compression of them. The other sign
in CT studies is the “white cerebellum sign”
(Fig. 16), in which the cerebral hemispheres are
hypodense due to diffuse edema, making the cer-
ebellum and brainstem appear relatively
hyperdense. This finding may be related to blood
flow redistribution to the posterior circulation that
occurs during anoxia [44]. Both the “reversal
sign” and “white cerebellum sign” are associated
with a worst outcome [44–46].
In MRI studies, the abnormalities on DWI are
evident in the first 12–24 h. First, images show
high intensity in the posterolateral lentiform
nuclei, and if the thalami are involved, usually
the ventrolateral nuclei will be most affected. In
the next 48 h, there is progression and involve-
ment of the rest of the basal ganglia and cortex.
Conventional T1- and T2-weighted images are
usually normal during the first day and may
remain normal for 48 h [43]. After this,
T2-weighted images show diffuse basal ganglia
and cortical hyperintensities with a relative spar-
ing of the thalami and peri-rolandic cortex
(Fig. 17). These T2 hyperintensities are believed
to represent edema [3]. The differential diagnosis
M. Castillo and F. Chiang
Fig. 17 FLAIR image demonstrates affectation of the
basal ganglia with sparing of the thalami. The occipital
cortex is also slightly hyperintense
Table 4 HII in postnatal infants and young children
Severe 1–2 years: injury to the caudate nuclei,
putamina, lateral geniculate nuclei,
hippocampi, and cerebral cortex
(especially anterior frontal and parieto-
occipital)
<1 year: mixture of the features of
neonatal asphyxia and later infantile
asphyxia
Mild to Cortical watershed zones
moderate
of bilateral basal ganglia abnormalities is wide
and includes inborn errors of metabolism, hypo-
glycemia, osmotic myelinolysis, hemolytic ure-
mic syndrome, toxic exposure, and infectious
encephalitis.
Mild-to-Moderate Asphyxia
In mild hypoxic insults to older infants, watershed
zone abnormalities in the cortex and subcortical
white matter are seen. White matter lesions may
also be seen but are more common in younger
children (under 1 year of age) [7]. Relative spar-
ring of the periventricular white matter is common
[31] (Table 4).
22 Hypoxic–Ischemic Encephalopathy (Preterm, Term, and Adult)
HII in Older Children and Adults
In older children, just as in younger ones, the most
common causes of asphyxia are drowning and
choking accidents. In adults on the other hand,
the causes of HII are different. The leading causes
for hypoxic adult events are related to chronic
diseases as cerebrovascular disease and/or cardiac
arrest with secondary hypoxemia. In this group of
patients, mild-to-moderate injury will manifest as
watershed infarcts. More severe insults have del-
eterious effects in the cortical gray matter and
deep gray structures. The cortex is usually dif-
fusely affected predominantly in the peri-rolandic
and visual areas, and the cerebellum and hippo-
campi may also be affected. The pathophysiology
of this gray matter injury pattern is probably
related to the fact that the gray matter contains
most of the postsynaptic glutamate receptors and
therefore is most susceptible to the excitotoxic
effects of glutamate. Also, gray matter is more
metabolically active than white matter.
Cerebellar injury is common in older patients.
The reason for this is not completely understood,
but it is believed that the relative immaturity of the
Purkinje cells in neonates has a protective. These
cells are particularly prone to ischemia because
Fig. 18 NECT in severe HII injury after a cardiac arrest shows total loss of gray-white matter differentiation with
hypoattenuation of the cortex and basal ganglia. There is also effacement of the CSF-containing spaces
485
they are not able to generate energy during an
anoxic event. The Purkinje cells die along with
cells in the granular layer [13].
In older patients, the first imaging study
performed for a suspected brain anoxic injury is
commonly a CT. In CT, the most common find-
ings are diffuse hypoattenuation of the cortex and
basal ganglia, consistent with edema, effacement
of the CSF-containing spaces, and loss of gray-
white matter differentiation (Fig. 18). In older
patients, as in children, the “reversal sign” and
“white cerebellum” sign may be present and also
indicate a poor prognosis (Fig. 19) [3].
As in other age groups, DWI is the earliest to
show abnormalities, usually starting after a few
hours after the injury. During the first 24 h,
hyperintensity in DWI is evident in the basal
ganglia, cerebellar hemispheres (Fig. 20), and cere-
bral cortex (especially in peri-rolandic and occipital
areas), and the thalami, brains tem and basal
ganglia may also be involved (Fig. 21) [7, 47, 48].
Abnormalities on T1- and T2-weighted images
may be delayed in comparison with DWI. After
the first 24 h, T2 images begin to show abnormal-
ities that consist of hyperintensity and swelling of
gray matter structures that may persist until the
end of the second week (Figs. 22, 23 and 24).
486 M. Castillo and F. Chiang

DWI images usually pseudo-normalize by the end

of the first week. In the chronic stage, T2-weighted
images demonstrate some residual hyperintensity
in basal ganglia, and the T1 may show
hyperintensity in the affected cortex, representing
cortical laminar necrosis in addition to diffuse atro-
phy (Figs. 25 and 26) [48] (Table 5).

Delayed White Matter Injury

Delayed white matter injury, also known as

Fig. 19 NECT showing the “white cerebellum sign.” The
cerebellum appears “white” when compared with the
abnormal hypodense brain
postanoxic leukoencephalopathy, is a rare compli-
cation of global hypoxia and takes place weeks
after the injury. It is an uncommon syndrome
which may be seen in approximately 3 % of
carbon monoxide intoxications. Patients usually
present with a period of relative clinical stability

Fig. 20 (a–d) DWI in

severe HII shows diffuse
diffusion restriction of the
cerebellar parenchyma with
the corresponding ADC
map hypointensity
22 Hypoxic–Ischemic Encephalopathy (Preterm, Term, and Adult) 487

Fig. 21 (a–f) DWI showing diffuse diffusion restriction throughout the cortex, caudate nuclei, and left lentiform nucleus
in a severe HII

or improvement, after which they develop acute

Fig. 22 T2WI shows hyperintensity and swelling of the
parieto-occipital cortex in HII
neurologic deterioration 2–3 weeks after the ini-
tial hypoxic event. Symptoms include delirium,
personality changes, intellectual impairment,
movement disorders, and/or seizures [49, 50].
This condition has a relatively good prognosis
with approximately 75 % of patients showing
near-complete or complete recovery in 6–12
months. In the rest of patients, there maybe resid-
ual dementia or rarely the condition may progress
to paresis, coma, or even death [3]. It is important
to identify patients with this condition recogniz-
ing their clinical and imaging features to be able to
provide them time to recover without abrupt with-
drawal of care [51].
MR is the study of choice in suspected delayed
white matter injury. DWI can fail to demonstrate
abnormalities when performed immediately after
the insult, but when performed during the period
488
Fig. 23 (a, b) DWI in HII shows diffusion restriction in the cortex and basal ganglia, with the corresponding ADC map
hypointensity
Fig. 24 T2WI of the same patient shows diffuse cortical
and basal ganglia hyperintensity and swelling
of delayed neurologic decline demonstrate diffuse
confluent areas of restricted diffusion in the white
matter [3, 47–49, 52]. T2-weighted images also
show corresponding hyperintensities in these areas.
M. Castillo and F. Chiang
Although this syndrome is typically a delayed
process, it can present as a progressive leukoence-
phalopathy not separable from the primary injury
and without an intervening lucid period. In these
cases, the abnormalities seen in DWI appear as
early as 2 days. In other cases, clinical findings
may not be apparent and MRI may be the first to
detect white matter injury. Clinical improvement
appears to be associated with improvement of the
signal intensity abnormalities in the white matter;
however, some residual hyperintense areas may
persist beyond 18 months [3, 53].
The pathophysiology of this delayed injury is
not known. It is believed that the findings repre-
sent a process similar to Wallerian degeneration
although apoptotic cell death may also play a role
and help explain the delay in the presentation of
the disease [47].
Role of Proton MR Spectroscopy
in the Evaluation of HII
Magnetic resonance spectroscopy (MRS) is a
technique that allows the identification and quan-
tification of some metabolites in tissues. It differs
from the conventional MR imaging in that it pro-
vides physiological and pathophysiological data
22 Hypoxic–Ischemic Encephalopathy (Preterm, Term, and Adult) 489

Fig. 25 (a, b) FLAIR images in HII demonstrate hyperintensity in the temporoparietal cortex

Fig. 26 (a–c) T1WI show areas of hyperintense cortex representing cortical laminar necrosis

instead of anatomic information. MRS has
become a valuable tool in the identification of
HII, especially in the perinatal period. Together
with DWI, they are the most sensitive modalities
for detecting HII in the acute period [3, 17, 54].
Furthermore, MRS is very useful in the first 24 h,
being more sensitive than the conventional
MR sequences and DWI in this period. MRS
also has value in predicting the severity of injury
[10, 17, 42].
The main alteration seen in MRS when HII
ensues is elevation of lactate. Lactate is not nor-
mally present or seen only as a very small peak in
normal newborns and only the first few hours of
life [55]. In HII, MRS shows increased lactate
levels which appear as a doublet at 1.3 ppm at
1.5 T in the deep gray matter, parieto-occipital
regions, or watershed zones by 2–8 h [17, 54,
56]. A glutamine–glutamate peak may be detected
at 2.3 ppm [54, 55] and could reflect the glutamate
release that occurs in HII (Fig. 27). Evidence
suggests that the elevation of lactate takes place
in two different phases. A first elevation occurs
very early in the acute stage of the injury and is
probably due to hypoxemia and the secondary
anaerobic glycolysis that takes place. Next, it
returns to normal as perfusion is restored. Then,
a second peak occurs over the following 24–48 h
490
Table 5 HII in older children and adults
Severe The cortex usually diffusely affected,
deep gray structures (basal ganglia and
thalami), hippocampi, and cerebellum
Mild to Cortical watershed zones
moderate
which is believed to be caused by a process known
as “secondary energy failure” in which surviving
neurons suffer delayed energy depletion most
probably due to mitochondrial failure. Injury
caused by this mechanism results in lactate eleva-
tion after 24 h and is associated with poor prog-
nosis [5, 6]; this partially explains the
underestimation of injury seen in DWI during
the first hours.
To correctly interpret the MRS studies in pre-
term and term infants, it is important to realize that
the spectra of premature babies differ significantly
from the one seen in term infants and adults. This
is the result of the differences in the concentration
of metabolites in the developing brain which does
not become similar to the adult brain until 2 years
of age [55]. In newborns, it is normal to have a
lower N-acetylaspartate (NAA) and higher myo-
inositol (Myo) and choline (Cho) levels compared
with older children and adults. In preterm babies,
a small peak of lactate is present and makes the
interpretation of MRS in suspected HII more dif-
ficult. Lactate diminishes and NAA increases as
the brain matures, but trace amounts of lactate
may be present even in term infants
[3, 55–57]. Because of this, it is very useful to
take into consideration the gestational age of the
infant at birth when interpreting the MRS to avoid
false-positive interpretations.
One useful tool is using the lactate–NAA ratio
in preterm neonates. The normal ratio in the thal-
ami for normal control neonates (term and pre-
term) is around 0.25 [56]. In preterm and term
neonates with HII, lactate–NAA ratios are greater
than 0.4, while in more severe injury, ratios are
greater than 0.5. When ratios are above the 95 %
confidence limits, there is a clear association with
major impairments or death at 1 year [56]. NAA is
usually normal in acute HII. If it descends in the
subacute phase, it is also associated with a poor
neurologic outcome [5, 58]. In older children, the
M. Castillo and F. Chiang
findings at MRS in HII are reduced NAA, ele-
vated lactate, and elevated glutamine–glutamate.
These alterations in the spectra correlate with a
bad prognosis including persistent vegetative
state or death [27]. Although HII cannot be
completely ruled out with normal examinations
until day 3–4, a normal MRS study and no abnor-
mal findings in conventional MRI sequences cor-
relate with a good outcome [6].
Imaging Choices in HII
As HII characteristics are variable, highly depen-
dent on the time since onset and patient’s age
(CNS maturity), many factors must be taken into
account when choosing the type of imaging study.
Some of the most important considerations are the
patient’s condition, age, availability of imaging
modalities, and concerns about ionizing radiation
exposure.
Many neonates with suspected HII are in inten-
sive care units and are hemodynamically unstable,
making the transportation difficult. In this regard,
performing cranial US may be best, taking in
consideration that it is fast and easy to perform
in a bedside setting, does not involve ionizing
radiation, and may provide useful information.
The major limitations of US are its operator
dependence and its low sensitivity (especially in
some areas such as the convexities). When US
demonstrates HII, it is very helpful but if it is
negative, another study is required.
In neonates, the next study should be MRI. CT
is not recommended in these patients because it
involves the use of ionizing radiation and is no
more sensitive than US due to the high water
content of the neonatal brain [3]. The MR protocol
should include at least DWI, ADC map, and
T1–T2-weighted images. MR spectroscopy can
be performed in the acute stage, especially when
DWI is negative and a clinical suspicion for HII is
high. In these situations, long echo time
(135 msec) and short echo time (30 msec) MRS
with voxels positioned in the basal ganglia and
centra semiovale are most useful. If the study is
negative in the first 24 h, it should be repeated
within 2–4 days. If subtle abnormalities are
22 Hypoxic–Ischemic Encephalopathy (Preterm, Term, and Adult) 491

Fig. 27 (a, b) MR I : Integral

spectroscopy in a term
neonate with HII. Short Cho
(top) and long (bottom) Ins dd1 I : 93.9
echo times show lactate a I : 72.5
doublet centered at 1.3 ppm
20
in the short echo time and an NAA
inverted lactate doublet in I : 140
the long echo time. There is
also elevated Cr
glutamine–glutamate (red 15 I : 59.0
arrow) in the short echo Cr2
I : 74.2
time and reduced NAA in
the long echo time
10 Lac
I : 72.4

ppm
4 3 2 1

I : Integral
Cho
b
I : 49.2
6

Cr
I : 32.3 NAA
4
I : 29.4

Lac
I : 11.5

ppm
3 2 1

present on MRI in the acute setting, follow-up In adults and infants with closed anterior fon-
imaging at the end of the first week is tanelles, unenhanced head CT is the initial exam-
recommended to define progression and overall ination of choice. If CT is positive, performing an
extent of injury [6]. MRI to assess the extent of the injury can be
492
considered. Currently, there is no clear role for
MRI perfusion imaging in this clinical setting.
Summary
HII is a highly variable entity that can pose sig-
nificant difficulties to the diagnosis from a clinical
and neuroimaging standpoint especially in
newborns.
Different patterns are observed in these
patients and depend mainly on brain maturity,
severity and length of the injury, and type and
timing of the imaging studies. A clear understand-
ing of the pathophysiological events related to the
global hypoxic–ischemic injury may help in
understanding its different imaging patterns. The
concept of selective vulnerability allows imagers
to focus their attention in specific areas that are
most likely injured. The most sensitive imaging
modalities in the detection of HII in the
hyperacute stage are DWI and MR spectroscopy.
DWI can underestimate the extent of injury in
very young children, so performing delayed imag-
ing is recommended to assess the extent of injuries
and shed light on the long-term prognosis of these
patients.
Key Points
– Imaging findings in HII are highly variable and
depend mainly on brain maturity, severity and
duration of the injury, and timing and imaging
modality used to study these patients.
– Conventional T1- and T2-weighted images are
most useful at the end of the first week when
DWI abnormalities have pseudo-normalized.
– DWI is useful in the first week. When used for
the evaluation of the neonatal brain, it should
be reviewed carefully in conjunction with
ADC maps because true restriction may be
masked by intrinsic T2 hyperintensity of the
unmyelinated brain at this age.
– In preterm neonates, profound ischemia pre-
sents with damage to the deep gray matter
and brainstem. Moderate-to-mild insults pre-
sent as germinal matrix hemorrhage, intraven-
tricular hemorrhage, and/or periventricular
white matter damage (PVL).
M. Castillo and F. Chiang
– MRS and DWI are the most sensitive modali-
ties for detecting HII in the acute period.
References
1. Dugan LL, Choi DW (1999). Hypoxia-ischemia and
brain infarction. In: Siegel GJ, Agranoff BW, Albers
RW et al (eds) Basic neurochemistry: molecular, cellu-
lar and medical aspects, 6th edn. Lippincott-Raven,
Philadelphia. Available from: http://www.ncbi.nlm.
nih.gov/books/NBK28046/
2. Armstrong-Wells J, Bernard TJ, Boada R, Manco-
Johnson M (2010) Neurocognitive outcomes following
neonatal encephalopathy. NeuroRehabilitation
26(1):27–33. doi:10.3233/NRE-2010-0533
3. Huang BY, Castillo M (2008) Hypoxic-ischemic brain
injury: imaging findings from birth to adulthood.
Radiographics 28(2):417–439. doi:10.1148/rg.2820
75066; quiz 617
4. Khurshid F, Lee KS, McNamara P, Whyte H, Mak W
(2011) Lessons learned during implementation of ther-
apeutic hypothermia for neonatal hypoxic ischemic
encephalopathy in a regional transport program in
Ontario. Paediatr Child Health 16(3):153–156
5. Barkovich AJ, Baranski K, Vigneron D et al (1999)
Proton MR spectroscopy for the evaluation of brain
injury in asphyxiated, term neonates. AJNR Am J
Neuroradiol 20:1399–1405
6. Grant PE, Yu D (2006) Acute injury to the immature
brain with hypoxia with or without hypoperfusion.
Radiol Clin North Am 44:63–77, viii
7. Christophe C, Fonteyne C, Ziereisen F et al (2002)
Value of MR imaging of the brain in children with
hypoxic coma. AJNR Am J Neuroradiol 23:716–723
8. Wijdicks EF, Campeau NG, Miller GM (2001) MR
imaging in comatose survivors of cardiac resuscitation.
AJNR Am J Neuroradiol 22:1561–1565
9. Johnston MV, Trescher WH, Ishida A, Nakajima W
(2001) Neurobiology of hypoxic-ischemic injury in the
developing brain. Pediatr Res 49(6):735–741
10. Barkovich AJ (2005) Brain and spine injuries in
infancy and childhood. In: Barkovich AJ (ed) Pediatric
neuroimaging, 4th edn. Lippincott Williams &
Wilkins, Philadelphia, pp 190–290
11. O’Shea TM (2002) Cerebral palsy in very preterm
infants: new epidemiological insights. Ment Retard
Dev Disabil Res Rev 8:135–145
12. Barkovich AJ, Sargent SK (1995) Profound asphyxia
in the premature infant: imaging findings. AJNR Am J
Neuroradiol 16:1837–1846
13. Castillo M (2007) Selective vulnerability and the cere-
bellum in neonates. AJNR Am J Neuroradiol 28:20–21
14. Chugani HT, Phelps ME, Mazziotta JC (1987) Positron
emission tomography study of human brain functional
development. Ann Neurol 22:487–497
15. Hasegawa M, Houdou S, Mito T, Takashima S,
Asanuma K, Ohno T (1992) Development of
22 Hypoxic–Ischemic Encephalopathy (Preterm, Term, and Adult)
myelination in the human fetal and infant cerebrum: a
myelin basic protein immunohistochemical study.
Brain Dev 14:16
16. Whyte HE, Blaser S (2013) Limitations of routine
neuroimaging in predicting outcomes of preterm
infants. Neuroradiology 55(Suppl 2):3–11.
doi:10.1007/s00234-013-1238-6
17. Barkovich AJ, Westmark K, Partridge C, Sola A,
Ferriero DM (1995) Perinatal asphyxia: MR findings
in the first 10 days. AJNR Am J Neuroradiol
16:427–438
18. Volpe JJ (1989) Intraventricular hemorrhage in the
premature infant: current concepts – I. Ann Neurol
25:3–11
19. Paneth N, Pinto-Martin J, Gardiner J et al (1993) Inci-
dence and timing of germinal matrix/intraventricular
hemorrhage in low birth weight infants. Am J
Epidemiol 137:1167–1176
20. Merrill JD, Piecuch RE, Fell SC, Barkovich AJ, Gold-
stein RB (1998) A new pattern of cerebellar hemor-
rhages in preterm infants. Pediatrics 102:E62
21. Correa F, Enríquez G, Rossello J, Lucaya J et al (2004)
Posterior Fontanelle Sonography: An Acoustic Win-
dow into the Neonatal Brain. AJNR Am J Neuroradiol
25:1274–1282
22. Ballabh P, Braun A, Nedergaard M (2004) Anatomic
analysis of blood vessels in germinal matrix, cerebral
cortex, and white matter in developing infants. Pediatr
Res 56:117–124
23. Back SA (2006) Perinatal white matter injury: the
changing spectrum of pathology and emerging insights
into pathogenetic mechanisms. Ment Retard Dev
Disabil Res Rev 12(2):129–140
24. Back SA, Luo NL, Borenstein NS, Levine JM, Volpe
JJ, Kinney HC (2001) Late oligodendrocyte progeni-
tors coincide with the developmental window of vul-
nerability for human perinatal white matter injury.
J Neurosci 21:1302–1312
25. Ferriero D, Miller S (2010) Imaging selective vulnera-
bility in the developing nervous system. J Anat
217:429–435. doi:10.1111/j.1469-7580.2010.01226.x
26. Flodmark O, Lupton B, Li D et al (1989) MR imaging
of periventricular leukomalacia in childhood. AJR Am
J Roentgenol 152:583–590
27. Dubowitz LM, Bydder GM, Mushin J (1985) Devel-
opmental sequence of periventricular leukomalacia:
correlation of US, clinical, and nuclear magnetic reso-
nance functions. Arch Dis Child 60:349–355
28. Inder TE, Anderson NJ, Spencer C, Wells S, Volpe JJ
(2003) White matter injury in the premature infant: a
comparison between serial cranial sonographic and
MR findings at term. AJNR Am J Neuroradiol
24:805–809
29. Murgo S, Avni EF, David P, Muller MF, Golzarian J,
Balériaux D, Struyven J (1999) Periventricular
leukomalacia in premature infants: prognostic role of
ultrasonography and MRI. J Radiol 80(7):715–720
30. Felderhoff-Mueser U, Rutherford MA, Squier WV
et al (1999) Relationship between MR imaging and
493
histopathologic findings of the brain in extremely sick
preterm infants. AJNR Am J Neuroradiol
20:1349–1357
31. Barkovich AJ, Truwit CL (1990) Brain damage from
perinatal asphyxia: correlation of MR findings with
gestational age. AJNR Am J Neuroradiol
11:1087–1096
32. Vannucci RC, Perlman JM (1997) Interventions for
perinatal hypoxic-ischemic encephalopathy. Pediatrics
100:1004–1014
33. Martinez-Biarge M, Diez-Sebastian J, Wusthoff CJ,
Mercuri E, Cowan FM (2013) Antepartum and
intrapartum factors preceding neonatal hypoxic-
ischemic encephalopathy. Pediatrics 132(4):
e952–e959. doi:10.1542/peds.2013-0511
34. Ferriero DM (2004) Neonatal brain injury. N Engl J
Med 351:1985–1995
35. Nelson KB (2002) The epidemiology of cerebral palsy
in term infants. Ment Retard Dev Disabil Res Rev
8:146–150
36. Barkovich AJ (1992) MR and CT evaluation of pro-
found neonatal and infantile asphyxia. AJNR Am J
Neuroradiol 13:959–972
37. Blankenberg F, Loh N, Bracci P, D’Arceuil H
et al (2000) Sonography, CT, and MR imaging: a
prospective comparison of neonates with suspected
intracranial ischemia and hemorrhage. AJNR Am J
Neuroradiol 21:213–218
38. Babcock DS, Ball W Jr (1983) Postasphyxial enceph-
alopathy in full-term infants: US diagnosis. Radiology
148:417–423
39. Hertzberg BS, Pasto ME, Needleman L, Kurtz AB,
Rifkin MD (1987) Postasphyxial encephalopathy in
term infants: sonographic demonstration of increased
echogenicity of the thalamus and basal ganglia. J Ultra-
sound Med 6:197–202
40. Connolly B, Kelehan P, O’Brien N et al (1994) The
echogenic thalamus in hypoxic ischaemic encephalop-
athy. Pediatr Radiol 24:268–271
41. Stark JE, Seibert JJ (1994) Cerebral artery Doppler
ultrasonography for prediction of outcome after peri-
natal asphyxia. J Ultrasound Med 13:595–600
42. Robertson RL, Ben-Sira L, Barnes PD et al (1999) MR
line-scan diffusion-weighted imaging of term neonates
with perinatal brain ischemia. AJNR Am J Neuroradiol
20:1658–1660
43. Dubowitz DJ, Bluml S, Arcinue E, Dietrich RB (1998)
MR of hypoxic encephalopathy in children after near
drowning: correlation with quantitative proton MR
spectroscopy and clinical outcome. AJNR Am J
Neuroradiol 19:1617–1627
44. Harwood-Nash DC (1992) Abuse to the pediatric cen-
tral nervous system. AJNR Am J Neuroradiol
13:569–575
45. Bird CR, Drayer BP, Gilles FH (1989) Pathophysiol-
ogy of “reverse” edema in global cerebral ischemia.
AJNR Am J Neuroradiol 10:95–98
46. Han BK, Towbin RB, De Courten-Myers G,
McLaurin R, Ball WS Jr (1989) Reversal sign on CT:
494
effect of anoxic/ischemic cerebral injury in children.
AJNR Am J Neuroradiol 10:1191–1198
47. Arbelaez A, Castillo M, Mukherji S (1999) Diffusion
weighted MR imaging of global cerebral anoxia. AJNR
Am J Neuroradiol 20:999–1007
48. Takahashi S, Higano S, Ishii K et al (1993) Hypoxic
brain damage: cortical laminar necrosis and delayed
changes in white matter at sequential MR imaging.
Radiology 189:449–456
49. Roychowdhury S, Maldjian JA, Galetta SL, Grossman
RI (1998) Postanoxic encephalopathy: diffusion MR
findings. J Comput Assist Tomogr 22:992–994
50. Bass E (1985) Cardiopulmonary arrest: pathophysiol-
ogy and neurologic complications. Ann Intern Med
103:920–927
51. Salazar R, Dubow J (2012) Delayed posthypoxic
leukoencephalopathy following a morphine overdose.
J Clin Neurosci 19(7):1060–1062. doi:10.1016/j.
jocn.2012.01.001
52. Kim JH, Chang KH, Song IC et al (2003) Delayed
encephalopathy of acute carbon monoxide intoxica-
tion: diffusivity of cerebral white matter lesions.
AJNR Am J Neuroradiol 24:1592–1597
53. Inagaki T, Ishino H, Seno H, Umegae N, Aoyama T
(1997) A long-term follow-up study of serial magnetic
M. Castillo and F. Chiang
resonance images in patients with delayed encephalop-
athy after acute carbon monoxide poisoning. Psychia-
try Clin Neurosci 51:421–423
54. Hanrahan JD, Sargentoni J, Azzopardi D et al (1996)
Cerebral metabolism within 18 hours of birth asphyxia:
a proton magnetic resonance spectroscopy study.
Pediatr Res 39:584–590
55. Bertholdo D, Watcharakorn A, Castillo M (2013) Brain
proton magnetic resonance spectroscopy: introduction
and overview. Neuroimaging Clin N Am
23(3):359–380. doi:10.1016/j.nic.2012.10.002
56. Penrice J, Cady EB, Lorek A et al (1996) Proton
magnetic resonance spectroscopy of the brain in nor-
mal preterm and term infants, and early changes after
perinatal hypoxia-ischemia. Pediatr Res 40:6–14
57. Leth H, Toft PB, Pryds O, Peitersen B, Lou HC,
Henriksen O (1995) Brain lactate in preterm
and growth retarded neonates. Acta Paediatr
84:495–499
58. Groenendaal F, Veenhoven RH, van der Grond J, Jan-
sen GH, Witkamp TD, De Vries LS (1994) Cerebral
lactate and N-acetyl-aspartate/choline ratios in asphyx-
iated full-term neonates demonstrated in vivo using
proton magnetic resonance spectroscopy. Pediatr Res
35:148–151
 Part V
Aneurysms
 Aneurysmal Subarachnoid
Hemorrhage 23
Frédéric Clarençon, Nader-Antoine Sourour, Vincent Degos,
Aurélien Nouet, Federico Di Maria, Eimad Shotar,
Joseph Gabrieli, Lise Le Jean, and Jacques Chiras

Contents Surgical Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 506

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498
Intracranial Aneurysms: Epidemiology . . . . . . . . . . 498
Clinical Symptoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499
Conservative Management . . . . . . . . . . . . . . . . . . . . . . . . . . 508
Complications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 508
SAH-Related Complications . . . . . . . . . . . . . . . . . . . . . . . . 508
Iatrogenic Complications . . . . . . . . . . . . . . . . . . . . . . . . . . . . 512

Diagnostic Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501 Prognostic Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513

Diagnosis of SAH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501 Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515
Imaging Modalities for the Depiction of IC
Aneurysms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 502 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 517
DSA-Negative SAH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 503 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 517
Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 504
General Considerations on Ruptured Intracranial
Aneurysms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 504
Endovascular Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 504

F. Clarençon (*) • N.-A. Sourour • F. Di Maria •

E. Shotar • J. Gabrieli • J. Chiras
Department of Interventional Neuroradiology,
Pitié-Salpêtrière Hospital, Paris VI University, Paris,
France
e-mail: fredclare5@msn.com; nsourour@gmail.com;
federico.dimaria@gmail.com; eimad.shotar@gmail.com;
wjosephw@libero.it; jacques.chiras@psl.aphp.fr
V. Degos • L. Le Jean
Department of Anesthesia and Critical Care, Neuro-
intensive Care Unit, Pitié- Salpêtrière Hospital, Paris VI
University, Paris, France
e-mail: degosv@gmail.com; lise.lejean@psl.aphp.fr
A. Nouet
Department of Neurosurgery, Pitié-Salpêtrière Hospital,
Paris VI University, Paris, France
e-mail: aurelien.nouet@psl.aphp.fr

# Springer Science+Business Media New York 2016 497

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_19
498 F. Clarençon et al.

Abstract
Intracranial Aneurysms: Epidemiology
Intracranial aneurysms are a non-exceptional
Intracranial (IC) aneurysms are non-exceptional
disease affecting 1–3 % of the general popula-
disease that, according to angiographic and
tion. The incidence of aneurysmal subarach-
autopsy studies, affects, in the Western countries,
noid hemorrhage (SAH) is about 1/10,000/
about 1–3 % of the population [1, 2]. A higher
year. Aneurysmal SAH is a devastating disease
prevalence of IC aneurysms has been reported in
leading to death approximately in 50 % of the
Norway and Japan [3, 4], suggesting genetic
cases and to neurological sequels in about 25 %
predisposing factors for this disease [5]. The IC
of the cases. Management of aneurysmal SAH
aneurysms’ rupture incidence has been evaluated
consists in early exclusion of the aneurysm’s
at 1/10 000/year [3, 6–9].
sac either by endovascular means or by micro-
The cumulative individual rupture risk per year
surgical clipping. Intensive care management
is difficult to calculate and has been estimated
is of tremendous importance to treat
around 0.5–1 %/year [4]. Most of the IC aneu-
SAH-related complications like hydrocephalus
rysms are located within the anterior circulation
and to prevent delayed complications such as
(85 %) [10]. Anterior communicating artery
vasospasm. Current perspectives for intracra-
(AComm) aneurysms account for 30 % of IC
nial aneurysms management consist in a better
aneurysms, posterior communicating artery
selection of the unruptured aneurysms to treat,
(PComm) aneurysms 25 %, middle cerebral artery
in order to propose a preventive treatment of
(MCA) aneurysms 20 %, internal carotid artery
exclusion before they bleed.
(ICA) termination 7.5 %, pericallosal artery
(PA) 4.5 %, basilar artery (BA) 7–10 %, and
Keywords
posterior inferior cerebellar artery 3–4 %
Subarachnoid hemorrhage • Intracranial aneu-
[11]. Saccular bifurcation aneurysms account for
rysm • Rupture • Coiling • Clipping • Hydro-
more than 90 % of IC aneurysms, the
cephalus • Arterial vasospasm
remaining being sidewall or dissecting aneu-
rysms [12]. Intracranial aneurysms have a female
predominance. Indeed, females have a relative
Introduction
risk of 1.24 for IC aneurysms, compared to
males [3]. Some genetic diseases may also be
Subarachnoid hemorrhage (SAH) consists in blood
associated with arterial wall abnormalities that
effusion in the subarachnoid space. The most fre-
may promote the formation of IC aneurysms.
quent cause of SAH is head trauma. In case of
Among them, the most frequently encountered is
spontaneous SAH, intracranial aneurysm is the
the polycystic kidney disease (PCKD) [13, 14].
leading cause and therefore should be carefully
In this specific population, the relative risk to
looked for with appropriate imaging modalities.
have intracranial aneurysm is 4.4-fold higher
Due to the high morbidity and mortality rate of
than in the general population [15]. Type IV
aneurysmal subarachnoid hemorrhage, early aneu-
Ehlers-Danlos disease (anomaly of type III colla-
rysm sac exclusion should be performed either by
gen) [16], Marfan disease (anomaly of fibrillin)
microsurgical technique or by endovascular means.
[17], and alpha-1 antitrypsin deficit [18] are other
Specific SAH-related complications that may occur
diseases of the elastic tissue that may be associ-
within the following days after the hemorrhage
ated with higher incidence of IC aneurysms.
require hospitalization in intensive care unit and
Familial forms of IC aneurysms are rare; com-
adapted management. We present an overview of
pared to incidental aneurysms, their size is fre-
the intracranial aneurysms’ epidemiology, of the
quently over 10 mm, and they are often multiple,
clinical and radiological signs suggesting an aneu-
located on the MCA, and prone to rupture earlier
rysmal SAH, and of the therapeutic management of
[19, 20].
this severe disease.
23 Aneurysmal Subarachnoid Hemorrhage
Clinical Symptoms
Most of the IC aneurysms are asymptomatic and
are discovered incidentally during imaging work-
up. The most severe and feared complication of IC
aneurysms is the rupture, leading to SAH
(Fig. 1a). Indeed, about 50 % of the patients who
Fig. 1 Different patterns of intracranial hemorrhage
related to aneurysmal rupture. (a) Typical subarachnoid
hemorrhage in a patient with an AComm aneurysm. Spon-
taneous hyperdensities are seen within the cortical sulci.
Note a thickener amount of blood in the interhemispheric
499
presented an aneurysmal SAH will die from the
rupture or from the rupture-related complications
[21]. Among the survivors, half will present
sequels from the SAH at long-term follow-up
[22]. Clinically, SAH is revealed by a typical
thunderclap headache and is associated with clin-
ical signs of meningeal irritation (neck stiffness,
fissure, close to the ruptured aneurysm (arrow). (b)
Intraparenchymal hemorrhage related to a right MCA
aneurysm rupture. (c) Acute subdural hematoma related
to a calcified left MCA aneurysm rupture
500 F. Clarençon et al.

photophobia, phonophobia). In case of massive sometimes be responsible for intraparenchymal

SAH, focal neurological deficit may be seen. hemorrhage [23] (Fig. 1c) or even subdural hema-
Finally, SAH may lead to comatose state (Tables 1 toma [24] (Fig. 1b). Rarely, the rupture may pre-
and 2). Although the great majority of IC aneu- sent as an acute and severe headache without
rysms’ rupture leads to SAH, this rupture may blood leakage within the subarachnoid space.
This entity corresponds to a “pre-rupture” status
and is described as a “cracking syndrome”
Table 1 World Federation of Neurological Surgeons
(WFNS) scale [25]. PComm aneurysms associated with sudden
headache and cranial nerve (CN) III palsy should
Grade Glasgow Coma Scale Motor deficit
be considered as ruptured aneurysms and treated
1 15 Absent
in emergency (Fig. 2).
2 13–14 Absent
3 13–14 Present
In case of large or giant aneurysms, especially
4 7–12 Present or absent on the carotid siphon, IC aneurysms may be
5 3–6 Present or absent revealed by progressive CN palsy (oculomotor
palsy or trigeminal neuralgia) [26]. These symp-
toms are related to the mass effect and the
edema that may compress/irritate the CNs of the
cavernous sinus (Fig. 3) [27]. ICA giant/large
Table 2 Hunt and Hess scale
aneurysms may also be revealed by epistaxis
I Asymptomatic or minimal headache and slight
when an erosion of the wall of the sphenoid
nuchal rigidity
II Moderate to severe headache, nuchal rigidity, no
sinus is present – the aneurysm presenting a risk
neurological deficit (except cranial nerve palsy) of rupture in the nasal cavities, leading to a poten-
III Drowsiness, confusion, or mild focal deficit tially cataclysmic and life-threatening
IV Stupor, moderate or severe hemiparesis, possible nosebleeding [28]. Finally, headache may reveal
early decerebrate rigidity, and vegetative intracranial aneurysms. However, the relationship
disturbance between the pain and the aneurysm is often diffi-
V Deep coma, decerebrate rigidity, moribund
cult to establish.

Fig. 2 A 59-year-old female presenting with severe head- after embolization showing a satisfactory occlusion of the
ache and acute CN III palsy. (a) Left ICA digital subtrac- aneurysm’s sac and the patency of the PComm. The CN III
tion angiography in lateral projection showing a 17 mm palsy completely recovered 5 weeks after the symptoms’
PComm aneurysm. (b) Control DSA in lateral projection onset
23 Aneurysmal Subarachnoid Hemorrhage
Fig. 3 Giant left cavernous ICA aneurysm revealed by a
CN VI palsy in a 68-year-old female. (a) and (b) photo-
graphs showing the left CN VI palsy: (a) right position
Diagnostic Imaging
Diagnosis of SAH
Unenhanced computed tomography (CT) is the
best imaging option at the acute phase of the
SAH. Indeed, the sensitivity of CT scan to detect
SAH is over 98 % within the first 48 h following
the hemorrhage [29]. It appears with a typical
pattern of spontaneous hyperdensity in the sub-
arachnoid space that insinuates inside the cortical
sulci. In case of massive hemorrhage, intraven-
tricular inundation may also be seen. It is note-
worthy that the thickest amount of blood may
orientate to the location of the aneurysm. Finally,
CT scan may also help to depict associated hydro-
cephalus. However, this sensitivity of CT for SAH
progressively decreases with time (58 % at 5 days,
50 % at 1 week) [30]. Brain magnetic resonance
imaging (MRI) may increase the sensitivity of CT
scan for the detection of SAH, especially few days
after the rupture.
MRI is rarely performed in suspected SAH.
Indeed, the accessibility of MRI is far lower than
501
gaze and (b) left position gaze. Abduction deficit of the left
eye is seen in left position gaze. (c) Left ICA DSA in AP
projection showing the giant cavernous ICA aneurysm
that of CT scan. The best sequence to depict
SAH is the T2*WI. SAH will appear as a
hypointense signal within the cortical sulci
(Fig. 4a). The interest of this sequence is also to
depict SAH at the late phase [31]. Fluid-
attenuated inversion recovery (FLAIR) sequence
is also helpful do depict SAH showing a
hyperintense signal within the cortical sulci
(Fig. 4b). However, even if FLAIR-WI may be
positive for SAH at the acute phase whereas CT
scan is negative, this sequence is not the panacea
for SAH diagnosis. Indeed, the FLAIR-WI will be
positive in case of negative CT scan in only less
than 20 % of the cases [32]. A recent study has
nonetheless showed that an MRI protocol com-
bining FLAIR-WI and susceptibility-WI (SWI)
had a high sensitivity for the depiction of
SAH [33].
Despite the high sensitivity of these various
imaging techniques, the gold standard exam for
the positive diagnosis of SAH remains the lumbar
puncture (LP) [34, 35].
Indeed, LP will find incoagulable blood in the
CSF and products of hemoglobin breakdown,
confirming the diagnosis of SAH [36].
502
Fig. 4 Brain MRI in a 69-year-old female presenting with
an aneurysmal subarachnoid hemorrhage. (a) Axial
T2*WI showing hypointense signal within the subarach-
noid space (black arrows) corresponding to subarachnoid
hemorrhage and hypointense signal in the occipital horn of
Finally, even if aneurysmal rupture is the most
frequent cause of spontaneous SAH, one should
keep in mind that the most frequent cause of SAH
is trauma. Other causes of SAH should not be
missed when the presence of an aneurysm has
been ruled out, like cerebral vein thrombosis,
amyloid angiopathy, or rupture of an intracranial
vascular malformation [37]. It is noteworthy that
in case of brain arteriovenous malformation
(bAVM) rupture, the SAH is often associated
with an intraparenchymal hemorrhage [38].
F. Clarençon et al.
the right lateral ventricle (white arrow head). (b) Axial
fluid-attenuated inversion recovery-WI showing
hyperintense signal in the subarachnoid space of both
sylvian fissures (black arrows). (c) MR angiography; 3D
TOF showing an AComm aneurysm (arrow)
Imaging Modalities for the Depiction
of IC Aneurysms
The aim of vascular diagnostic imaging is to
depict the IC aneurysm responsible for the SAH.
The presence of other IC aneurysms on the circle
of Willis should always be looked for, since IC
aneurysms are multiple in more than 25 % of the
cases. Additionally, cerebral vasospasm should be
assessed on vascular imaging, especially for
patients admitted few days after the SAH. Finally,
23 Aneurysmal Subarachnoid Hemorrhage
Fig. 5 Comparison of 3D time-of-flight (TOF) MR angi-
ography (a) and digital subtraction angiography (DSA) (b)
in left oblique anterior projection for visualization of a left
differential diagnosis, like brain AVM or cerebral
vein thrombosis, will be searched on vascular
imaging if no IC aneurysm is found.
CT angiography (CTA) is a valuable tool for the
detection of IC aneurysms [39–41]. This exam can
be performed just after the CT scan that authenti-
cated the SAH. It requires an iodinated contrast
media injection via a venous access (most of time
antebrachial). Its sensitivity for the detection of IC
aneurysms is about 95 % [39]. However, very
small aneurysms (<3 mm) and aneurysms close
to the skull base may be missed by CTA [42]. It is
noteworthy that the continuous improvement of
CT technologies increases the sensitivity of CTA
for the depiction of IC aneurysms [43]. The other
limitations for CTA are due to the fact that it
delivers X-rays, which limits its use for pregnant
women or children, and due to the potential risk of
allergy to the iodinated contrast media.
Three-dimensional (3D) time-of-flight (TOF)
magnetic resonance angiography (MRA) has a
lower spatial resolution than CTA. However, due to
the fact that it does not require contrast media injec-
tion, this sequence may be interesting for patients
with renal insufficiency or in pregnant patients [44].
Finally, digital subtraction angiography (DSA)
is the gold standard exam for the visualization of
IC aneurysms. Indeed, its spatial resolution is very
high: 200 μm. Moreover, 3D rotational
503
bifurcation MCA aneurysm. Note that on DSA, the small
branches close to the aneurysm are more precisely seen
than on 3D TOF MRA
angiography developed in the early 2000s, by
providing a manipulable volume, increased dra-
matically the understanding of IC aneurysms’
anatomy and is very helpful to tailor their treat-
ment [45]. Thus, DSA is the best imaging modal-
ity to depict very small aneurysms and to see
accurately and precisely small branches originat-
ing from the neck or from the parent artery close to
the aneurysm (Fig. 5). Except in case of extreme
emergency (e.g., ruptured MCA aneurysm with
clinically non-tolerated hematoma requiring sur-
gical evacuation), DSA should always be
performed before treatment (either endovascular
or surgical) of an IC aneurysm. Indeed, DSA helps
to clearly demonstrate the shape, size, neck, and
relationships of the aneurysm with the parent
artery. DSA should always be performed via the
four supra-aortic arterial vessels (i.e., both ICAs
and both vertebral arteries) in order not to miss
multiple aneurysms on the circle of Willis.
DSA-Negative SAH
DSA-negative SAHs are defined as proven SAHs
with negative initial DSA for the diagnosis of IC
aneurysm or other IC vascular malformations [46].
The frequency of such SAH is as high as about
20 % of the cases [47]. This subgroup of SAH is
504
divided into perimesencephalic and
non-perimesencephalic (i.e., diffuse, cortical, or
xanthochromia only) SAHs. In perimesencephalic
SAH overall diagnostic yield of repeat angiogra-
phy is 0 %, while it reaches 12.5 % in
non-perimesencephalic ones [48]. Even if
DSA-negative SAHs have a better prognosis than
aSAH, and despite the very low rate of rebleeding,
in such SAH delayed complications may be
observed like hydrocephalus or arterial
vasospasm [47].
Treatment
General Considerations on Ruptured
Intracranial Aneurysms
The first-line treatment for ruptured IC aneu-
rysms, since the publication of the ISAT study
[49], became the endovascular one. Indeed,
when eligible for both treatments, a patient with
ruptured aneurysm is less likely to have poor
clinical outcome (death or severe disability) at
1-year follow-up when treated by endovascular
means than by surgical clipping (23.7 % vs.
30.6 %). However, the treatment option is largely
dependent on the experience of the operator of
each specialty in a given center. Moreover, neuro-
surgical clipping has still some consensual
indications like very large-necked aneurysms
or aneurysms associated with clinically
non-tolerated parenchymal hematoma, for which
hematoma evacuation and aneurysm clipping will
be performed during the same treatment.
Endovascular Treatment
The aim of the endovascular treatment is to
“secure” the aneurysm sac by avoiding
blood circulation in the sac. The treatment
should be performed in emergency due to the
potential risk of acute rebleeding of the ruptured
IC aneurysm within the first days following the
SAH [50].
F. Clarençon et al.
Selective Coiling
Since they have been developed by Guglielmi
et al. in the early 1990s [51], detachable coils
have revolutionized the treatment of IC aneu-
rysms. Most of these coils are made of platinum,
and all of them are detachable. They are deployed
inside the aneurysm via a microcatheter posi-
tioned in the sac and detached from a stainless-
steel microguidewire by an electrical current
(Fig. 6). The objective is to fill the aneurysm by
means of decreasing size coils until total angio-
graphic occlusion, while preserving the patency of
the parent artery and other adjacent branches.
Even if safe and efficient for the treatment of
most IC aneurysms, large-necked aneurysms
may be difficult to treat by regular coiling due to
the risk of coil protrusion in the parent artery that
may be responsible for thromboembolic
complication.
“Remodeling” Technique
The remodeling technique has been developed in
the mid-1990s to overcome the limitation of the
presence of a large neck for the endovascular
treatment of IC aneurysms [52]. The principle of
this technique consists in inflating temporarily a
balloon in front of the aneurysm’s neck during the
coil deployment. This balloon helps to constrain
the positioning of the coil’s loops in the aneu-
rysm’s sac during its delivery (Fig. 7). With this
technique, almost all IC aneurysms are now eligi-
ble for endovascular treatment. Despite some con-
troversies [53], there are no evidences that the
balloon remodeling technique increases the
thromboembolic complication rate [54, 55].
Parent Artery Occlusion
Parent artery occlusion is a potential therapeutic
option for ruptured aneurysms, but can only be
performed in case of sufficient collaterals,
especially for large parent artery [26]. Moreover,
the potential risks of delayed vasospasm that
can lead to ischemic complications in patients
with ruptured IC aneurysms may prevent from
occluding a large vessel to preserve the circle of
Willis.
23 Aneurysmal Subarachnoid Hemorrhage
Fig. 6 Regular coiling in an AComm ruptured aneurysm.
(a) and (b) right ICA DSA: (a) 3D rotational angiography,
(b) working projection showing the saccular AComm
Treatment with Biological Glue
Biological glue (n-butyl-cyanoacrylate [NBCA])
may be used in ruptured IC aneurysms for
parent artery occlusion in dissecting aneurysms
(e.g., posterior inferior cerebellar artery [PICA]
dissecting aneurysm). The aim of this
treatment is to perform a segmental occlusion by
means of glue of the portion of the parent
505
aneurysm. (c) Unsubtracted snapshot during the aneurysm
sac coiling. (d) Post-coiling DSA in working projection
showing the complete occlusion of the aneurysm
artery harboring the dissecting aneurysm
[56]. The other main indication for the use
of glue in aSAH is the presence of
infectious IC aneurysms (aka mycotic aneu-
rysms). As for dissecting IC aneurysms, the
aim of the treatment is to obtain a segmental
occlusion of the vessel harboring the
aneurysm [57].
506
Fig. 7 Ruptured basilar tip aneurysm treated by
remodeling technique in a 49-year-old female. (a) Left
vertebral artery DSA showing the basilar tip aneurysm
with a relatively large neck (arrow). (b) 3D rotational
angiography reconstruction showing more precisely the
anatomy of the aneurysm. (c) Snapshot of the road map
Intracranial Stenting
The use of intracranial stents is not recommended
for the treatment of ruptured IC aneurysms at the
acute phase. Indeed, these devices require a dual
anti-aggregation platelet therapy that may be
responsible for hemorrhagic complications. How-
ever, in some selected cases, IC stents may be
used for the ruptured IC aneurysms. Indeed, in
case of coil protrusion in the parent artery, the
deployment of a stent in front of the aneurysm’s
neck may be on option to avoid thromboembolic
complication. This technique is called the “bailout
technique” [58]. In case of uncoilable blood
F. Clarençon et al.
during the coiling under balloon protection (remodeling
technique). Note the positioning of the balloon in front of
the aneurysm’s neck to constrain the coils in the aneu-
rysm’s sac (arrow). (d) Final DSA control showing the
total angiographic exclusion of the aneurysm
blister-like aneurysms, the use of flow diverter
stents may be a valuable option, either at the
acute phase or few days after the SAH
(Fig. 8) [59].
Surgical Treatment
The aim of the surgical treatment is the same as
that for the endovascular one: to avoid aneurysm
sac circulation. The main technique used is the
surgical clipping.
23 Aneurysmal Subarachnoid Hemorrhage
Fig. 8 Aneurysmal subarachnoid hemorrhage in a
39-year-old female. (a) Unenhanced CT scan demonstrat-
ing a prepontine spontaneous hyperdensity (arrow). (b)
DSA in AP projection showing a “blood blister-like” aneu-
rysm located on the left P1 segment (arrow). (c) 3D rota-
tional angiography, confirming the presence of a “blood
507
blister-like” aneurysm of the P1 segment of the left PCA
(arrow). (d) Flat-panel CT-like acquisition after the stent
deployment showing its satisfactory apposition on the par-
ent artery wall. (e) DSA performed at 6 months showing
the complete occlusion of the aneurysm covered by the
flow diverter stent
508
Walter Dandy first reported the surgical clip-
ping of an aneurysm in 1937.
After having performed a craniotomy and a
microsurgical dissection around the aneurysm’s
sac and the parent artery, this technique
consists in excluding the aneurysm from the intra-
cranial circulation by positioning a surgical clip at
the neck of the aneurysm (Fig. 9). In case of
complex shape aneurysms, few clips may be nec-
essary to exclude the aneurysm and reconstruct
the parent artery. Most of IC aneurysms can be
treated by surgical means. However, IC aneu-
rysms located in the posterior circulation are
more difficult to treat, leading to a higher compli-
cation rate [60].
In complex aneurysms, for which a clipping
seems too risky, or for blood blister-like aneu-
rysms, the “wrapping technique” may be an
option to prevent rebleeding. This technique con-
sists in wrapping the aneurysm and the parent
artery by surgical gauze or plastic sheeting in
order to prevent further rebleeding [61]. Finally,
bypass associated with parent artery sacrifice may
be an option for the treatment of some very com-
plex IC aneurysms, especially for giant/large
aneurysms of the MCA [62].
Conservative Management
Conservative management and imaging follow-
up may be a valuable option for unruptured intra-
cranial aneurysms, especially for small ones
[63]. In ruptured IC aneurysms, conservative
management should not be proposed, since
rebleeding rate is high. However, for untreatable
aneurysms in patients in poor clinical condition, a
“wait and see” strategy could be proposed, the
treatment being to be discussed few weeks later,
if the patient recovers from the aSAH and its
complications. It is noteworthy that, in aSAH,
the spontaneous rebleeding rate within 2 weeks
ranges from 15 % to 25 % and is about 50 %
within 6 months. Six months after the SAH, the
cumulative annual risk for rebleeding is 2–3 % per
year [11, 64–66].
F. Clarençon et al.
Complications
Complications that may be observed in aneurys-
mal SAH are divided into those related to the SAH
and those related to the treatment.
SAH-Related Complications
SAH-related complications may be severe and life
threatening. Their consequences should thus not
be underestimated. Among these complications,
hydrocephalus, arterial vasospasm, and neurolog-
ical/neuropsychological deficits may be seen at
the acute phase of the SAH or with a delay
[67]. Their management requires a close coordi-
nation between neurosurgical, interventional neu-
roradiology, and intensive care teams.
Spontaneous Rebleeding
Spontaneous rebleeding is the most severe com-
plication in aSAH. Indeed, the clinical conse-
quences of the rebleeding are frequently worse
than the ones from the first bleeding [68]. The
rebleeding is related to the lysis of the clot that
stopped the first hemorrhage. The early rebleeding
has been reported with a rate around 14 % [69]
and is higher within the first 6 h [68]. This high
rate of spontaneous rebleeding and the poor clin-
ical consequences of such rebleeding are the ratio-
nale for an early treatment of ruptured IC
aneurysms.
Hydrocephalus
Hydrocephalus may be observed at the acute
phase of the hemorrhage or can be delayed. The
mechanisms that may lead to hydrocephalus are
acute obstruction of the cerebrospinal fluid path-
way by a clot and abnormal blood resorption due
to the SAH. Hydrocephalus may occur in the
following weeks after the SAH. It is related to a
fibrosis of the subarachnoid spaces and an inflam-
matory proliferation of arachnoid villosities’ cells.
Such delayed hydrocephalus is more frequent in
patients in poor initial grade (WFNS). Clinically,
consciousness disorders and psychiatric disorders
23 Aneurysmal Subarachnoid Hemorrhage
Fig. 9 A 60-year-old female with a ruptured left MCA
bifurcation treated by surgical clipping. (a) Left ICA DSA
in working projection showing a small, large-necked,
MCA bifurcation aneurysm (arrow). (b) Photographs dur-
ing the surgical clipping. Exposition of the aneurysm
(arrow) by microsurgical dissection. (c) Photograph after
509
the positioning of the clip at the aneurysm’s neck. (d) DSA
in working projection after the clipping showing a satis-
factory exclusion of the aneurysm (arrow). (e) On
non-subtracted image, the clip can be seen at the aneu-
rysm’s neck (arrows)
510
Table 3 Fisher grading scale [99] evaluating the appear-
ance of SAH on CT scan
1 No hemorrhage evident
2 Subarachnoid hemorrhage less than 1 mm thick
3 Subarachnoid hemorrhage more than 1 mm thick
4 Subarachnoid hemorrhage of any thickness with
intraventricular hemorrhage (IVH) or parenchymal
extension
may be observed. Communicant hydrocephalus
may require a ventriculoatrial or ventriculo-
peritoneal shunting and can be temporarily treated
by serial lumbar punctures [70].
Arterial Vasospasm
Arterial vasospasm may occur few days after the
hemorrhage (> day 3) until day 21. Its frequency
and severity increase with the volume of blood in
the subarachnoid space (see Fisher score;
Table 3). Arterial vasospasm is seen on angio-
graphic exams in about 70 % of the cases in
aSAH. In 30 % of the cases, arterial vasospasm
may lead to ischemic lesions [71]. Thus, vaso-
spasm should be prevented by aggressive thera-
pies and treated, when present, to prevent
potential severe complications. The clear patho-
physiology of arterial vasospasm related to aSAH
remains unclear, some data suggesting an inflam-
matory vasculitis rather than a true
vasospasm [72].
The “H therapy” is one of the first-line thera-
pies in the armamentarium of the treatment of
aSAH-related arterial vasospasm. This therapy
associates hypertension while keeping
normovolemia and normalized cardiac flow.
Vasoconstrictor therapy with noradrenalin is
administered in an intensive care unit, with con-
tinuous monitoring comprising, at least, an inva-
sive measurement of the blood pressure. The
“triple-H therapy” that associates hemodilution,
hypervolemia, and arterial hypertension is no lon-
ger used to prevent arterial vasospasm because it
did not prove its efficacy for the treatment of
cerebral vasospasm. The use of a vasodilator
with cerebral tropism like nimodipine is system-
atically performed during the first 3 weeks after
the hemorrhage either per os or intravenously.
F. Clarençon et al.
Daily low doses of statins have proven their effec-
tiveness in preventing cerebral vasospasm
[73]. The diagnosis of vasospasm should be
suspected in case of fever, disorders of conscious-
ness, and motor or sensitive deficit or on ultraso-
nographic findings (accelerated mean velocities
on arterial vessels). Such signs should incite to
perform brain CT scan to rule out ischemic infarct,
hydrocephalus, parenchymal hematoma, or
rebleeding and a vascular imaging (CTA or
DSA). On vascular imaging, arterial vasospasm
typically appears as narrowing of the ICA termi-
nation, A1 and M1 segments (Fig. 10). Cortical
defect may also be seen on capillary phase on
DSA.
In case of symptomatic vasospasm, aggressive
treatment should be performed in emergency.
Intra-arterial chemodilatation with various drugs
(nimodipine [74], nicardipine [75], milrinone
[76]) has proven its effectiveness. However, the
effect of such intra-arterial chemodilatation is
only temporary (<24 h). The balloon angioplasty
is another option for the treatment of proximal
vasospasm; it consists in the inflation of a
nondetachable balloon at the level of the arterial
narrowing. This treatment is more effective and
definitive, but presents some risks of severe com-
plications (dissection, artery rupture, etc.) [77].
Epilepsy
Seizures may occur at the acute phase or lately.
Seizures are present at the acute phase from 6 % to
25 % [78] and may be related to hypoperfusion of
the brain parenchyma during the bleeding; they
are associated with a poor prognosis. Late seizures
may be related to rebleeding. Persistent seizures
after discharge are observed in about 5 % of the
cases [79]; independent predicting factors for such
seizures are the presence of a subdural hematoma
and ischemic infarct [79]. Finally, it is noteworthy
that the incidence of post-SAH epilepsy is higher
in patients treated by clipping than for those
treated by endovascular means [79].
Dysnatremia
The incidence of hyponatremia ranges from 30 %
to 40 %. Most of the time, this hyponatremia is
related to the so-called cerebral salt-wasting
23 Aneurysmal Subarachnoid Hemorrhage
Fig. 10 A 53-year-old male who presented an AComm
aneurysm rupture. Increasing of blood velocities on Dopp-
ler sonography at day 5. Left ICA DSA in anteroposterior
projection: arterial phase (a) and capillary phase (b).
Narrowing of A1 and M1 segments is seen (a, arrows)
consistent with a proximal arterial vasospasm. Note the
delayed opacification of the left ACA (a, arrow head). On
syndrome. This syndrome, first described in the
1950s, is characterized by an important loss of
water and sodium in the urine. This syndrome is
the consequence of a central hypersecretion of
511
capillary phase, extensive parenchymal defect is seen in the
left ACA and MCA territories, corresponding to a severe
distal vasospasm (b, arrow heads). (c) Unenhanced CT
scan performed the same day, showing multiple
vasospasm-related ischemic lesions in both hemispheres,
predominating in the left side
natriuretic peptide, analog of the atrial natriuretic
peptide. The cerebral salt-wasting syndrome
associates a depletive hyponatremia and a
hypovolemia. These conditions may increase the
512 F. Clarençon et al.

vulnerability of the brain for vasospasm. The

“syndrome of inappropriate antidiuretic hormone
secretion” (SIADH) is less frequent in SAH. This
syndrome associates hyponatremia and
hypervolemia. Hypernatremia, independently
from its mechanism, may worsen cerebral edema.
Finally, hypernatremia may be related to dys-
function of the hypothalamic-pituitary axis. Dia-
betes insipidus is characterized by a hypo-osmolar
polyuria (urine density close to the one of water),
hypernatremia, and hyponatriuresis, responsible
for intracellular dehydration and hypovolemic sta-
tus. Diabetes insipidus is treated by compensation
of hydric loss using hypotonic solute and by
antidiuretic hormone.

Cardiac Failure
The stimulation of the sympathetic nervous sys-
tem during the aneurysm’s rupture is responsible
for an “adrenergic storm” responsible for cardiac
disease [80]. Arrhythmia may be observed as well
as myocardial infarction and congestive heart fail-
ure. These cardiac complications may be life
threatening.
A massive release of catecholamines in the
myocardial wall, which may occur during the
aSAH, will result in myocardial stunning poten-
tially responsible for cardiac failure. This syn-
drome leading to a left ventricle insufficiency is
known as the “adrenergic myocarditis.” Clini-
cally, this syndrome presents as cardiogenic
shock. The duration of this cardiac failure is var-
iable: from few hours to 5 days. Its evolution is
most of the time favorable under tonicardiac treat-
ment, with ad integrum recovery of the cardiac
function.
Fig. 11 Brain CT scan in a 46-year-old female who expe-
rienced acute severe aneurysmal subarachnoid hemor-
rhage. Note the hyperdensity (arrow) in the posterior
compartment of the eye, corresponding to blood as part
of the so-called Terson syndrome
Terson Syndrome
The so-called Terson syndrome corresponds to
vitreous hemorrhage secondary to acute SAH
(Fig. 11) [82]. This complication may lead to
severe decrease of visual acuity. The mechanism
leading to this complication is still unclear. Sud-
den increase of intracranial pressure may hamper
the venous drainage leading to a rupture of the
hematoretinal barrier. In bilateral non-remitting
cases, vitrectomy can be proposed [83].

Neurological/Neuropsychological Iatrogenic Complications

Deficits
Neurological sequels may be seen secondary to
aSAH; they may be related to the hemorrhage
itself, to iatrogenic complications, or to ischemic
vasospasm. Neuropsychological deficits, like
memory loss or frontal syndrome, may be seen,
especially in ruptured aneurysms of the AComm
[81]. These neuropsychological disorders may be
responsible for severe handicap and may limit the
daily living and the return to work.
Perforation/Periprocedural Rebleeding
Perforation during the coiling is the most feared
complication during the treatment of ruptured IC
aneurysms by endovascular means. Indeed, this
hemorrhagic complication may lead to death
[84]. Rupture during endovascular treatment is
reported from 2 to 5 % [85–89]. The rupture
may be related to perforation with the
microguidewire or the microcatheter during the
23 Aneurysmal Subarachnoid Hemorrhage
microcatheterization of the aneurysms’ sac or dur-
ing the deployment of the coils’ loops (usually the
first one) (Fig. 12). This complication, if not ade-
quately and quickly managed, may be fatal, even
in the angiographic suite [90]. If a perforation
occurs during endovascular treatment, the coiling
must be performed or pursued in emergency to
stop the bleeding. It is noteworthy that this com-
plication is more frequent in the endovascular
coiling of ruptured IC aneurysms than in
unruptured ones. The mortality rate related to
aneurysm perforation is relatively high
(33 %) [91].
During surgical clipping, rebleeding may also
occur, either during the microsurgical dissection
or during the positioning of the clip. The fre-
quency of such rebleeding during surgery is esti-
mated for 10–20 %. Ruptured aneurysms are more
likely to rupture during the surgical clipping than
unruptured ones [92]. Even if the clinical conse-
quences of intraoperative rupture in surgical clip-
ping are less pejorative than in endovascular
coiling, the clinical outcome of patients who expe-
rienced a rupture during the clipping is worse than
the other ones [86]. PCA, AComm and PComm
aneurysms were more liable to rupture
intraoperatively. The occurrence of intraoperative
rupture for early surgery is not significantly higher
than for surgery performed more than 3 days after
subarachnoid hemorrhage [92]. To avoid such
complication, a temporary occlusion of the parent
vessel may be used [92].
Thromboembolic/Ischemic
Complications
Symptomatic thromboembolic complication may
occur in 5–10 % of the cases in embolization of IC
aneurysms [93]. Even if small ischemic infarcts
are relatively frequently seen on diffusion-
weighted images after endovascular treatment
(as frequently as in half of the cases [94]), most
of them remain asymptomatic, and among the
symptomatic ones, only a few will be responsible
for permanent deficit. Technical difficulties that
are associated with a higher risk of thromboem-
bolic complications are microcatheter
repositioning, coil removal and repositioning,
and size of the aneurysmal neck [55]. Most of
513
these complications are easily managed during
the procedure by dropping the blood pressure or
by IVaspirin injection (Fig. 13). Sometimes, more
aggressive managements are required like IA
injection of GPIIb/IIIa inhibitors [95], mechanical
thrombectomy [96], or even “bailout” stenting
[58]. Rarely, delayed (within the few hours after
the coiling) thromboembolic complications may
occur. Such complication may be observed when
the contact surface between the coils’ cast and the
parent artery is large.
Ischemic complications during the surgical
clipping may be related to the inopportune clip-
ping of a normal branch close to the aneurysm’s
neck, to hypoperfusion in the territory of an arte-
rial branch narrowed by a clip not satisfactorily
positioned, or to compression of surgical retrac-
tors on the brain parenchyma. Most of ischemic
complications are silent (i.e., non-symptomatic):
8 %. Symptomatic ischemic complications related
to surgical clipping only account for about
2 % [97].
Prognostic Factors
Some factors may predict the clinical outcome of
patients who had aSAH or the occurrence of
delayed complications. Indeed, the World Feder-
ation of Neurological Surgeons (WFNS) scale
(Table 1) and the Hunt and Hess score (Table 2)
are predictive, when high (4–5) for poor clinical
outcome [98].
The Fisher score is based on the evaluation of
the SAH-related blood burden on initial CT scan
(Table 3). When the Fisher score is high, the risk
and potential severity of delayed cerebral vaso-
spasm increase [99]. Age is a controversial prog-
nostic factor for clinical outcome. Indeed, even if
elderly patients (70 years) present comorbidities
that may hamper the clinical outcome, 37 % of
them will have a good clinical outcome at long-
term follow-up after endovascular coiling of a
ruptured aneurysm, independently from the initial
WFNS score at admission [100]. However, hydro-
cephalus in elderly patients may be associated
with a worse clinical outcome [101]. Finally,
some scores combining clinical and biomarker
514
Fig. 12 Example of perforation during the coiling in a
ruptured left PComm aneurysm. (a) Left ICA DSA in
working projection showing a 4 mm saccular PComm
aneurysm (arrow). Also note the presence of a second
small anterior choroidal artery aneurysm (arrow head).
(b). DSA after the deployment of the first coil loop show-
ing an acute rebleeding (arrows). The heparin was quickly
reversed and the coiling pursued while the balloon was
F. Clarençon et al.
kept inflated in front of the aneurysm. (c) DSA performed
after the coiling showing the complete occlusion of the
PComm aneurysm and the stop of the bleeding. (d)
Unenhanced CT scan performed after the procedure show-
ing a diffuse SAH and contrast media extravasation. (e)
Unenhanced CT scan performed at day 10 showing a total
regression of the SAH. The patient eventually discharged
3 weeks after the hemorrhage without any deficit
23 Aneurysmal Subarachnoid Hemorrhage
data, like the ABC score [102], have been pro-
posed to predict the 1-year clinical outcome in
patients with coiled ruptured aneurysms. Using
the ABC score, high GCS and elevated serum
levels of S100β protein (a marker of glial destruc-
tion) and troponin I were predictive of 1-year
mortality in this cohort.
Fig. 13 (continued)
515
Perspectives
Even if the management of ruptured intracranial
aneurysms has dramatically improved during the
past decades, the major challenge for the future in
the treatment of IC aneurysms will be to treat them
before they bleed and thus to accurately select
516
Fig. 13 Example of thromboembolic complication during
the coiling of a ruptured basilar tip aneurysm. (a) DSA in
working projection showing the basilar tip aneurysm with
a large neck. (b) 3D rotational angiography showing more
precisely the anatomy of the aneurysm. (c) Snapshot of the
road map in anteroposterior projection during the treatment
of the aneurysm. The microcatheter tip is positioned in the
aneurysm’s sac (black arrow) and the remodeling balloon
in front of the neck (white arrow). (d) Snapshot of the road
map in lateral projection during the coil deployment (black
those which are prone to rupture. At this
moment, trials designed to compare the
endovascular treatment vs. conservative manage-
ment for unruptured aneurysms have failed,
mostly due to the difficulty to include patients in
these studies [103].
F. Clarençon et al.
arrow) under balloon inflation (white arrow). (e) DSA in
working projection at the end of the coiling showing clot
formation at the origin of the P1 segment of the left PCA
(white arrow) and of the left superior cerebellar artery
(black arrow). (f) After increasing of the blood pressure
and IV aspirin injection, the control DSA showed a nearly
complete disappearance of the clots. (g) Unenhanced CT
scan performed at D7 showing no infarct in the left PCA
territory
Although some rupture risk factors have been
recognized and are used in daily practice to dis-
cuss the indications for the treatment of
unruptured IC aneurysms [104], no absolute
markers for further rupture of IC aneurysms
have been identified. Some tools are currently
23 Aneurysmal Subarachnoid Hemorrhage
under development to individualize, among all the
IC aneurysms, those that are prone to rupture in
their natural history. The most promising of these
tools is the computational fluid dynamics (CFD)
technologies [105]. CFD is a computational sim-
ulation of the flow in the parent artery and the
aneurysm sac. Many parameters can be obtained
from this simulation like the wall shear stress,
blood velocity, and impingement patterns, some
of them being possibly predictive for aneurysm
rupture [106].
Summary
Despite the advances in the management of aneu-
rysmal subarachnoid hemorrhage, this disease
remains associated with a poor prognosis. The
management of ruptured IC aneurysms can only
be multidisciplinary, with trained diagnostic neu-
roradiologists, experienced interventional neuro-
radiologists and neurosurgeons, and intensivists
specialized in neurointensive care.
References
1. Johnston SC, Higashida RT, Barrow DL et al (2002)
Recommendations for the endovascular treatment of
intracranial aneurysms: a statement for healthcare
professionals from the Committee on Cerebrovascu-
lar Imaging of the American Heart Association Coun-
cil on Cardiovascular Radiology. Stroke J Cereb Circ
33:2536–2544
2. Bardach NS, Zhao S, Gress DR et al (2002) Associ-
ation between subarachnoid hemorrhage outcomes
and number of cases treated at California hospitals.
Stroke J Cereb Circ 33:1851–1856
3. de Rooij NK, Linn FH, van der Plas JA et al (2007)
Incidence of subarachnoid haemorrhage: a systematic
review with emphasis on region, age, gender and time
trends. J Neurol Neurosurg Psychiatry 78:1365–1372
4. Muller TB, Sandvei MS, Kvistad KA et al (2013)
Unruptured intracranial aneurysms in the Norwegian
Nord-Trondelag Health Study (HUNT): risk of rup-
ture calculated from data in a population-based cohort
study. Neurosurgery 73:256–261, discussion 260;
quiz 261
5. Caranci F, Briganti F, Cirillo L et al (2013) Epidemi-
ology and genetics of intracranial aneurysms. Eur J
Radiol 82:1598–1605
6. Lafuente J, Maurice-Williams RS (2003) Ruptured
intracranial aneurysms: the outcome of surgical
517
treatment in experienced hands in the period prior to
the advent of endovascular coiling. J Neurol
Neurosurg Psychiatry 74:1680–1684
7. Fraser JF, Riina H, Mitra N et al (2006) Treatment of
ruptured intracranial aneurysms: looking to the past to
register the future. Neurosurgery 59:1157–1166, dis-
cussion 1166–1157
8. Wermer MJ, van der Schaaf IC, Algra A et al (2007)
Risk of rupture of unruptured intracranial aneurysms
in relation to patient and aneurysm characteristics: an
updated meta-analysis. Stroke J Cereb Circ
38:1404–1410
9. (1998) Unruptured intracranial aneurysms–risk of
rupture and risks of surgical intervention. Interna-
tional study of unruptured intracranial aneurysms
investigators. N Engl J Med 339:1725–1733
10. Bonneville F, Sourour N, Biondi A (2006) Intracra-
nial aneurysms: an overview. Neuroimaging Clin N
Am 16:371–382
11. Brisman JL, Song JK, Newell DW (2006) Cerebral
aneurysms. N Engl J Med 355:928–939
12. Mizutani T, Miki Y, Kojima H et al (1999) Proposed
classification of nonatherosclerotic cerebral fusiform
and dissecting aneurysms. Neurosurgery 45:253–259,
discussion 259–260
13. Niemczyk M, Gradzik M, Niemczyk S et al (2013)
Intracranial aneurysms in autosomal dominant poly-
cystic kidney disease. AJNR Am J Neuroradiol
34:1556–1559
14. Rozenfeld MN, Ansari SA, Shaibani A et al (2014)
Should patients with autosomal dominant polycystic
kidney disease be screened for cerebral aneurysms?
AJNR Am J Neuroradiol 35:3–9
15. Rinkel GJ, Djibuti M, Algra A et al (1998) Prevalence
and risk of rupture of intracranial aneurysms: a sys-
tematic review. Stroke J Cereb Circ 29:251–256
16. Schievink WI (2004) Cerebrovascular involvement in
Ehlers-Danlos syndrome. Curr Treat Options
Cardiovasc Med 6:231–236
17. Schievink WI (1999) Marfan syndrome and intracra-
nial aneurysms. Stroke J Cereb Circ 30:2767–2768
18. Schievink WI, Katzmann JA, Piepgras DG
et al (1996) Alpha-1-antitrypsin phenotypes among
patients with intracranial aneurysms. J Neurosurg
84:781–784
19. Ruigrok YM, Rinkel GJ (2008) Genetics of intracra-
nial aneurysms. Stroke J Cereb Circ 39:1049–1055
20. Ruigrok YM, Rinkel GJ, Algra A et al (2004) Char-
acteristics of intracranial aneurysms in patients with
familial subarachnoid hemorrhage. Neurology
62:891–894
21. Wiebers DO, Whisnant JP, Huston J 3rd et al (2003)
Unruptured intracranial aneurysms: natural history,
clinical outcome, and risks of surgical and
endovascular treatment. Lancet 362:103–110
22. Nieuwkamp DJ, Setz LE, Algra A et al (2009)
Changes in case fatality of aneurysmal subarachnoid
haemorrhage over time, according to age, sex, and
region: a meta-analysis. Lancet Neurol 8:635–642
518
23. Stapleton CJ, Walcott BP, Fusco MR, Butler WE,
Thomas AJ, Ogilvy CS (2015) Surgical management
of ruptured middle cerebral artery aneurysms with
large intraparenchymal or sylvian fissure hematomas.
Neurosurgery. 76(3):258–64
24. Boujemaa H, Gongora-Rivera F, Barragan-Campos H
et al (2006) Bilateral acute subdural hematoma
from ruptured posterior communicating artery aneu-
rysm. A case report. Interv Neuroradiol J
Peritherapeutic Neuroradiol Surg Proc Relat Neurosci
12:37–40
25. Feely M, Kapoor S (1987) Third nerve palsy due to
posterior communicating artery aneurysm: the impor-
tance of early surgery. J Neurol Neurosurg Psychiatry
50:1051–1052
26. Clarencon F, Bonneville F, Boch AL et al (2011) Par-
ent artery occlusion is not obsolete in giant aneurysms
of the ICA. Experience with very-long-term follow-
up. Neuroradiology 53:973–982
27. Roccatagliata L, Guedin P, Condette-Auliac S
et al (2010) Partially thrombosed intracranial aneu-
rysms: symptoms, evolution, and therapeutic man-
agement. Acta Neurochir 152:2133–2142
28. Davies A, Dale O, Renowden S (2011) Spontaneous
rupture of an intra-cavernous internal carotid artery
aneurysm presenting with massive epistaxis.
J Laryngol Otol 125:1070–1072
29. Cortnum S, Sorensen P, Jorgensen J (2010) Determin-
ing the sensitivity of computed tomography scanning
in early detection of subarachnoid hemorrhage. Neu-
rosurgery 66:900–902, discussion 903
30. Latchaw RE, Silva P, Falcone SF (1997) The role of
CT following aneurysmal rupture. Neuroimaging
Clin N Am 7:693–708
31. Sohn CH, Baik SK, Lee HJ et al (2005) MR imaging
of hyperacute subarachnoid and intraventricular hem-
orrhage at 3T: a preliminary report of gradient echo
T2*-weighted sequences. AJNR Am J Neuroradiol
26:662–665
32. Mohamed M, Heasly DC, Yagmurlu B et al (2004)
Fluid-attenuated inversion recovery MR imaging and
subarachnoid hemorrhage: not a panacea. AJNR Am J
Neuroradiol 25:545–550
33. Verma RK, Kottke R, Andereggen L et al (2013)
Detecting subarachnoid hemorrhage: comparison of
combined FLAIR/SWI versus CT. Eur J Radiol
82:1539–1545
34. Sidman R, Connolly E, Lemke T (1996) Subarach-
noid hemorrhage diagnosis: lumbar puncture is still
needed when the computed tomography scan is nor-
mal. Acad Emerg Med 3:827–831
35. Davis SM, Donnan GA (2008) Thunderclap head-
ache: CT and lumbar puncture but occasionally
more! Stroke J Cereb Circ 39:1396
36. Cruickshank A, Auld P, Beetham R et al (2008)
Revised national guidelines for analysis of
cerebrospinal fluid for bilirubin in suspected sub-
arachnoid haemorrhage. Ann Clin Biochem
45:238–244
F. Clarençon et al.
37. Marder CP, Narla V, Fink JR et al (2014) Subarach-
noid hemorrhage: beyond aneurysms. AJR Am J
Roentgenol 202:25–37
38. Geibprasert S, Pongpech S, Jiarakongmun P
et al (2010) Radiologic assessment of brain arteriove-
nous malformations: what clinicians need to know.
Radiogr Rev Publ Radiol Soc North Am Inc
30:483–501
39. Li Q, Lv F, Yao G et al (2014) 64-section
multidetector CT angiography for evaluation of intra-
cranial aneurysms: comparison with 3D rotational
angiography. Acta Radiol 55(7):840–6
40. Taschner CA, Thines L, Lernout M et al (2007) Treat-
ment decision in ruptured intracranial aneurysms:
comparison between multi-detector row CT angiog-
raphy and digital subtraction angiography.
J Neuroradiol 34:243–249
41. Teksam M, McKinney A, Casey S et al (2004) Multi-
section CT angiography for detection of cerebral
aneurysms. AJNR Am J Neuroradiol 25:1485–1492
42. Zhang H, Hou C, Zhou Z et al (2012) Evaluating of
small intracranial aneurysms by 64-detector CT angi-
ography: a comparison with 3-dimensional rotation
DSA or surgical findings. J Neuroimag Off J Am Soc
Neuroimag
43. Luo Z, Wang D, Sun X et al (2012) Comparison of the
accuracy of subtraction CT angiography performed
on 320-detector row volume CT with conventional
CT angiography for diagnosis of intracranial aneu-
rysms. Eur J Radiol 81:118–122
44. Yanamadala V, Sheth SA, Walcott BP et al (2013)
Non-contrast 3D time-of-flight magnetic resonance
angiography for visualization of intracranial aneu-
rysms in patients with absolute contraindications to
CT or MRI contrast. J Clin Neurosci 20:1122–1126
45. Missler U, Hundt C, Wiesmann M et al (2000) Three-
dimensional reconstructed rotational digital subtrac-
tion angiography in planning treatment of intracranial
aneurysms. Eur Radiol 10:564–568
46. Hui FK, Tumialan LM, Tanaka T et al (2009) Clinical
differences between angiographically negative, dif-
fuse subarachnoid hemorrhage and perimesen-
cephalic subarachnoid hemorrhage. Neurocrit Care
11:64–70
47. Lin N, Zenonos G, Kim AH et al (2012) Angiogram-
negative subarachnoid hemorrhage: relationship
between bleeding pattern and clinical outcome.
Neurocrit Care 16:389–398
48. Dalyai R, Chalouhi N, Theofanis T et al (2013) Sub-
arachnoid hemorrhage with negative initial catheter
angiography: a review of 254 cases evaluating patient
clinical outcome and efficacy of short- and long-term
repeat angiography. Neurosurgery 72:646–652, dis-
cussion 651–642
49. Molyneux A, Kerr R, Stratton I et al (2002) Interna-
tional Subarachnoid Aneurysm Trial (ISAT) of neu-
rosurgical clipping versus endovascular coiling in
2143 patients with ruptured intracranial aneurysms:
a randomised trial. Lancet 360:1267–1274
23 Aneurysmal Subarachnoid Hemorrhage
50. Wong GK, Boet R, Ng SC et al (2012) Ultra-early
(within 24 hours) aneurysm treatment after subarach-
noid hemorrhage. World Neurosurg 77:311–315
51. Guglielmi G, Vinuela F, Dion J et al (1991)
Electrothrombosis of saccular aneurysms via
endovascular approach. Part 2: preliminary clinical
experience. J Neurosurg 75:8–14
52. Moret J, Cognard C, Weill A et al (1997) Reconstruc-
tion technic in the treatment of wide-neck intracranial
aneurysms. Long-term angiographic and clinical
results. Apropos of 56 cases. J Neuroradiol 24:30–44
53. van Rooij WJ, Sluzewski M, Beute GN et al (2006)
Procedural complications of coiling of ruptured intra-
cranial aneurysms: incidence and risk factors in a
consecutive series of 681 patients. AJNR Am J
Neuroradiol 27:1498–1501
54. Shapiro M, Babb J, Becske T et al (2008) Safety and
efficacy of adjunctive balloon remodeling during
endovascular treatment of intracranial aneurysms: a
literature review. AJNR Am J Neuroradiol
29:1777–1781
55. Albayram S, Selcuk H, Kara B et al (2004) Throm-
boembolic events associated with balloon-assisted
coil embolization: evaluation with diffusion-weighted
MR imaging. AJNR Am J Neuroradiol 25:1768–1777
56. Crowley RW, Albuquerque FC, Ducruet AF
et al (2012) Technical considerations in the
endovascular management of aneurysms of the pos-
terior inferior cerebellar artery. Neurosurgery
71:ons204–ons217, discussion ons217–ons208
57. Chapot R, Houdart E, Saint-Maurice JP et al (2002)
Endovascular treatment of cerebral mycotic aneu-
rysms. Radiology 222:389–396
58. Yoo E, Kim DJ, Kim DI et al (2009) Bailout stent
deployment during coil embolization of
intracranial aneurysms. AJNR Am J Neuroradiol
30:1028–1034
59. Gonzalez AM, Narata AP, Yilmaz H et al (2014)
Blood blister-like aneurysms: single center experi-
ence and systematic literature review. Eur J Radiol
83:197–205
60. Spetzler RF, McDougall CG, Albuquerque FC
et al (2013) The barrow ruptured aneurysm trial:
3-year results. J Neurosurg 119:146–157
61. Cudlip SA, Kitchen ND, McKhahn GM et al (1998)
Wrapping of solitary ruptured intracranial aneurysms,
outcome at five years. Acta Neurochir
140:1167–1170, discussion 1170–1161
62. Kivipelto L, Niemela M, Meling T et al (2014) Bypass
surgery for complex middle cerebral artery aneu-
rysms: impact of the exact location in the MCA tree.
J Neurosurg 120:398–408
63. Valenca MM (2013) “Sit back, observe, and wait.” Or
is there a pharmacologic preventive treatment for
cerebral aneurysms? Neurosurg Rev 36:1–9, discus-
sion 9–10
64. Edlow JA, Malek AM, Ogilvy CS (2008) Aneurysmal
subarachnoid hemorrhage: update for emergency
physicians. J Emerg Med 34:237–251
519
65. Rosengart AJ, Schultheiss KE, Tolentino J
et al (2007) Prognostic factors for outcome in patients
with aneurysmal subarachnoid hemorrhage. Stroke J
Cereb Circ 38:2315–2321
66. Feigin VL, Findlay M (2006) Advances in
subarachnoid hemorrhage. Stroke J Cereb Circ
37:305–308
67. Koenig MA (2012) Management of delayed cerebral
ischemia after subarachnoid hemorrhage. Continuum
18:579–597
68. Inagawa T, Kamiya K, Ogasawara H et al (1987)
Rebleeding of ruptured intracranial aneurysms in the
acute stage. Surg Neurol 28:93–99
69. Ohkuma H, Tsurutani H, Suzuki S (2001) Incidence
and significance of early aneurysmal rebleeding
before neurosurgical or neurological management.
Stroke J Cereb Circ 32:1176–1180
70. Hasan D, Lindsay KW, Vermeulen M (1991) Treat-
ment of acute hydrocephalus after subarachnoid hem-
orrhage with serial lumbar puncture. Stroke J Cereb
Circ 22:190–194
71. Dorsch NW (1995) Cerebral arterial spasm–a clinical
review. Br J Neurosurg 9:403–412
72. Dumont AS, Dumont RJ, Chow MM et al (2003)
Cerebral vasospasm after subarachnoid hemorrhage:
putative role of inflammation. Neurosurgery
53:123–133, discussion 133–125
73. Sanchez-Pena P, Nouet A, Clarencon F et al (2012)
Atorvastatin decreases computed tomography and
S100-assessed brain ischemia after subarachnoid
aneurysmal hemorrhage: a comparative study. Crit
Care Med 40:594–602
74. Biondi A, Ricciardi GK, Puybasset L et al (2004)
Intra-arterial nimodipine for the treatment of symp-
tomatic cerebral vasospasm after aneurysmal sub-
arachnoid hemorrhage: preliminary results. AJNR
Am J Neuroradiol 25:1067–1076
75. Linfante I, Delgado-Mederos R, Andreone V
et al (2008) Angiographic and hemodynamic effect
of high concentration of intra-arterial nicardipine in
cerebral vasospasm. Neurosurgery 63:1080–1086,
discussion 1086–1087
76. Shankar JJ, dos Santos MP, Deus-Silva L et al (2011)
Angiographic evaluation of the effect of intra-arterial
milrinone therapy in patients with vasospasm from
aneurysmal subarachnoid hemorrhage. Neuroradiol-
ogy 53:123–128
77. Dorsch NW (2002) Therapeutic approaches to vaso-
spasm in subarachnoid hemorrhage. Curr Opin Crit
Care 8:128–133
78. Hart Y, Sneade M, Birks J et al (2011) Epilepsy after
subarachnoid hemorrhage: the frequency of seizures
after clip occlusion or coil embolization of a ruptured
cerebral aneurysm: results from the International Sub-
arachnoid Aneurysm Trial. J Neurosurg
115:1159–1168
79. Claassen J, Peery S, Kreiter KT et al (2003) Predictors
and clinical impact of epilepsy after subarachnoid
hemorrhage. Neurology 60:208–214
520
80. Ahmadian A, Mizzi A, Banasiak M et al (2013) Car-
diac manifestations of subarachnoid hemorrhage.
Heart Lung Vessels 5:168–178
81. Proust F, Martinaud O, Gerardin E et al (2009) Qual-
ity of life and brain damage after microsurgical clip
occlusion or endovascular coil embolization for rup-
tured anterior communicating artery aneurysms: neu-
ropsychological assessment. J Neurosurg 110:19–29
82. Terson A (1900) De l’hémorrhagie dans le corps vitre
au cours de l’hémorrhagie cerebrale. Clin Ophthalmol
6:309–312
83. Stienen MN, Lucke S, Gautschi OP et al (2012)
Terson haemorrhage in patients suffering aneurysmal
subarachnoid haemorrhage: a prospective analysis of
60 consecutive patients. Clin Neurol Neurosurg
114:535–538
84. Chen M (2013) A checklist for cerebral aneurysm
embolization complications. J Neurointerv Surg
5:20–27
85. Ricolfi F, Le Guerinel C, Blustajn J et al (1998) Rup-
ture during treatment of recently ruptured aneurysms
with Guglielmi electrodetachable coils. AJNR Am J
Neuroradiol 19:1653–1658
86. Elijovich L, Higashida RT, Lawton MT et al (2008)
Predictors and outcomes of intraprocedural rupture in
patients treated for ruptured intracranial aneurysms:
the CARAT study. Stroke J Cereb Circ 39:1501–1506
87. Vinuela F, Duckwiler G, Mawad M (1997) Guglielmi
detachable coil embolization of acute
intracranial aneurysm: perioperative anatomical and
clinical outcome in 403 patients. J Neurosurg
86:475–482
88. Cognard C, Weill A, Castaings L et al (1998) Intra-
cranial berry aneurysms: angiographic and clinical
results after endovascular treatment. Radiology
206:499–510
89. Doerfler A, Wanke I, Egelhof T et al (2001) Aneurys-
mal rupture during embolization with Guglielmi
detachable coils: causes, management, and outcome.
AJNR Am J Neuroradiol 22:1825–1832
90. Ryu CW, Lee CY, Koh JS et al (2011) Vascular
perforation during coil embolization of an intracranial
aneurysm: the incidence, mechanism, and clinical
outcome. Neurointervention 6:17–22
91. Cloft HJ, Kallmes DF (2002) Cerebral aneurysm per-
forations complicating therapy with Guglielmi
detachable coils: a meta-analysis. AJNR Am J
Neuroradiol 23:1706–1709
92. Leipzig TJ, Morgan J, Horner TG et al (2005) Anal-
ysis of intraoperative rupture in the surgical treatment
of 1694 saccular aneurysms. Neurosurgery
56:455–468, discussion 455–468
93. Brooks NP, Turk AS, Niemann DB et al (2008) Fre-
quency of thromboembolic events associated with
endovascular aneurysm treatment: retrospective case
series. J Neurosurg 108:1095–1100
F. Clarençon et al.
94. Altay T, Kang HI, Woo HH et al (2011) Thromboem-
bolic events associated with endovascular treatment
of cerebral aneurysms. J Neurointerv Surg 3:147–150
95. Aggour M, Pierot L, Kadziolka K et al (2010)
Abciximab treatment modalities for thromboembolic
events related to aneurysm coiling. Neurosurgery
67:503–508
96. Fourie P, Duncan IC (2003) Microsnare-assisted
mechanical removal of intraprocedural distal middle
cerebral arterial thromboembolism. AJNR Am J
Neuroradiol 24:630–632
97. Krayenbuhl N, Erdem E, Oinas M et al (2009) Symp-
tomatic and silent ischemia associated with microsur-
gical clipping of intracranial aneurysms: evaluation
with diffusion-weighted MRI. Stroke J Cereb Circ
40:129–133
98. Wang T, Zhang JH, Qin X (2011) Analysis on death-
associated factors of patients with subarachnoid hem-
orrhage during hospitalization. Acta Neurochir Suppl
110:219–223
99. Fisher CM, Kistler JP, Davis JM (1980) Relation of
cerebral vasospasm to subarachnoid hemorrhage
visualized by computerized tomographic scanning.
Neurosurgery 6:1–9
100. Iosif C, Di Maria F, Sourour N et al (2013) Is a high
initial World Federation of Neurosurgery (WFNS)
grade really associated with a poor clinical outcome
in elderly patients with ruptured intracranial aneu-
rysms treated with coiling? J Neurointerv Surg
101. Degos V, Gourraud PA, Tursis VT et al (2012) Elderly
age as a prognostic marker of 1-year poor outcome for
subarachnoid hemorrhage patients through its inter-
action with admission hydrocephalus. Anesthesiol-
ogy 117:1289–1299
102. Degos V, Apfel CC, Sanchez P et al (2012) An admis-
sion bioclinical score to predict 1-year outcomes in
patients undergoing aneurysm coiling. Stroke J Cereb
Circ 43:1253–1259
103. Raymond J, Molyneux AJ, Fox AJ et al (2008) The
TEAM trial: safety and efficacy of endovascular treat-
ment of unruptured intracranial aneurysms in the pre-
vention of aneurysmal hemorrhages: a randomized
comparison with indefinite deferral of treatment in
2002 patients followed for 10 years. Trials 9:43
104. Lall RR, Eddleman CS, Bendok BR et al (2009)
Unruptured intracranial aneurysms and the assess-
ment of rupture risk based on anatomical and mor-
phological factors: sifting through the sands of data.
Neurosurg Focus 26:E2
105. Cebral JR, Mut F, Weir J et al (2011) Quantitative
characterization of the hemodynamic environment in
ruptured and unruptured brain aneurysms. AJNR Am
J Neuroradiol 32:145–151
106. Cebral JR, Sheridan M, Putman CM (2010) Hemody-
namics and bleb formation in intracranial aneurysms.
AJNR Am J Neuroradiol 31:304–310
 Blister Aneurysms
24
James V. Byrne and Svein Harald Mørkve

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522 Blister aneurysms comprise about 1 % of rup-
tured intracranial aneurysms. They are associ-
Etiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522
ated with higher rates of rebleeding and, as a
Demography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522 result, with greater mortality and morbidity
Pathology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522 than typical aneurysms. Because of rarity, the
etiology and pathology are poorly understood.
Presentation and Incidence . . . . . . . . . . . . . . . . . . . . . . . . 523
Diagnosis relies on a stereotypical appearance
Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 523 and location at non-branch points on the inter-
Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525 nal carotid artery at angiography, together with
Surgical Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 526 inspection at surgery. They are increasingly
seen involving other intracranial arteries and
Endovascular Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . 528
generally considered to be caused by athero-
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 531 sclerotic mural disease.
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 532 Surgical clipping is difficult because of their
small size, sessile shape, and thin sac walls. It
is frequently complicated by operative rupture,
and alternative wrapping and trapping are often
required. For these reasons, endovascular treat-
ments are increasingly used as an alternative.
The recent introduction of high mesh density
stents, designed to redirect blood flow away
from aneurysms and reinforce the parent artery,
has stimulated the development of new pro-
cedures to manage these lesions. Initial expe-
riences with flow-diverting stents have shown
J.V. Byrne (*) a substantial improvement in outcomes, and
Nuffield Department of Surgical Sciences, Oxford stent procedures are rapidly being accepted as
University, John Radcliffe Hospital, Oxford, UK
the intervention of choice.
e-mail: james.byrne@nds.ox.ac.uk
S.H. Mørkve
Department of Neurosurgery, Haukeland University
Hospital, Bergen, Norway
e-mail: snhm@helse-bergen.no; smorkve@gmail.com

# Springer Science+Business Media New York 2016 521

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_20
522
Keywords
Aneurysm • Intracranial aneurysm • Blister
aneurysm • Blood blister aneurysm • Blister-
like aneurysms • Subarachnoid hemorrhage •
Stent • Flow diverter • Wrapping • Clip wrap-
ping • Microsurgical clipping • Endovascular
therapy
Introduction
These are a well-recognized but poorly defined
subgroup of intracranial aneurysms. They were
first described in the neurosurgical literature
30 years ago [1, 2] as small focal arterial wall
dilations occurring at non-branching points of the
internal carotid artery. About 5 % of intracranial
aneurysms are found at non-branch points, of
which a minority look like a blister of the arterial
wall on inspection at surgery [3, 4]. This type of
aneurysm has been reported to arise on the
anteromedial, dorsal, anterior, and superior internal
carotid artery wall [5]. The term blister aneurysm
has therefore been applied to describe a small
aneurysm with a thin wall found on the wall of
the supraclinoid internal carotid artery. Aneurysms
with the same appearance, on the basis of inspec-
tion at surgery or on angiography, have also been
found arising from other basal cerebral arteries.
Another characteristic feature of blister aneu-
rysms is their behavior. This is a potential source
of confusion since relying on small size and stereo-
typical angiographic appearance to distinguish this
type of aneurysm is unreliable because serial angio-
grams have shown that they may enlarge within
days of presentation. This behavior means that
their appearance on angiograms may vary from a
small bulge of the arterial wall to sessile or saccular
sac shapes [6]. Changing size and shape together
with frequent rebleeding are considered their hall-
mark behavior and the best criteria for
distinguishing them from other types of intracranial
aneurysms. Because of the thin wall evident at sur-
gery, the terms blister, blister-like [3, 4], or blood
blister aneurysm [7] have been used synonymously
by different authors. In this chapter, the simple term
blister aneurysm will be used throughout.
J.V. Byrne and S.H. Mørkve
Etiology
The primary cause is probably atherosclerosis of
the arterial wall combined with particular hemo-
dynamic forces generated by blood flow in the
supraclinoid internal carotid artery [1, 8]. The lat-
ter is most likely related to the shape of the
supraclinoid section and forces generated at
curves of the arterial wall [4]. Other mechanisms
have been proposed, including local dissection
[9], although no classic flap has been demon-
strated, or the presence of occult small branch
arteries. Abe et al. postulated that mural hemor-
rhages at sites of atheroma were important etiol-
ogy factors and that subsequent growth into a
saccular shape was caused by organization of the
hematoma overlying a wall defect [7]. Such a
mechanism would explain differences in histol-
ogy and natural history reported for this type of
aneurysm.
Demography
They are more commonly found in women than
men [9–11], and patients are generally younger in
age than those with typical intracranial aneurysms
[12]. They are reported to occur more frequently
on the right side [13]. Systemic hypertension, [9,
12] and atherosclerosis [7, 9] are associated risk
factors.
Pathology
There have been very few reports of histological
examinations of blister aneurysms. Ishikawa
et al. [5] described the autopsy findings of a man
after fatal bleeding from an aneurysm of the supe-
rior surface of the internal carotid artery. This
report described the histological examination as
follows: “The internal elastic lamina and media
had disappeared at the border between the eccen-
trically sclerotic and normal carotid wall. The gap
in the internal elastica was covered with normal
adventitia and fibrinous tissue. This portion was
not composed of collagenous tissue as ordinarily
24 Blister Aneurysms
seen in aneurysm walls. Neither infiltration of
inflammatory cells nor dissection of the artery
were observed.” Thus, ruling out dissection, they
concluded that the aneurysm was the result of
degeneration of the internal elastic lamina. In
another case report by Kim et al. [6], the histology
of a resected blister aneurysm showed smooth
muscle in the wall, and the authors considered it
to be the same as a “true aneurysm.” In this
patient, clipping was performed 12 days after
SAH and after a second angiogram had shown
the aneurysm to have grown from sessile to a
saccular shape.
These are the only two reports quoted in the
translated English literature though there have
been sufficient reports of in vivo observations at
surgery to conclude that the blister aneurysm has a
distinct pathology and no clear evidence that they
are the result of focal dissection. The observation
that a blister aneurysm can enlarge within days of
SAH suggests a dynamic process, which may be
associated with the proximity and organization of
an acute blood clot [7]. Local exposure to endog-
enous lytic agents could be a factor in the process.
Presentation and Incidence
Most patients present with acute subarachnoid
hemorrhage (80–90 %), and the rest are diagnosed
coincidental with other aneurysms. In the reports
Fig. 1 Axial CT (a) showing acute subarachnoid hemor-
rhage caused by rupture of a blister aneurysm. The aneu-
rysm was not diagnosed on preliminary CTA, but
subsequent DSA (b and c) showed a typical “thorn-
523
of large series of ruptured intracranial aneurysms,
their frequency is around 1 %. Yasargil reported
3 (0.9 %) in a series of 319 [2], Nakagawa et al.
8 (1.7 %) in a series of 460 [4], and Meling et al.’s
14 (1.5 %) in a series of 912 [13] ruptured aneu-
rysms. In Meling et al.’s series, all blister aneu-
rysms occurred at non-branching points of the
internal carotid artery, and these constituted
6.6 % of ruptured internal carotid artery
aneurysms [13].
Imaging
Because of their small size, blister aneurysms can
be difficult to identify on CT or magnetic reso-
nance imaging (MRI) scans, and catheter angio-
gram (DSA) is the best imaging modality to
identify these lesions. On DSA, a focal arterial
dilation is seen with a variety of possible shape
configurations, including fusiform, sessile, or sac-
cular dilations of the artery at non-branch points.
These changes are often difficult to distinguish
from mural dissections (see Fig. 1). The location
on the internal carotid artery helps to suggest the
diagnosis, but as discussed above, the classic
in vivo description of this lesion was made by
direct inspection at surgery, and this remains the
best method of confirming the angiogram diagno-
sis. However, small focal dilations demonstrated
in similar circumstances on arteries other than the
shaped” dilation of the internal carotid artery on the
anterior-superior surface between the posterior communi-
cating and anterior choroidal artery origins (arrows)
524
Fig. 2 Axial CT (a) and lateral DSA (b) at presentation
5 days after spontaneous subarachnoid hemorrhage caused
by a blister aneurysm of the right internal carotid artery.
The aneurysm was treated with a combination of an
endovascular stent and endosaccular coils after the patient
suffered repeat subarachnoid hemorrhage 12 days later.
internal carotid artery may also be blister aneu-
rysms, and proposing this diagnosis after angiog-
raphy alone seems appropriate.
Sequential angiograms often show the aneu-
rysm to enlarge within days of the presenting
subarachnoid hemorrhage, which helps to distin-
guish them from very small non-blister aneurysms
[6, 9] (see Fig. 2). This transformation into a larger
saccular shape on serial DSA is usually
interpreted to indicate dissection or mural hemor-
rhage as the initiating event for aneurysms at sites
other than the internal carotid artery. It creates a
potential difficulty in classification for future
authors, since blister aneurysms are increasingly
managed by endovascular treatment and
J.V. Byrne and S.H. Mørkve
Control angiograms after deployment of the stent (c) and
after insertion of coils (d) show how the aneurysm enlarged
in the interval between angiograms and developed a more
saccular shape. Small markers can be seen at the two ends
of the stent (arrows)
confirmation of the diagnosis by surgical inspec-
tion is not available (see Fig. 3).
For the time being, it can be argued that the
blister aneurysm is a precursor of a saccular aneu-
rysm, which spontaneous subarachnoid hemor-
rhage brings to medical attention and after
rupture is, for a short period of time, in an unstable
state. Whether such aneurysms left untreated
inevitably evolve to become saccular aneurysms
is unknown, but it has been reported that sponta-
neous complete regression can occur [14]. In the
future, higher resolution imaging than available
on current magnetic resonance imagers may be
able to better detect mural hemorrhage and pro-
vide noninvasive in vivo confirmation of the
24 Blister Aneurysms
Fig. 3 Unruptured tandem small aneurysms (large
arrows) of the right internal carotid artery. DSA performed
before (a) and after (b) placing a stent to cover the necks of
the aneurysms, which arise from the superior and lateral
postulated etiology and diagnosis at other arterial
locations.
Management
Blister aneurysms are notoriously difficult to treat
regardless of the approach employed, and inter-
ventions are associated with overall rates of mor-
tality and morbidity that are significantly higher
than those for typical ruptured intracranial aneu-
rysms [9, 15, 16]. The thin, fragile wall of this
type of aneurysm, in combination with their small
sizes and broad necks, are some of the features
that render them particularly prone to
intraprocedural rupture when treated using either
surgical or endovascular techniques [12, 17].
Blister aneurysms are thus considered particularly
dangerous lesions and carry an overall
re-rupture risk that is greater than that of most
saccular aneurysms [12, 17]. Another feature of
these lesions that adds to the high mortality and
morbidity rates reported after spontaneous
subarachnoid hemorrhage is a frequently
observed rapid regrowth of the aneurysmal sac
within weeks of apparently initial successful treat-
ment [15]. Blister aneurysms consequently may
require multiple treatments, all carrying similar
525
surface of the artery. Their location is typical of blister
aneurysms, but inspection at surgery was not available to
confirm the diagnosis. The small arrows show the markers
at the ends of the stent (Courtesy of Dr. M. Cellerini)
high risks, which significantly increases the total
risk of iatrogenic morbidity [15]. Adding to the
total risk is the possibility that these small lesions
are frequently not diagnosed on primary CT angi-
ography or initial catheter angiograms and are
only evident on repeat angiograms after growth
of the aneurysmal sac. This delay prolongs the
period that patients are at risk of repeat bleeding
and additional morbidity if rebleeding
occurs [15].
The best treatment of blister aneurysms has,
since their first recognition, been controversial,
and various strategies have over time been advo-
cated. Most recently, this has centered on the
relative strength and weaknesses of open micro-
surgery compared to endovascular treatments.
Parent artery sacrifice (i.e., a deconstructive tech-
nique) was for a long time considered the best
primary treatment for these lesions and is still by
many used as a rescue therapy if other techniques
fail. Advances in both microsurgical and
endovascular techniques, however, have allowed
an increasing proportion of blister aneurysms to
be treated preserving the parent artery (i.e., recon-
structive techniques) or preserving the blood flow
carried through the parent artery to its distal terri-
tory through various bypass procedures. These
techniques generally appear to offer better
526
anatomical results and patient outcomes than par-
ent artery sacrifice [13, 18].
Other measures in addition to direct interven-
tions to occlude the aneurysm are important adju-
vant treatments, which should be considered in
order to prevent aneurysm re-rupture. These
include tight control of arterial blood pressure,
careful drainage of cerebrospinal fluid, and the
use of antifibrinolytic drugs such as aminocaproic
acid [18].
Most treatment techniques proposed and
described in the literature have been directed
toward the classical blister aneurysms arising
from the internal carotid artery. Similar lesions
can, however, as already mentioned sometimes
be identified on angiography at other locations.
The typical “thorn-shaped” arterial wall dilation
(Fig. 1) has for example been demonstrated in
aneurysms of the basilar artery, and because of
this appearance, these aneurysms have also been
termed blister aneurysms [19]. However, since
endovascular therapists do not have the possibility
of direct visual inspection, it can only be assumed
that aneurysms with these appearances at other
locations have the same characteristics as those
of the ICA.
Surgical Treatment
The thin, frail walls of blister aneurysms easily
rupture during surgical dissection or during
attempted aneurysm repair [9, 12, 13, 20]. Appli-
cation of clips directly onto the aneurysm is
treacherous, and gentle dissection needs to be
performed with careful planning of aneurysm
clip placement. Ideally, the clip should be applied
completely parallel to the parent vessel, which
may be facilitated by clipping under systemic
hypotension [9, 12]. Parallel clip placement
ensures an even distribution of traction from the
aneurysm clip on the vessel wall, minimizing
distortion of the wall and allowing a more secure
and stable hold of the clip on the wall surrounding
the thin-walled and fragile sac of the blister aneu-
rysm [4, 7, 10]. Despite surgical manipulation
being performed with the utmost care, avulsion
and rupture of the aneurysm wall is not an
J.V. Byrne and S.H. Mørkve
uncommon occurrence during attempted direct
clipping of these lesions [9, 13]. Intraoperative
rupture may create a situation where sacrifice of
the parent vessel (ICA) to control bleeding is the
only possible way to regain control. However,
parent artery sacrifice has been shown by Meling
et al. to be associated with poor outcomes if
performed in SAH patients within 48 h of their
bleeding even if collaterals appeared to be ade-
quate on preoperative angiography [13]. These
authors suggested that this was due to vasospasm
secondary to subarachnoid hemorrhage reducing
the available blood flow support to the distal vas-
cular territory via collateral arteries and that parent
artery occlusion in the acute phase after hemor-
rhage should be avoided if at all possible [13].
Due to the high risk of intraoperative rupture of
blister aneurysms during attempted clipping,
alternative methods for surgical repair have been
proposed. Several authors have advocated treat-
ment with various wrapping or combined clipping
and wrapping techniques [7, 21]. The vessel-
encircling clip graft, or Sundt clip graft, was intro-
duced over 40 years ago, and recently, Park and
Meyer described the application of Sundt clip
grafts as a rescue technique when vessel dehis-
cence occurred perioperatively. They achieved
good outcomes in 76 % of patients, most of
whom were treated for blister aneurysms
[22]. Sundt clip grafts have also been successfully
applied as a salvage operation after failed
endoluminal stenting [23], and their application
remains an important tool in the arsenal of surgical
treatment techniques for selected patients. How-
ever, the encircling clip grafts have important
limitations when it comes to the anatomy of the
vessel segment around which they are to be
applied. The clip grafts cannot be applied to blister
aneurysms arising close to branch arteries without
occluding them and thereby causing ischemic
infarctions and potentially permanent neurologi-
cal deficits.
Although generally less effective than clipping
in preventing re-rupture of saccular aneurysms,
different wrapping techniques have been
employed successfully by several authors to treat
unclippable blister aneurysms [24–26]. Various
materials have been used for wrapping including
24 Blister Aneurysms
muscle [27], muslin gauze followed by a layer of
temporal fibrous fascia [28], and cotton [26]. Blis-
ter aneurysms have also been coated with various
plastic adhesives [29] including formulated cya-
noacrylates, such as Biobond or Aron Alpha [30,
31], and methyl methacrylate [29]. Wrapping
using muscle and fascia seems to be of limited
value when it comes to preventing re-rupture,
most likely because of the gradual necrosis and
disappearance of the tissue over time [27, 31, 32].
Wrapping with cotton or gauze, however, either
alone or combined with coating the wrapped
aneurysm with plastic adhesives, offers better pro-
tection over time [32]. A potential disadvantage of
using plastic adhesives relates to their potential to
cause chronic inflammatory changes with necro-
sis and fibrosis of the media layer of the vessel
wall. Used alone, cotton and muslin gauze cause
fibrosis in the adventitia but no changes in the
arterial wall media [33]. This selective reaction
of the outer layer of the vessel wall is thought to
better reinforce the aneurysm sac [27]. Several
authors therefore consider cotton used without
additional adhesive coating the best material for
wrapping ruptured aneurysms [21, 26, 33].
In one study, re-rupture rates for aneurysms
wrapped between 1965 and 1975 using muslin
gauze with or without additional muscle or adhe-
sives were found to be 8.6 % for early re-rupture,
i.e., within the first 6 months, and 1.5 % per
annum for late re-ruptures up to 10 years postop-
eratively [24]. This does not compare favorably to
an annual re-rupture risk of around 0.3–0.4 % per
year for aneurysms that were clipped in the same
period [24]. However, in the cooperative study
performed prior to the widespread availability of
CT scanning, Nishioka reported a confirmed fatal
rebleeding rate of 25 % in untreated patients with
all types of intracranial aneurysms [34], and it is
thus clear that wrapping of aneurysms offers a
significant degree of protection compared to the
re-rupture risk of untreated aneurysms. The use-
fulness of this technique has also been confirmed
in the era of microsurgical neurosurgery, and one
more recent study reported no rebleeding during a
follow-up interval of 5 years for patients with
solitary aneurysms wrapped using muslin
gauze [25].
527
A further consideration in treatment planning
is the possible late development of chronic gran-
ulation tissue after wrapping. Granulomas or
so-called gauzomas have been found to some-
times form around wrapped aneurysms [26].
These may cause cranial neuropathies, which is
the most frequent complication of aneurysm
wrapping apart from re-rupture [26]. To minimize
the risk of this happening, the wrapping material
should therefore never be placed in direct contact
with cranial nerves [21].
Wrapping of blister aneurysms is now usually
performed in combination with some form of clip
placement. Aneurysmal clips can be placed across
the lesion after it has been wrapped with gauze or
cotton (e.g., Surgicel) in a technique called “wrap-
holding clipping.” Here the wrapping material both
reinforces the arterial wall and prevents slippage
and displacement of the clip which may cause
aneurysm re-rupture. The wrapping material can
also be used to encase the arterial segment
containing the aneurysm, and an aneurysm clip
can then be placed over the wrapping material to
hold it in place without incorporating the aneurysm
or vessel wall in the clip. This technique is known
as “wrap clipping” [35]. Both techniques have been
reported to achieve better results for difficult blister
aneurysms than clipping or wrapping alone
[9]. Wrapping materials commonly in use are cot-
ton (e.g., Bemsheet or Surgicel) [35], rayon [21],
Gore-Tex [18, 26], or Dacron [36]. The latter two
fabrics are mostly used for wrap clipping.
Delayed aneurysm growth and branch occlu-
sions have been reported after wrap clipping, and
frequent follow-up imaging is therefore
recommended in particular for patients treated
with this technique [35]. Another way to combine
wrapping and clipping techniques is to put the
wrapping material around a primary clipped
lesion. Similar to wrap-holding clipping, this is
also performed in order to prevent slippage of the
clip, to strengthen the arterial wall, and to provoke
formation of scar tissue over time [18]. This tech-
nique has been used as a salvage operation
together with arterial suturing or tying the blades
of a clip to the carotid trunk with silk sutures
following perioperative rupture of blister aneu-
rysms during primary clip placement [20].
528
Over recent years, it has become increasingly
clear that blister aneurysms of the supraclinoid
internal carotid artery may involve a large portion
of the vessel circumference. Often nearly half
(i.e., 180 ) of the vessel’s circumference is
directly involved in the aneurysm with its charac-
teristic diseased thin and fragile vessel wall. This
leaves a very limited amount of normal vessel
wall onto which it is possible to place aneurysm
clips and can make clip reconstruction, with pres-
ervation of a vessel diameter sufficient to sustain
adequate perfusion in the territory supplied by the
artery, impossible [12]. Extracranial to intracra-
nial (EC-IC) bypass followed by parent artery
occlusion or trapping has therefore been increas-
ingly advocated by several authors as a means of
treating blister aneurysms [13, 18, 37, 38]. Various
bypass techniques are used depending on the
magnitude of the blood flow the bypass needs to
sustain and the inherent availability of collateral
blood flow. A low-flow bypass from the superfi-
cial temporal artery (STA) to the middle cerebral
artery (MCA), a so-called STA-MCA bypass, can
only provide limited supportive blood flow. A
high-flow bypass between the external carotid
artery (ECA) and the MCA with an interposed
arterial or venous graft (e.g., radial artery
(RA) or saphenous vein (SV)) can provide much
more substantial blood flow. When treating aneu-
rysms of the supraclinoid internal carotid artery
(ICA), a high-flow bypass such as the ECA-RA-
MCA bypass is usually required [37, 38].
In one small study, 4 out of 5 patients with
ruptured internal carotid artery blister aneurysms
had good outcomes after treatment with ECA-
RA-MCA bypass surgery and parent artery occlu-
sion [38]. A temporary low-flow STA-MCA
bypass was used during the procedure to provide
adequate perfusion to the MCA territory while the
permanent anastomosis between the ICA and
MCA using a RA graft was established
(so-called double-insurance bypass) [38]. This is
a complex procedure, but the authors’ results were
in line with the results from a large systematic
review of different aneurysms in the anterior cir-
culation treated with EC-IC bypass surgery, in
which good outcomes (mRS 0–1) were reported
in 81 % of the patients [39]. The complication rate
J.V. Byrne and S.H. Mørkve
with this kind of bypass surgery, however, espe-
cially in the setting of acute subarachnoid hemor-
rhage, is significant, and 25 % of patients in the
same review had greater neurological deficits after
surgery than before, mostly due to thromboem-
bolic events [39]. Because of this high complica-
tion rate, Spetzler and colleagues have recently
stated that they consider the EC-IC high-flow
bypass and parent artery occlusion as the last
option for treating blister aneurysms to be used
only after failure of other techniques, i.e., primary
clipping or wrap clipping [18]. They point out that
although good results are obtainable using bypass
surgery, the complication rates of this kind of
surgery immediately after subarachnoid hemor-
rhage remain high in clinical practice [18].
Endovascular Treatment
The standard endovascular treatment for ruptured
intracranial aneurysm is performed by packing the
aneurysm sac with platinum coils (i.e.,
endosaccular coiling). This approach, however,
like standard microsurgical clipping is not optimal
for treating blister aneurysms. Attempted
endosaccular coiling of these lesions has at best
yielded mixed results and is often not possible
because of their broad-based shapes and small
sizes [13, 17]. Both these features make it difficult
for coils to be retained within the aneurysm sac,
and treatment of small-sized aneurysms is gener-
ally associated with higher rates of periprocedural
rupture. Thus endosaccular coil embolization
carries a high risk of coil herniation into the parent
artery as well as intraoperative aneurysm rupture
[16]. The wide dome and inherent recurrence ten-
dency of blister aneurysms also prevent balloon
remodeling from achieving more permanent aneu-
rysm occlusion [17, 40]. Consequently, most blis-
ter aneurysms are unsuitable for endosaccular
coiling, and parent artery occlusion has histori-
cally been considered the more practical and safer
endovascular treatment option [17].
However, for the reasons already mentioned,
outcomes after parent artery sacrifice are often
poor in the acute period after subarachnoid hem-
orrhage unless supported by bypass surgery
24 Blister Aneurysms
[13, 15], even if the native collateral blood flow
capacity appeared sufficient when assessed before
treatment [13]. Parent artery occlusion also carries
the inherent risk of occluding important side
branch arteries of the internal carotid artery [15],
and therefore in many centers microsurgical repair
remained the primary treatment approach for blis-
ter aneurysms until self-expanding endovascular
stents for intracranial use became available on the
commercial market.
Combining endosaccular coil embolization
with stenting has several potential advantages.
The most important in the context of blister aneu-
rysms is that stents deployed in the parent artery
act to retain coils in the aneurysm sac. Coil and
stent treatments are performed either by first plac-
ing the stent and then introducing coils through
the struts of the stent or by catheterizing the aneu-
rysm sac and then placing the stent over the cath-
eter before introducing endosaccular coils into the
aneurysm (so-called “jailing” technique) (see
Fig. 2). Stent-assisted coiling has allowed a sig-
nificantly larger proportion of blister aneurysms to
be treated successfully with an endovascular
approach, and good clinical outcomes have been
reported in 62–89 % of patients [16, 40–42]. A
majority of treated patients, however, still had to
have more than one procedure, either for place-
ment of a second stent within the first or place-
ment of additional coils after aneurysm regrowth
becomes evident on follow-up imaging [40, 41].
Early postoperative or perioperative hemorrhage,
most often fatal, has also been reported to occur in
13–28 % of blister aneurysm patients treated
using stent and coil techniques [16, 41, 42]. The
results of initial attempts to utilize stents to aug-
ment endovascular treatment by endosaccular
coiling were mixed.
So-called flow-diverting stents were intro-
duced in 2007 and provide an alternative
endovascular treatment option for aneurysms
with a difficult anatomy such as the blister aneu-
rysm. These stents work by reconstructing lami-
nar flow in the parent vessel and redirecting blood
flow away from the aneurysm sac [43, 44]. Inflow
jets of blood into the aneurysm are reduced, and
the shear stress on the aneurysm wall is eliminated
immediately after deployment [43, 44]. The effect
529
is to induce progressive aneurysm thrombosis,
which usually occurs within weeks of treatment,
and neointimal growth covering the stent and
closing the aneurysm neck over the following
months [43, 44].
The intra-aneurysmal reduction in blood flow
induced by flow-diverting stents is closely related
to the porosity and pore density of the device
[44]. Porosity is a measure of the proportion of
open area to the total area of the stent wall, and for
flow-diverting stents, this is typically in the range
of 65–70 % (i.e., a metal coverage of 30–35 %). In
comparison, conventional stents have higher
porosities (>90 %) and therefore considerably
less metal coverage [43, 44]. Flow-diverting
stents appear to offer a better endovascular treat-
ment alternative for blister aneurysms in part
because they can be used without placement of
any adjunct endosaccular coils. This theoretically
reduces the risk of periprocedural aneurysm rup-
ture compared to the stent and coil technique but
needs to be proven in practice [45, 46].
There are currently four commercially avail-
able flow-diverting stents that are in common use.
These are the Silk/Leo flow diverter (Balt Extru-
sion, Montmorency, France), the Pipeline Embo-
lization Device (PED - Covidien, Dublin,
Ireland), the Flow Redirection Endoluminal
Device (FRED – Microvention, Tustin, CA,
USA), and the Surpass flow diverter (Surpass,
Stryker Neurovascular, Fremont, CA, USA).
Flow-diverting stents have been used to treat blis-
ter aneurysms with encouraging results [45, 47].
Good outcomes (mRS 0–2 or unchanged/
improved neurological status) and combined mor-
bidity and mortality rates of 0–10 % were reported
in 86–100 % of patients treated with flow-
diverting stents for blister aneurysms not
presenting acutely after subarachnoid hemorrhage
[46, 48]. Recently, good outcomes (mRS 0–2)
were also reported in 92 % of patients treated for
ruptured blister aneurysms using Silk stents with a
combined morbidity and mortality rate of
18 % [19]. Similar results have been obtained by
others treating ruptured blister aneurysms with
both Silk stents and PEDs with good clinical out-
comes (mRS 0–2 or GOS 4–5) in 66–94 % of
patients [49–51]. Procedure-related morbidity
530
and mortality rates, however, remain as high as
15–25 %, and thromboembolic or hemorrhagic
complications in the postoperative period have
been reported in 5–33 % [49–51]. Aneurysms
treated with flow-diverting stents tend to
gradually occlude as endosaccular thrombosis
occurs over a period of time after treatment,
and complete occlusion has been reported in
75–100 % at 3–6 months and in 94–100 % at
12 months [19, 49, 51, 52].
Deployment of any stent, and especially flow-
diverting stents with their larger metal content,
requires the use of drugs to prevent spontaneous
in-stent thrombosis. Patients are therefore obliged
to receive antiplatelet drugs with attendant risks of
spontaneous hemorrhage and more severe bleed-
ing should the aneurysm rupture or re-rupture. In
elective treatments of unruptured aneurysms,
most centers today start treatment with two
antiplatelet drugs, typically aspirin and
clopidogrel, 5–17 days before treatment
[53–55]. Such dual antiplatelet therapy is
employed in part to overcome potential resistance
to one or the other drug, which occurs in about
20 % of people. The heterogeneity of individuals’
responses to antiplatelet drugs is idiosyncratic for
aspirin, but for clopidogrel, it is attributable to this
being a prodrug. Clopidogrel is administered
orally and though usually well absorbed requires
activation by cytochrome P450 enzymes in the
liver. Its effectiveness is thus liable to be affected
by interactions with other commonly used drugs
such as atorvastatin and verapamil [54]. Whereas
15–20 % of the population are hyperresponders to
clopidogrel, as many as 25–32 % are
hyporesponders to the same drug [53–55]. Platelet
function tests are therefore usually recommended
at least once before surgery to make sure patients
are not under- or overtreated and to make any
necessary adjustments in dose or switch from
clopidogrel to an alternative antiplatelet drug
such as ticagrelor or prasugrel before the proce-
dure [54]. To further complicate matters, up to
75 % of patients will also show an increasing
response to clopidogrel during long-term treat-
ment and may thus become overtreated. It has
therefore been suggested to perform follow-up
platelet function tests in all patients treated over
J.V. Byrne and S.H. Mørkve
some time with clopidogrel [54]. This precaution
has been recommended in order to reduce the
incidence of hemorrhagic complications that
have been shown to occur preferentially in
patients overtreated with clopidogrel [54]. Treat-
ment with two antiplatelet drugs (e.g., aspirin and
clopidogrel) is typically continued for a minimum
of 3–6 months after the procedure, and then aspi-
rin is usually continued for at least an additional
6 months or in some instances for the remainder of
the patient’s life [54, 55].
The most widespread method of performing
platelet function tests in endovascular therapy
units is the point-of-care system VerifyNow
(Accumetrics, San Diego, CA). This system can
measure a patient’s response to all three main
classes of antiplatelet agents in current use.
These are blocking the formation of thromboxane
A2 in platelets (aspirin), platelet receptor P2Y12
inhibitors (thienopyridines such as clopidogrel),
and platelet glycoprotein IIb/IIIa receptor inhibi-
tors (e.g., abciximab). The system measures plate-
let function qualitatively by exposing patient
whole-blood samples to a mixture of agonists
causing platelet activation and fibrinogen-coated
microparticles. Platelet function is measured
based upon the activated platelets’ ability to bind
to the microparticles causing them to aggregate.
Aspirin inhibition is reported as aspirin reaction
units (ARU), whereas clopidogrel inhibition (and
other thienopyridines) is reported as P2Y12 reac-
tion units (PRU). Importantly, when it comes to
neuroendovascular treatments using stents, PRU
values <60 or >240 have been shown to be a
strong predictor for thromboembolic and hemor-
rhagic complications, respectively, up to 6 months
postoperatively [53, 54].
In the setting of acute subarachnoid hemor-
rhage and the use of stents to treat ruptured aneu-
rysms, administration of antiplatelet drugs
requires careful consideration. Though various
therapy regimens have been used for prophylaxis
against stent thrombosis in this situation, there is
still no consensus on the optimal treatment. In
most centers, however, loading doses of
clopidogrel and aspirin or ticagrelor are used
[16, 54, 55]. Typically, emergency loading doses
in the range of 325–500 mg aspirin and
24 Blister Aneurysms
300–600 mg clopidogrel are administered at the
time of, or up to 8 h prior to, endovascular treat-
ment. Both drugs are then continued in lower
maintenance doses from the day after treatment
as for elective cases [49, 55]. Platelet function
tests are under these circumstances not commonly
recommended prior to treatment, as they will not
affect the planned treatment strategy, but they can
be of value at a later stage to ensure adequate post-
procedural antiplatelet therapy.
It has been suggested that the risk of intracere-
bral hemorrhage in stented patients on antiplatelet
drugs is independent of a preexisting subarach-
noid hemorrhage [47, 56], and similar delayed
hemorrhage rates have been reported among
patients treated for ruptured or unruptured aneu-
rysms using flow-diverting stents [45–52]. How-
ever, dual antiplatelet therapy in the acute period
after subarachnoid hemorrhage is a delicate bal-
ance and complicates, in particular, the manage-
ment of hydrocephalus. Operative intracranial
procedures, such as insertion of an external ven-
tricular drain (EVD) or placement of a ventricu-
loperitoneal shunt (VP shunt), are associated with
significantly increased risks of causing intracra-
nial hemorrhages in this setting [15, 57, 58], and
particular attention to surgical hemostasis is there-
fore warranted during such procedures.
Gonzales et al. recently published their experi-
ence with ruptured blister aneurysms together
with a systematic review of the existing literature
on treatment of these aneurysms [15]. The authors
reported good outcomes (mRS 0–2) in 91 % of
their own patients treated from 2003 to 2013.
Endovascular therapy was their primary approach
and was initially performed using conventional
stents combined with endosaccular coiling but
was later changed to flow-diverting stents (single
or multiple stents telescoped within each other)
without adjunctive endosaccular coils. If uncer-
tainty as to the likely best approach for the patient
existed after imaging by CT angiography, catheter
angiography was performed in a hybrid theater
where endovascular or surgical treatment could
proceed directly after assessment of the aneurysm
anatomy by DSA performed in the operating
room. In their review of the literature, they found
63 original papers including 322 patients with
531
ruptured blister aneurysms. A majority of the
patients were treated surgically, but 28 % required
more than one therapeutic approach. Both the
overall morbidity (mRS >2 or new neurological
deficit following treatment) and mortality were
higher in patients who received surgical treat-
ment, 18 % and 14 % respectively, as compared
to 3.4 % and 11.5 % respectively in patients
receiving endovascular treatment [15]. Thus
despite the need for antiplatelet drugs being a
major constraint for the endovascular treatment
of ruptured blister aneurysms with flow-diverting
stents, this technique appears the most promising
prospect for improving treatment in these difficult
aneurysms.
Summary
Blister aneurysms are rare and comprise less than
1 % of ruptured intracranial aneurysms. Various
terms, such as blood blister and blister-like, have
been used for this type of aneurysm. A common
feature is a higher frequency of rebleeding than
occurs in the usual type of ruptured intracranial
aneurysms and, as a result, greater mortality and
morbidity. The pathological cause for this differ-
ent behavior is poorly understood because of their
rarity. In vivo diagnosis relies on a stereotypical
appearance on angiography or after inspection at
surgery and a typical location at non-branch
points of the internal carotid artery. However, the
same appearances and natural histories are
increasingly reported for aneurysms involving
other intracranial arteries. They are generally con-
sidered to be caused by atherosclerotic mural
disease.
Surgical clipping is difficult because of their
small size, sessile shape, and thin sac walls. It is
frequently complicated by operative rupture, and
alternative but less secure treatments by wrapping
or trapping are often considered a safer option. For
these reasons, endovascular treatments have been
proposed and used as an alternative. The recent
introduction of high mesh density endovascular
stents, which are designed to redirect blood flow
away from the aneurysm lumen and reinforce the
parent artery wall, has stimulated the development
532
of new procedures to manage these lesions. Initial
experiences of treatment with endovascular stents
have shown a substantial improvement in out-
comes, and stent procedures are rapidly being
accepted as the intervention of choice.
References
1. Ohara H, Sakamoto T, Suzuki J (1979) Sclerotic brain
aneurysms. In: Suzuki J (ed) Cerebral aneurysms. Neu-
ron, Tokyo, pp 673–682
2. Yasargil MG (1984) Clinical considerations, surgery of
the intracranial aneurysm and results. Georg Thieme
Verlag, Stuttgart/New York
3. Yoshimoto Y, Ochiai C, Nagai M (1996) Cerebral
aneurysms unrelated to arterial bifurcations. Acta
Neurochir (Wien) 138:958–963
4. Nakagawa F, Kobayashi S, Takemae T, Sugita K
(1986) Aneurysms protruding from the dorsal
wall of the internal carotid artery. J Neurosurg
65:303–308
5. Ishikawa T, Nakamura N, Houkin K, Nomura N (1997)
Pathological consideration of a “blister-like” aneurysm
at the superior wall of the internal carotid artery: case
report. Neurosurgery 40:403–405
6. Kim JH, Kwon TH, Kim JH, Park YK, Chung HS
(2006) Internal carotid artery dorsal wall aneurysm
with configurational change: are they all false aneu-
rysms? Surg Neurol 66:441–443
7. Abe M, Tabuchi K, Yokoyama H, Uchino A (1998)
Blood blisterlike aneurysms of the internal carotid
artery. J Neurosurg 89:419–424
8. Sano K (1979) Internal carotid posterior communicat-
ing anterior choroidal region aneurysms. Microsurgical
treatment. In: Pia HW, Langmaid C, Zierski J (eds)
Cerebral aneurysms. Advances in diagnosis and ther-
apy. Springer, Berlin/Heidelberg/New York, p 252
9. Ogawa A, Suzuki M, Ogasawara K (2000) Aneurysms
at nonbranching sites in the supraclinoid portion of the
internal carotid artery: internal carotid artery trunk
aneurysms. Neurosurgery 47:578–583
10. Shigeta H, Kyoshima K, Nakagawa F, Kobayashi S
(1992) Dorsal internal carotid artery aneurysms with
special reference to angiographic presentation and sur-
gical management. Acta Neurochir (Wien) 119:42–48
11. Kinouchi H, Mizoi K, Nagamine Y, Yanagida N,
Mikawa S, Suzuki A, Sasajima T, Yoshimoto T
(2002) Anterior paraclinoid aneurysms. J Neurosurg
96:1000–1005
12. Sim SY, Shin YS, Cho KG, Kim SY, Kim SH, Ahn YH,
Yoon SH, Cho KH (2006) Blood blister-like aneu-
rysms at nonbranching sites of the internal carotid
artery. J Neurosurg 105:400–405
13. Meling TR, Sorteberg A, Bakke SJ, Slettebø H,
Hernesniemi J, Sorteberg W (2008) Blood blister-like
aneurysms of the internal carotid artery causing
J.V. Byrne and S.H. Mørkve
subarachnoid haemorrhage: treatment and outcome.
J Neurosurg 108:662–671
14. Ueta T, Ichi S, Ochi T, Suzuki I (2004) Spontaneous
regression of an aneurysm at a nonbranching site of the
supraclinoid internal carotid artery. Case report.
J Neurosurg 101:1070–1072
15. Gonzalez AM, Narata AP, Yilmaz H, Bijlenga P,
Radovanic I, Schaller K, Lovblad KO, Pereira VM
(2014) Blood blister-like aneurysms: single center
experience and systematic literature review. Eur J
Radiol 83:197–205
16. Meckel S, Singh TP, Undrén P, Ramgren B, Nilsson
OG, Phatouros C, McAulife W, Cronqvist M (2011)
Endovascular treatment using predominantly stent-
assisted coil embolization and antiplatelet and
anticoagulation management of ruptured blood
blister-like aneurysms. AJNR Am J Neuroradiol
32:764–771
17. Park JH, Park IS, Han DH, Kim SH, Oh CW, Kim JE,
Kim HJ, Han MH, Kwon OK (2007) Endovascular
treatment of blood blister-like aneurysms of the inter-
nal carotid artery. J Neurosurg 106:812–819
18. Garrett M, Spetzler RF (2012) Surgical treatment of
blister-like aneurysms. World Neurosurg 77:76–77
19. Aydin K, Arat A, Sencer S, Hakyemez B,
Barburoglu M, Sencer A, Izgi N (2014) Treatment of
ruptured blood blister-like aneurysms with flow
diverter Silk stents. J Neurointerv Surg. doi:10.1136/
neurintsurg-2013-011090
20. Joo SP, Kim TS, Moon KS, Kwak HJ, Lee JK, Kim JH,
Kim SH (2006) Arterial suturing followed by clip
reinforcement with circumferential wrapping for
blister-like aneurysms of the internal carotid artery.
Surg Neurol 66:424–428
21. Figueiredo EG, Foroni L, Monaco BA, Gomes MQ,
Sterman Neto H, Teixeira MJ (2010) The clip-wrap
technique in the treatment of unclippable aneurysms.
Arq Neuropsiquiatr 68:115–118
22. Park PJ, Meyer FB (2010) The Sundt clip graft. Neu-
rosurgery 66:300–305
23. Cho TG, Hwang SN, Nam TK, Park SW (2012) Sal-
vage surgical treatment for failed endovascular proce-
dure of a blood blister-like aneurysm. J Cerebrovasc
Endovasc Neurosurg 22:295–303
24. Todd NV, Tocher JL, Jones PA, Miller JD (1989) Out-
come following aneurysm wrapping: a 10-year follow-
up review of clipped and wrapped aneurysms.
J Neurosurg 70:841–846
25. Cudlip SA, Kitchen ND, McKahn GM, Bell BA (1998)
Wrapping of solitary intracranial aneurysms, outcome
at five years. Acta Neurochir (Wien) 140:1167–1171
26. Desmukh VR, Kakarla UK, Udaya K, Figueiredo EG,
Zabramski M, Spetzler R (2005) Long-term clinical
and angiographic follow-up of unclippable wrapped
intracranial aneurysms. Neurosurgery 58:434–442
27. Sachs E Jr (1972) The fate of muscle and cotton
wrapped about intracranial carotid artery aneurysms.
A laboratory and clinico-pathological study. Acta
Neurochir (Wien) 26:121–137
24 Blister Aneurysms
28. Gillingham FJ (1958) The management of ruptured
intracranial aneurysms. Hunterian lecture. R Coll
Surg Engl 23:89–117
29. Dutton JEM (1956) Intracranial aneurysms. A new
method of surgical treatment. Br J Med 2:585–586
30. Cartoon CA, Kennedy JC, Heifetz MD et al (1965) The
use of a plastic adhesive (methyl-2-cyanoacrylate
monomer) in the management of intracranial aneu-
rysms and leaking cerebral vessels. A report of
15 cases. In: Sahs AL, Field WS (eds) Intracranial
aneurysms and subarachnoid haemorrhage. Charles
C. Thomas, Springfield
31. Minakawa T, Koike T, Fujii Y, Ishii R, Tanaka R, Arai
H (1987) Long term results of ruptured aneurysms
treated by coating. Neurosurgery 21:660–663
32. Mount LA, Antunes JL (1975) Results of treatment of
intracranial aneurysms by wrapping and coating.
J Neurosurg 42:189–193
33. Herrera O, Kawamura S, Yasui N, Yoshida Y (2005)
Histological changes in the rat common carotid artery
induced by aneurysmal wrapping and coating mate-
rials. Neurol Med Chir (Tokyo) 39:134–139
34. Nishioka H (1966) Report on the cooperative study of
intracranial aneurysms and subarachnoid
haemorrhage. Section VII. I. Evaluation of the conser-
vative management of ruptured intracranial aneurysms.
J Neurosurg 25:574–592
35. Kim YB, Hong CK, Chung J, Joo JY, Huh SK (2014)
Long-term clinical and angiographic outcomes of
wrap- clipping strategies for unclippable cerebral aneu-
rysms. Yonsei Med J 55:401–409
36. Nakano S, Iseda T, Yoneyama T, Goya T, Wakisaka S
(2000) A combination of wrapping and clipping using
a collagen-impregnated dacron fabric (Hemashield).
Surg Neurol 53:330–333
37. Kazumata K, Nakayama N, Nakamura T,
Kamiyama H, Terasaka S, Houkin K (2014) Changing
treatment strategy from clipping to radial artery graft
bypass and parent artery sacrifice in patients with rup-
tured blister-like internal carotid artery aneurysms.
Neurosurgery 10(Suppl 1):66–73
38. Murai Y, Mizunari T, Umeoka K, Tateyama K,
Kobayashi S, Teramoto A (2012) Ischemic complica-
tions after radial artery grafting and aneurysmal trap-
ping for ruptured internal carotid artery anterior wall
aneurysms. World Neurosurg 77:166–171
39. Schaller B (2008) Extracranial-intracranial bypass
to reduce the risk of ischemic stroke in intracranial
aneurysms of the anterior cerebral circulation:
a systematic review. J Stroke Cerebrovasc Dis
17:287–298
40. Lee BH, Kim BM, Park MS, Park SI, Cheung EC, Suh
SH, Choi CS, Won YS, Yu IK (2009) Reconstructive
endovascular treatment of ruptured blood blister-like
aneurysms of the internal carotid artery. J Neurosurg
110:431–436
41. Fang YB, Li Q, Yang PF, Huang QH, Zhao WY, Xu Y,
Hong B, Liu JM (2013) Treatment of blood blister-like
aneurysms of the internal carotid artery with stent-
533
assisted coil embolization. Clin Neurol Neurosurg
115:920–925
42. Ihn YK, Kim SH, Sung JH, Kim TG (2012) The
efficacy of endovascular treatment of ruptured blood
blister-like aneurysms using stent-assisted coil embo-
lization. Interv Neuroradiol 18:432–441
43. Sadasivan C, Cesar L, Seong J, Rakian A, Hao Q, Tio
FO, Wakhloo AK, Lieber BB (2009) An original flow
diversion device for the treatment of intracranial aneu-
rysms: evaluation in the rabbit elastase-induced model.
Stroke 40:952–958
44. Zanaty M, Chaloui N, Tjoumakaris SI, Rosenwasser
RH, Gonzalez LF, Jabbour P (2014) Flow-diversion
panacea or poison? Front Neurol. doi: 10.3389/
fneur.2014.00021
45. Martin AR, Cruz JP, Matouk CC, Marotta TR (2012)
The pipeline flow-diverting stent for exclusion of rup-
tured intracranial aneurysms with difficult morphol-
ogies. Neurosurgery 70:21–28
46. Chalouhi N, Zanati M, Tjoumakaris S, Gonzalez LF,
Hasan D, Kung D, Rosenwasser RH, Jabbour P (2014)
Treatment of blister-like aneurysms with the
pipeline embolization device. Neurosurgery
74:527–532
47. Rasskazoff S, Silvaggio J, Brouwer PA, Kaufmann A,
Nistor A, Iancu D (2010) Endovascular treatment of a
ruptured blood blister-like aneurysm with a flow-
diverting stent. Interv Neuroradiol 16:255–258
48. Wagner A, Cortsen M, Hauerberg J, Romner B, Wag-
ner MP (2012) Treatment of intracranial aneurysms.
Reconstruction of the parent artery with flow-diverting
SILK-stent. Neuroradiology 54:709–718
49. Cruz JP, O’Kelly C, Kelly M, Wong JH, Alshaya W,
Martin A, Spears J, Marotta TR (2013) Pipeline embo-
lization device in aneurysmal subarachnoid
haemorrhage. AJNR Am J Neuroradiol 34:271–276
50. Causin F, Pascarella R, Pavesi G, Marasconi R,
Zambon G, Battaglia R, Munari M (2011) Acute
endovascular treatment (<48 hours) of uncoilable rup-
tured aneurysms at non-branching sites using Silk
flow-diverting devices. Interv Neuroradiol 17:357–364
51. Nerva JD, Morton RP, Levitt MR, Osbun JW, Ferreira
MJ, Ghodke BV, Kim LJ (2014) Pipeline embolization
device as primary treatment for blister aneurysms and
iatrogenic pseudoaneurysms of the internal carotid
artery. J Neurointerv Surg. doi:10.1136/neurintsurg-
2013-011047
52. Brinjikji W, Murad M, Lanzino G, Cloft HJ, Kallmes
DF (2013) Endovascular treatment of intracranial
aneurysms with flow diverters. A metaanalysis. Stroke
44:442–447
53. Delgado Almandoz JE, Crandall BM, Scholz JM,
Fease JL, Anderson RE, Kadkhodayan Y, Tubman
DE (2013) Pre-procedure P2Y12 reaction units value
predicts perioperative thromboembolic and hemor-
rhagic complications in patients with cerebral aneu-
rysms treated with the Pipeline embolization device.
J Neurointerv Surg 5. doi:10.1136/neurintsurg-2012-
010582
534
54. Delgado Almandoz JE, Crandall BM, Scholz JM,
Fease JL, Anderson RE, Kadkhodayan Y, Tubman
DE (2014) Last-recorded P2Y12 reaction units value
is strongly correlated with thromboembolic and hem-
orrhagic complications occurring up to 6 months after
treatment in patients with cerebral aneurysms treated
with the Pipeline embolization device. AJNR Am J
Neuroradiol 35:128–135
55. Faught RW, Satti SR, Hurst RW, Pukenas BA, Smith
MJ (2014) Heterogeneous practice pattern regarding
antiplatelet medications for neuroendovascular
stenting in the USA: a multicenter survey.
J Neurointerv Surg. doi:10.1136/neurintsurg-2013-
010954
J.V. Byrne and S.H. Mørkve
56. Pearl M, Gregg L, Gailloud P (2010) Endovascular
treatment of aneurysmal subarachnoid haemorrhage.
Neurosurg Clin N Am 21:271–280
57. Mahaney KB, Chalouhi N, Viljoen S, Smietana J,
Kung DK, Jabbour P, Bulsara KR, Howard M, Hasan
DM (2013) Risk of hemorrhagic complication associ-
ated with ventriculoperitoneal shunt placement in
aneurysmal subarachnoid hemorrhage patients on
dual antiplatelet therapy. J Neurosurg 119:937–942
58. Kung DK, Policeni BA, Capuano AW, Rossen JD,
Jabbour PM, Torner JC, Howard MA, Hasan D
(2011) Risk of ventriculostomy-related hemorrhage in
patients with acutely ruptured aneurysms treated using
stent-assisted coiling. J Neurosurg 114:1021–1027
 Surgical Treatment of Aneurysms
25
Saul F. Morales-Valero, Shanna Fang, and Giuseppe Lanzino

Contents Abstract
Introduction and Historical Overview . . . . . . . . . . . . 536 Although an increasing number of aneurysms
are treated with endovascular therapy, tradi-
Indications for Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . 537
tional surgical treatment continues to be an
Choice of Treatment Modality . . . . . . . . . . . . . . . . . . . . . 537 important therapeutic option for both ruptured
Preoperative Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538 and unruptured aneurysms. The goal of aneu-
rysm treatment is to isolate the aneurysm from
Surgical Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538
Anesthetic Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540 the circulation while ensuring patency of par-
Critical Steps in Aneurysm Surgery (Table 1) . . . . . . 540 ent artery and/or perforator branches. With
Intraoperative Assessment of Aneurysm advanced neurosurgical approaches focused
Occlusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543 on minimizing brain manipulation, careful
Intraoperative Angiography . . . . . . . . . . . . . . . . . . . . . . . . . 543 arachnoid dissection, adjunctive intraoperative
Microvascular Doppler Ultrasonography . . . . . . . . . . . . 543 monitoring, and imaging techniques, aneu-
Indocyanine Green Fluorescent
Videoangiography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543 rysm surgery can reliably achieve definitive
occlusion with low complication rates. In the
Surgical Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543
Pterional Craniotomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 544
last decade, the widespread utilization of non-
Other Surgical Approaches for the Treatment invasive vascular imaging techniques has led
of Aneurysms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545 to an increased number of unruptured aneu-
Postoperative Care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545 rysms that are detected and treated. Noninva-
sive imaging studies such as MRA and CTA
Complications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 546
play an ever-increasing role not only for the
Follow-Up of Surgically Treated Aneurysms . . . . . 546 preoperative assessment but also postoperative
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 547 monitoring of patients with intracranial
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 547
aneurysms.

Keywords
S.F. Morales-Valero • G. Lanzino (*) Intracranial aneurysms • Cerebral aneurysms •
Department of Neurologic Surgery, Mayo Clinic, Vascular disorder • Surgical clipping • Surgical
Rochester, MN, USA
treatment • Carotid balloon occlusion • Digital
e-mail: saulfmv@gmail.com;
lanzino.giuseppe@mayo.edu subtraction angiography • Magnetic resonance
angiography • Computed tomography angiog-
S. Fang
Mayo Medical School, Mayo Clinic, Rochester, MN, USA raphy • Indocyanine green angiography •
e-mail: fang.shanna@mayo.edu Intraoperative angiography • Ruptured
# Springer Science+Business Media New York 2016 535
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_5
536
aneurysm • Unruptured aneurysm • Aneurys-
mal subarachnoid hemorrhage
Introduction and Historical Overview
Aneurysm rupture and subarachnoid hemorrhage
are associated with significant morbidity and mor-
tality. Patients who suffer aneurysmal subarach-
noid hemorrhage carry an average case fatality
rate as high as 52 %, while as many as 46 % of
those surviving the subarachnoid hemorrhage
have significant long-term cognitive impairment,
and a third of survivors are left dependent on
lifelong care [1–4]. For these patients, the immi-
nent danger is rebleeding from the ruptured aneu-
rysm, so the therapeutic goal is to isolate and
secure the aneurysm to prevent this catastrophic
complication. The mortality associated with rup-
tured aneurysms has decreased in the last three
decades with early surgical intervention, which
has been shown to reduce the risk of rebleeding
and improve the chance for good recovery in
patients with subarachnoid hemorrhage [5–7]. In
recent years, advances and widespread utilization
of noninvasive imaging have led to a large number
of unruptured and often incidental aneurysms to
be diagnosed, and this has been associated with an
increase in the number of invasive procedures to
treat intact aneurysms as well.
Surgical clipping has long been considered the
gold standard for the treatment of intracranial
aneurysms. Direct surgical clipping excludes the
aneurysm from the circulation and removes the
risk of growth and rupture. In recent years, paral-
lel to the development of endovascular tech-
niques, there has been a progressive reduction in
the number of surgically treated aneurysms. Nev-
ertheless, surgery continues to be an important
therapeutic strategy in the treatment of these
lesions.
Prior to 1970, the method of treating aneu-
rysms was carotid ligation, which was shown to
decrease the risk of rebleeding compared to con-
servative treatment [8–10]. The high rates of treat-
ment failure and ischemic complications
associated with this procedure were greatly
reduced with modifications that included the use
S.F. Morales-Valero et al.
of preoperative temporary carotid occlusion test-
ing, development of adjustable clamps for gradual
carotid occlusion, and introduction of adjuvant
extracranial-to-intracranial arterial bypass [8, 9,
11, 12]. Another well-accepted method of carotid
occlusion for aneurysm treatment at the time was
parent artery sacrifice with permanent balloon
occlusion, although the success of treatment was
similarly dependent on the degree of collateral
circulation available to prevent ischemia
[13–16]. Long-term complications of carotid and
parent artery occlusion include the risk of imme-
diate or late cerebral ischemia, as well as devel-
opment of flow-associated aneurysms [11,
17–20]. In a review of patients treated with ther-
apeutic carotid occlusion, 4.3 % of patients were
found to have de novo aneurysm formation on
average 9.1 years after treatment, thus suggesting
the need for extensive follow-up [21]. Nowadays,
carotid sacrifice is used only in selected cases of
large and giant aneurysms that are not amenable to
surgical clipping or endovascular treatment and
has shown good outcomes and low mortality rates
[20, 22–24].
In the late 1960s and early 1970s with the
introduction of the surgical microscope and
improvements in surgical clip technology, direct
clipping with the use of microsurgical techniques
became the modus operandi and led to the devel-
opment of modern aneurysm surgery as we know
it today. The next important step in the treatment
of intracranial aneurysms was the introduction of
the Guglielmi Detachable Coil in 1992, which
represented a dramatic paradigm shift with the
eventual establishment of endovascular therapy
as an acceptable alternative, and even first-line
option, to traditional clip ligation at many centers
[25]. However, microsurgery remains the funda-
mental gold standard for certain aneurysms with
complex morphological features such as small
size, broad-based neck, those located in small
diameter parent arteries. Surgical treatment is
also preferred for certain patients such as those
of younger age and good functional status, as it
provides immediate and long-term aneurysmal
obliteration. Given the variety of considerations
that factor into decision making for the best ther-
apeutic approach for each individual aneurysm,
25 Surgical Treatment of Aneurysms
a multidisciplinary cerebrovascular team is neces-
sary to work together to consider both surgical and
endovascular options in every case [26–28]. This
chapter provides an overview of the principles of
modern intracranial aneurysm surgery.
Indications for Treatment
Invasive treatment of an intracranial aneurysm is
dependent on many different conditions, the most
important one being the rupture status. A patient
who presents with intracranial hemorrhage from a
ruptured aneurysm has a greater than 60 % risk of
re-rupture over the next 6 months, with the risk
being much higher during the first 2–3 weeks after
the original bleed [29]. Therefore, treatment of a
ruptured aneurysm is indicated in the vast major-
ity of cases except in the very elderly who present
with impaired neurological condition and those
patients in deep coma who fail to improve after
maximal neurological and systemic resuscitation.
While the decision to treat is fairly clear in a
patient with a ruptured aneurysm, the decision-
making process is much more difficult in the
patient who is found to harbor an unruptured
intracranial aneurysm. Detailed analyses of the
factors that play a role in the decision to treat
unruptured aneurysms are beyond the scope of
this chapter.
In general, treatment of an unruptured aneu-
rysm is recommended for patients harboring
aneurysms that are symptomatic, those
documented to have increased in size on serial
imaging, those larger than 6 mm, and aneurysms
in certain locations that seem to harbor a worse
natural history (e.g., posterior communicating
artery (PComm), the posterior circulation, and
the anterior communicating artery complex).
Patient age, presence of risk factors such as
smoking and hypertension, as well as family his-
tory of a ruptured intracranial aneurysm also play
a pivotal role in the decision to treat an unruptured
aneurysm. The patient’s overall attitude about the
knowledge of having an aneurysm and whether or
not this has impacted negatively their quality of
life is yet another factor to take into account.
Although generalizations cannot be made and
537
each patient and aneurysm should be considered
individually, a consensus group of international
experts has recently come up with a scoring sys-
tem that can be used in deciding whether or not an
unruptured aneurysm should be considered for
invasive treatment [30].
Choice of Treatment Modality
Surgical clipping of intracranial aneurysms has
long been considered the modality of choice and
the gold standard for the treatment of intracranial
aneurysms. However, since the introduction of the
Guglielmi detachable coil in clinical practice in
1992, endovascular treatment has emerged as a
valid alternative to direct surgical clipping. Due to
technological advances, an increasing number of
aneurysms are treated nowadays by endovascular
means. Three prospective studies have been
conducted in patients with ruptured aneurysms
comparing surgical clipping to endovascular
coiling [31–33]. A meta-analysis of these studies
showed similar results in the three studies, dem-
onstrating unequivocally that the percentage of
patients with unfavorable outcome (modified
Rankin score of 4–6) is significantly lower in
patients treated endovascularly compared to
those treated with surgical clipping. This conclu-
sion indicates that for patients with ruptured intra-
cranial aneurysms amenable to either treatment,
endovascular coiling is associated with a better
functional outcome and should be the preferred
treatment modality [34]. However, concerns exist
about the long-term durability of endovascular
coiling with the risk of recurrence, need for
retreatment, and underlying risk of long-term
risk of re-rupture of the aneurysm. Because of
these concerns, surgical clipping might be pre-
ferred for very young patients with ruptured
aneurysms [35].
There are no available randomized studies
comparing endovascular to surgical treatment in
patients with unruptured intracranial aneurysms.
Overall, given the perceived less invasiveness of
endovascular treatment, an increasing number of
unruptured aneurysms are treated by
endovascular means rather than surgical clipping.
538
More recent developments in endovascular tech-
niques, such as stent assisted coiling [36] and flow
diversion [37], have further expanded the indica-
tions for endovascular treatment of unruptured
aneurysms. In general, endovascular treatment is
preferred for unruptured aneurysms that are ame-
nable to safe treatment using this modality. How-
ever, surgical clipping continues to have an
important role when treating certain aneurysms
of very small size or those with a geometry and
location (particularly MCA (middle cerebral
artery) bifurcation aneurysms) that is unfavorable
for endovascular treatment.
Preoperative Imaging
The diagnosis of a ruptured intracranial aneurysm
is often suggested by the presence of subarach-
noid hemorrhage on non-contrast head computed
tomography (CT) in patients with the typical clin-
ical presentations [38, 39]. However, many aneu-
rysms are detected incidentally during brain
imaging studies performed in the workup of neu-
rological complaints. In order to confirm the diag-
nosis and accurately characterize these lesions,
proper vascular imaging techniques are necessary
to accurately guide treatment decisions. From the
surgical point of view, it is of particular interest to
assess the anatomy of the aneurysm particularly
with attention to the orientation, diameter, neck
width, and presence of perforator branches
derived from the sac or neck of aneurysms. It is
also useful to determine the relationship of the
aneurysm to the parent vessel and surrounding
neural or osseous structures. Other features such
as the presence of calcifications at the neck level
are also important, since this can imply a more
challenging treatment.
Many different imaging techniques are avail-
able to evaluate intracranial aneurysms. Digital
subtraction angiography (DSA), computed
tomography angiography (CTA), and magnetic
resonance angiography (MRA) have been
implemented in the evaluation of these vascular
lesions [38]. Unquestionably, catheter angiogra-
phy remains the gold standard for the diagnosis of
intracranial aneurysms [39]. This modality
S.F. Morales-Valero et al.
provides high-resolution images and allows for
accurate characterization of the aneurysm anat-
omy (Fig. 1). Moreover, it allows immediate treat-
ment if endovascular therapy is considered
suitable. Despite its invasive nature, the compli-
cation rate in hands of experienced neuroradiolo-
gists is very low [40]. Classically, a
two-dimensional digital subtraction angiography
with three different projections is obtained to pro-
vide appropriate detail. However, three-
dimensional reconstruction (Fig. 1b) is increas-
ingly being used to obtain a better morphological
visual depiction with the advantage of manipula-
tion and free rotation of the vascular tree, which is
very useful for treatment planning. This technique
has also been shown to improve the detection of
additional small aneurysms when compared to the
classic two-dimensional angiography [41].
CTA (Fig. 2) is a fast and widely available
modality that can be performed quickly in the
acute setting after subarachnoid hemorrhage.
The sensitivity and specificity of this modality
are reported as 98 % and 100 %, respectively
[42]. It is possible to derive three-dimensional
renderings of the aneurysm, and this is particu-
larly useful in depicting the relationship of the
aneurysm to its surrounding bony structures
[38]. The use of CTA has greatly increased during
the last years [43], and some authors support its
role as the sole preoperative study for intracranial
aneurysms [42]. Technical advances including
increased field strength (from 1.5 to 3-T) and
refinement of techniques such as time of flight
(TOF) have improved the diagnostic value of
MRA over time [44]. Although this modality is
not appropriate for the acute setting, it is widely
utilized in the diagnosis and characterization of
unruptured aneurysms and is the method of choice
in patients with absolute contraindications (preg-
nancy, anaphylactic reactions, etc.) to contrast
administration.
Surgical Principles
The goal of surgical treatment of intracranial
aneurysms is complete obliteration with minimal
brain manipulation, while preserving flow in
25 Surgical Treatment of Aneurysms
Fig. 1 (a) Selective right internal carotid artery (ICA)
digital subtraction angiography injection, oblique projec-
tion, showing a complex aneurysm arising from the takeoff
of the anterior choroidal artery (arrows). (b) 3-dimensional
reconstruction of the digital subtraction angiogram show-
ing the irregular shape of the aneurysm. (c) Unsubtracted
right ICA angiogram, oblique projection, showing an
Fig. 2 (a) 3D reconstruction of computed tomography
angiogram in a patient with a small, 2.6 mm, posterior
inferior cerebellar artery aneurysm (arrows), showing a
detailed profile of the aneurysm. (b) Rotational
539
aneurysm remnant after clipping (arrows). (d) Selective
right ICA injection shows the postoperative remnant. At
surgery, it was found that the origin of the anterior choroi-
dal artery was duplicated (arrows), and therefore a small
remnant was intentionally left to protect the more distal
duplicated anterior choroidal artery
manipulation shows a different projection of the aneurysm.
Here, the relationship of the neck of the aneurysm and the
origin of the posterior inferior cerebellar artery is well
identified (arrows)
540
parent artery and associated perforator vessels.
The success of the surgical procedure depends
on a variety of factors, including appropriate
anesthetic management and an organized
surgical technique. All of the procedures,
regardless of the location, involve certain critical
steps that allow adequate exposure and
obliteration of the aneurysm. Further principles
have to be taken into account depending on the
particular anatomic location. Once the
aneurysm has been surgically treated, there are
tools available in the intraoperative setting to con-
firm proper obliteration of the aneurysm and
ensure patency of the parent artery and perforator
branches.
Anesthetic Management
Appropriate anesthetic management plays a very
important role in the surgical treatment of intra-
cranial aneurysms. The baseline clinical condition
of the patient and the rupture status of the aneu-
rysm often dictate the anesthetic plan. While early
surgical treatment is the best measure to reduce
the risk of rebleeding in patients with aneurysmal
subarachnoid hemorrhage, tight control of hemo-
dynamics and certain precautions must be
implemented to minimize the risk of
periprocedural rupture. The transmural pressure
gradient should be minimized while an adequate
cerebral perfusion pressure should be maintained.
Appropriate prophylactic measures should be
adopted to prevent excessive hypertension that
can be present during procedures such as tracheal
intubation or surgical stimuli such as skin inci-
sions or pinning for the craniotomy. During the
surgical procedure, pharmacologic measures can
be established in order to increase brain
relaxation, minimize brain retraction, and facili-
tate dissection. Osmotic diuresis with mannitol
reduces the brain tissue volume, while changes
in the ventilatory parameters such as controlled
hyperventilation can achieve mild hypocapnia
and reduced cerebral blood volume to provide
additional brain relaxation. These latter maneu-
vers are especially important in patients with
acutely ruptured aneurysms who often have a
S.F. Morales-Valero et al.
tight and edematous brain. Additional measures
have to be considered when a prolonged
temporary clipping is necessary, such as pharma-
cologic metabolic suppression. The role of
moderate hypothermia to extend the duration of
tolerable occlusion was felt to be protective as it
theoretically reduces the release of excitotoxic
amino acids. However, a randomized trial
failed to show an effect of mild intraoperative
hypothermia on neurologic outcome in patients
with favorable clinical grade subarachnoid
hemorrhage [45].
Critical Steps in Aneurysm Surgery
(Table 1)
Table 1 Critical steps for the surgical treatment of intra-
cranial aneurysms
1. Subarachnoid Opening of cisterns and arachnoid
dissection planes during initial exposure can
help drain CSF
Dissection of arachnoid fissures
exposes proximal parent artery
and allows proximal vascular
control
Sharp dissection along artery frees
aneurysm neck with focus on
avoiding damage to arterial
perforators
Trans-sylvian veins should be
preserved in order to limit venous
congestion and edema
2. Minimize CSF drainage can help brain
brain retraction relaxation via opening of basal
cisterns, fenestration of lamina
terminalis, or placement of external
ventricular drain in patients with
acute hemorrhage from aneurysm
rupture
Osmotic diuretics and steroid
administration minimizes cerebral
edema
Suction device should be used for
applying gentle retraction and
countertraction during aneurysm
dissection
Avoid excessive parenchymal
retraction to decrease risk of
ischemic injury or avulsion of an
adherent aneurysm
(continued)
25 Surgical Treatment of Aneurysms
Table 1 (continued)
3. Bony exposure Advanced skull base approaches
and modification of traditional
approaches remove bony structures
(e.g., sphenoid ridge, anterior and
posterior clinoid, etc.) to extend the
available operative corridor and
minimize brain retraction that is
required
4. Vascular Avoidance of dome manipulation
control and careful exposure of afferent and
efferent arteries are important to
establish proximal and distal
control for temporary clipping
5. Temporary Temporary clipping can soften the
clipping aneurysm and reduce risk of rupture
during aneurysm manipulation
6. Permanent Clip selection depends on size and
clipping direction of the aneurysm dome and
neck
Select proper blade length to avoid
poor visibility and inadvertent
occlusion of perforating vessels or
nearby cranial nerves
Booster clips can be utilized in
atherosclerotic calcified aneurysms
7. Clip inspection Clip blades should be placed
completely across neck and parallel
to branch vessel or bifurcations to
avoid small remnants
1. Subarachnoid dissection (Fig. 3a–c): Intra-
cranial aneurysms are commonly located
along the circle of Willis within the subarach-
noid space, so the surgical treatment is entirely
extraparenchymal. By opening the arachnoidal
planes and respecting the pia-arachnoid cover-
ing the brain, there is no violation of the cortex.
Progressive opening of the subarachnoid cis-
terns en route to the aneurysm allows CSF
drainage, and a wide opening of the surgical
corridor allows mobilization of the brain with-
out transmitting traction or undue forces on the
surrounding structures. The dissection usually
follows the path of a major artery with special
caution to avoid damage of small perforators.
2. Brain retraction: Proper view of the surgical
corridor occasionally requires some degree of
brain retraction. However, this should be
applied in a delicate fashion to avoid local
ischemia or damage to neurovascular
541
structures. The suction device can be used to
apply gentle retraction as the dissection pro-
gresses as well as to provide countertraction
when cutting tissue (Fig. 3a, c). If fixed retrac-
tors are necessary, these should be placed only
after thorough dissection to maximize brain
relaxation and minimize retraction pressure.
Drainage of cerebrospinal fluid by opening
the basal cisterns, fenestrating the lamina
terminalis, lowering an existing external ven-
tricular drain (EVD), or placement of a new
EVD intraoperatively (in ruptured cases) are
all acceptable maneuvers. In addition, osmotic
diuretics, as previously described, can be given
to decrease the brain tissue volume, and ste-
roids can be administered to minimize edema
from retraction.
3. Maximizing exposure through bone resec-
tion: Bone removal might be necessary to
maximize the exposure and decrease brain
manipulation, for example, the drilling of the
sphenoid ridge for pterional approaches. A
variety of advanced skull base approaches
have been used in the treatment of intracranial
aneurysms. While these are not necessary in
the majority of cases, they continue to have a
role in selected patients with aneurysms at
challenging locations such as the mid-basilar
trunk.
4. Vascular control (Fig. 3d–f): The afferent
and efferent arteries that supply anterograde
or retrograde blood flow to an aneurysm
should be exposed in case temporary
clipping is necessary. Proximal control
should be established as early as possible, and
the most appropriate point is defined according
to the location of the aneurysm; the most
complete and preferable is that located adja-
cent to the aneurysm. Avoidance of the
aneurysm dome while establishing proximal
and distal control is critical in order to
prevent intraoperative rupture. If vascular con-
trol cannot be established, a careful dissection
focused on the aneurysm neck should be made,
and the dome should be avoided to prevent
rupture.
5. Temporary clipping (Fig. 3d, e): Temporary
clips are often placed to facilitate aneurysm
542
Fig. 3 Artist depiction of the basic steps of microsurgery
for intracranial aneurysms. (a) The arachnoid surrounding
the aneurysm and its parent artery is widely opened. (b)
The proximal portion of the feeding vessel is exposed in
order to obtain proximal control. (c) The aneurysm neck is
circumferentially separated from surrounding tissues and
carefully inspected prior to clipping. (d) Temporary clip-
ping can be used to “soften” the aneurysm and facilitate
definitive clipping. In addition, presence of the temporary
clip protects from massive bleeding in the case of
intraoperative aneurysm rupture. When applying the tem-
porary clip, the utmost care is exercised to avoid any
compromise of small perforating vessels. (e) The perma-
nent clip is applied while the temporary clip is in place; the
S.F. Morales-Valero et al.
manipulation and to optimize visualization of
its anatomy, including small perforating arter-
ies and efferent vessels. In the case of complex
aneurysms, temporary clipping is especially
helpful for softening of the neck to allow for
more aggressive permanent clipping. In a
minority of cases, such as thrombosed or pre-
viously coiled aneurysms, the use of both prox-
imal and distal temporary clipping allows for
complete trapping of the segment harboring
the aneurysm.
6. Permanent clipping (Fig. 3e, f): Permanent
clipping is certainly the most important and
critical part of the surgical treatment of intra-
cranial aneurysms. The goal is to achieve com-
plete occlusion of the aneurysm while
preserving adequate flow through the parent
vessel and small perforating branches. There-
fore, complete visualization of the aneurysm
and its surrounding structures is essential
before proceeding with the permanent clip
application (Fig. 3c). Small aneurysms with
narrow necks are often treated with a single
clip. However, large aneurysms with broad
necks and complex anatomy often require mul-
tiple clips in order to achieve complete
obliteration.
7. Clip inspection: Once the permanent clips are
in place, the aneurysm can be mobilized in order
to confirm the patency of the parent artery as
well as the efferent branches and perforating
vessels. In addition, complete aneurysm obliter-
ation can be confirmed with visual inspection
and by palpating the aneurysm sac looking for
the absence of pulsatile flow. Puncture of the
dome is often done at this stage to confirm
complete exclusion by the clip.
ä
Fig. 3 (continued) ideal placement of the clip blades will
completely cross the neck of the aneurysm while preserv-
ing patency of the parent artery. (f) The temporary clip is
removed and the blades of the permanent clip are inspected
to ensure that clip blades completely cross the aneurysm
neck without compromising the parent artery. It is impor-
tant to check for any small perforators that may have
inadvertently been included inside the clip blades
25 Surgical Treatment of Aneurysms
Intraoperative Assessment
of Aneurysm Occlusion
Visual inspection alone is often insufficient to
confirm proper clip placement. Therefore, various
techniques have been developed to assess aneu-
rysm occlusion as well as parent artery and perfo-
rating branch patency. In the intraoperative
setting, these techniques can be used alone or in
combination to further adjust surgical clips.
Adjustments are undertaken to avoid vessel ste-
nosis, occlusion, and subsequent cerebral ische-
mia or aneurysm remnant, which may predispose
to aneurysm recurrence and rupture.
Intraoperative Angiography
Intraoperative digital subtraction angiography
(DSA) is considered by many as the gold standard
for assessment of aneurysm occlusion and vessel
patency. The rate of clip readjustment after
intraoperative angiography is highly variable and
ranges between 8 % and 34 %. Images in different
projections can be obtained for comparison with
the preoperative studies and to ensure proper clip
placement. Despite its invasive nature, the com-
plication rate is minimal in experienced hands,
and many authors advocate the routine use of
intraoperative angiography in all cases of aneu-
rysm clipping. However, it is not widely available,
requires the presence of trained personnel during
the surgical procedure, and is time-consuming.
Moreover, some small perforators are beyond its
resolution and cannot be visualized. With the
availability of newer techniques of intraoperative
imaging to verify proper clip placement, this tech-
nique is now used in selected cases, particularly
when treating large and complex aneurysms.
Microvascular Doppler
Ultrasonography
Micro-Doppler ultrasonography is a noninvasive
method often used during intracranial aneurysm
surgery for the assessment of parent vessel
patency and complete aneurysm occlusion. This
543
method is inexpensive, easily performed, does not
significantly prolong the operative time, and can
be used repeatedly if clip adjustments are
required. Although partial aneurysm occlusion
can be easily detected due to persistent flow in
the aneurysmal dome, the residual neck of a par-
tially obliterated or thrombosed aneurysm cannot
be detected with this modality because of the lack
of flow in the dome.
Indocyanine Green Fluorescent
Videoangiography
Near-infrared indocyanine green (ICG) fluores-
cent videoangiography is a relatively new method
for intraoperative assessment of blood flow during
aneurysm surgery. ICG is a near-infrared fluores-
cent compound that is injected intravenously and
binds to plasma proteins to reach the cerebral
vasculature within 30 s. A near-infrared light
source and a near-infrared-sensitive video camera,
which are included in most modern surgical
microscopes, are necessary to visualize ICG in
the vasculature. This technique allows for real-
time visualization of aneurysm filling and patency
of efferent vessels or perforating branches within
the surgical field (Fig. 4). This method of
intraoperative assessment is especially useful for
evaluating patency of very small perforating
branches (Fig. 4c–e), as many of these are
beyond the resolution of standard DSA. Recent
studies have shown similar rates of clip
readjustment when comparing intraoperative
ICG videoangiography to intraoperative DSA,
supporting the routine use of ICG
videoangiography while DSA is utilized only in
selected cases.
Surgical Approaches
The goal of aneurysm surgery is to attain occlu-
sion of the aneurysm while preserving flow in the
associated vasculature. The majority of aneu-
rysms involve proximal arterial branches at or
close to the base of the brain. A handful of surgical
approaches along with some of their variations are
544
Fig. 4 (a) Intraoperative high-magnification view of an
irregular anterior communicating artery aneurysm after
surgical exposure, with areas of lobulations (arrows) indi-
cating possible areas of weakness in the aneurysm wall. (b)
ICG fluorescent videoangiography view showing the aneu-
rysm prior to clipping. This image is obtained to establish
the baseline against which post-clipping ICG fluorescent
videoangiography can be compared. (c) Baseline ICG
fluorescent videoangiography showing presence of a
utilized in the treatment of these aneurysms: the
pterional (and its numerous variations), the
subtemporal, the interhemispheric, and the
retrosigmoid/far lateral. A detailed description of
these approaches and their pros and cons is
beyond the scope of this chapter.
Pterional Craniotomy
The pterional craniotomy, also known as the
frontotemporal craniotomy, is the most commonly
used approach in aneurysm surgery because of its
S.F. Morales-Valero et al.
small perforating vessel (arrows) that is incorporated in
the distal neck of the aneurysm. (d) Intraoperative view of
aneurysm after multiple-clip reconstruction of this com-
plex aneurysm, with sparing of the associated small perfo-
rating vessel (arrows). (e) Post-clipping ICG fluorescent
videoangiography shows no evidence of the previously
visualized aneurysm, while patent flow is preserved
through the small perforating vessel (arrows)
versatility and access to most anterior circulation
aneurysms and many posterior circulation aneu-
rysms. Not only is it the preferred exposure for
most aneurysms, but it is also widely utilized to
treat other vascular and nonvascular lesions of the
sellar/parasellar region, the orbital roof and fis-
sure, the cavernous sinus, as well as the frontal
and anterior temporal regions [46]. The patient is
positioned supine with the head fixed in three-
point fixation, slightly rotated to the contralateral
side, and extended so that the ipsilateral zygoma
becomes the highest point of the patient’s head.
This “classic” position maximizes the effects of
25 Surgical Treatment of Aneurysms 545

gravity in allowing the brain to “fall down” after
bony opening, thus minimizing the need for
mechanical retraction. The skin incision is curvi-
linear, starting 1 cm anterior to the tragus begin-
ning at the level of the zygomatic arch, and arcs
forward behind the hairline.
Over the years, numerous modifications of the
“classic” pterional approach have been described
and applied to the treatment of intracranial aneu-
rysms. With advances in skull base techniques,
some skull base-specific approaches have been
proposed to treat intracranial aneurysms. Of
these, the orbitozygomatic craniotomy has gained
traction in the past as a variation of the pterional
approach. With this approach, complete or
partial removal of the frontal or temporal
portion of the zygomatic arch provides the
required exposure to minimize the need for brain
retraction. On the other hand, the impetus for less
invasive techniques in recent years has led to a
progressive reduction of the size of the craniot-
omy for the treatment of most anterior circulation
aneurysms. These include development of the
“key-hole” approach and the lateral supraorbital
approach. There are a variety of different
approaches for the treatment of most anterior cir-
culation aneurysms from which to choose from,
and the selection of one modification of the
pterional approach over another is often related
to surgeon preference, degree of comfort with a
specific approach, and aneurysm characteristics or
position.
Other Surgical Approaches
for the Treatment of Aneurysms
Approaches other than the pterional and its mod-
ifications are utilized for aneurysms in specific
locations [47, 48]. These approaches and their
respective indications include the subtemporal
approach for some aneurysms of the upper basilar
artery and the proximal posterior cerebral artery,
the interhemispheric approach for aneurysms of
the pericallosal artery, and the retrosigmoid and
the far lateral craniotomy for aneurysms of the
lower basilar trunk and the origin of the posterior
inferior cerebellar artery (Table 2).
Table 2 Craniotomy selection for intracranial aneurysms
by location
Surgical approach Aneurysm location
Pterional approach and ICA, anterior
its modifications communicating artery
(AComm), middle cerebral
artery, proximal anterior
cerebral artery, posterior
cerebral artery, basilar
bifurcation, superior
cerebellar artery
Interhemispheric Distal anterior cerebral
approach artery, pericallosal artery,
callosal marginal artery
Far lateral approach/ Vertebral artery,
retrosigmoid and vertebrobasilar junction,
suboccipital approach mid-basilar artery, anterior
inferior cerebellar artery,
posterior inferior cerebellar
artery
Petrosal approach Mid-basilar artery, anterior
inferior cerebellar artery
Postoperative Care
Following craniotomy and clipping of aneurysms,
patients are typically admitted to the neurological
intensive care unit. In the absence of complica-
tions, patients are awakened and extubated in the
operating room. In the presence of immediate
postoperative deficits, an immediate CT scan
should be performed. If imaging studies show
evidence of vessel compromise responsible for
these deficits, return to the operating room may
be considered. However, this is exceedingly rare
with the available tools utilized intraoperatively
nowadays.
Patients are usually observed in the neurolog-
ical intensive care unit overnight, though the
patients with ruptured aneurysms are observed
for a longer period of time because of the risk of
SAH-related complications such as vasospasm
and hydrocephalus. With modern neurosurgical
techniques and high-power microscopes, surgery
for intracranial aneurysms is usually done within
the subarachnoid space, which minimizes the
amount of brain manipulation that is required.
With careful preservation of venous structures,
postoperative seizures are uncommon. Therefore,
546
the routine use of anticonvulsant medications is
not recommended. Anticonvulsant medications
delay recovery due to the high incidence of side
effects such as fatigue and drowsiness seen with
even the last-generation agents. If there are no
postoperative complications, intravenous fluids
are discontinued the day after the operation, and
the diet is advanced as tolerated. Similarly, Foley
catheter and arterial lines are disconnected early
and the patient is rapidly mobilized.
Complications
Complications related to surgical treatment of
intracranial aneurysms are uncommon due to the
use of advanced microsurgical techniques in com-
bination with intraoperative tools such as electro-
physiologic monitoring, microvascular Doppler
ultrasound, ICG videoangiography, and
intraoperative catheter angiography in select
cases. However, symptomatic epidural and sub-
dural hematomas can be seen postoperatively in
less than 1 % of patients, and ischemic complica-
tions are still a risk despite use of these advanced
intraoperative tools and microsurgical techniques.
Ischemic complications are often related to
manipulation of very small perforating vessels
during the dissection of the aneurysm or their
compromise during clip application. This occurs
more frequently in areas rich in small perforating
vessels, where the aneurysm sac is often adherent
to or in close relationship with vessels like the
basilar bifurcation or the anterior communicating
artery complex (Fig. 4d).
With advanced microsurgical techniques, post-
operative seizures are an uncommon occurrence,
but when encountered, an urgent CT scan should
be obtained to rule out the possibility of an intra-
cranial hematoma. They can also be the conse-
quence of a subpial dissection and compromise of
the cortical mantle. If the patient suffers a postop-
erative seizure, anticonvulsant medications are
started. In the absence of repeated seizures and
after an EEG rules out any evidence of an active
seizure focus, the anticonvulsants are
S.F. Morales-Valero et al.
discontinued after 6 months to 1 year. An aneu-
rysm remnant can occasionally be seen after clip-
ping of an aneurysm. Sometimes, the remnant is
left intentionally in order to decrease the risk of
compromise to the parent artery (Fig. 1d). At other
times, the remnant can be an unexpected finding
on postoperative imaging studies. In general,
small remnants after clipping of an aneurysm
have a benign natural history and can be followed
over time, especially if their correction is associ-
ated with a high risk of potential complications.
Follow-Up of Surgically Treated
Aneurysms
We usually obtain an immediate follow-up vascu-
lar study to document completeness of aneurysm
exclusion and to rule out complications. Although
this was traditionally done using catheter angiog-
raphy, CTA with 3D reconstruction has been
increasingly relied upon in recent years for post-
operative imaging. With modern software, arti-
facts related to the clips can be minimized, and
good correlation with postoperative angiography
has been shown. Long-term follow-up of surgi-
cally treated aneurysm has established surgery as
a durable treatment modality with an exceedingly
low incidence of recurrence after adequate clip-
ping. Patients who presented with subarachnoid
hemorrhage have a higher cumulative incidence
of recurrent hemorrhage when treated with
coiling as compared to clipped aneurysms
(2.6 % vs. 0.4 %). However, recurrent
hemorrhage can be also related to de novo aneu-
rysms [1, 49, 50]. Follow-up imaging every
5 years can be considered in patients who
presented with SAH in order to rule out de novo
aneurysm formation, although no study has defin-
itively shown the validity of this approach in
preventing future SAH. In the presence of a
known remnant, follow-up imaging study with
CTA should be performed approximately 1 year
after the original surgery. In the presence of a
stable remnant, further follow-up can be consid-
ered every 2–3 years thereafter.
25 Surgical Treatment of Aneurysms
Summary
In conclusion, despite rapid advances in
endovascular techniques, traditional surgical clip
ligation continues to be an important therapeutic
option for patients with ruptured and unruptured
aneurysms. Surgical clipping provides higher
degree of complete aneurysm exclusion when
compared to endovascular techniques and carries
a lower risk of aneurysm recurrence over time.
Strict adherence to microsurgical principles focus-
ing on minimizing brain manipulation, extensive
arachnoid dissection, and utilization of
intraoperative adjuncts are the key to low compli-
cation rates in aneurysm surgery. Noninvasive
vascular imaging studies such as MRA and CTA
play an ever-increasing role in the preoperative
and postoperative assessment of patients with
intracranial aneurysms undergoing surgical
treatment.
References
1. Rinkel GJ, Algra A (2011) Long-term outcomes of
patients with aneurysmal subarachnoid haemorrhage.
Lancet Neurol 10:349–356
2. Hackett ML, Anderson CS (2000) Health outcomes
1 year after subarachnoid hemorrhage: an international
population-based study. The Australian Cooperative
Research on Subarachnoid Hemorrhage Study Group.
Neurology 55:658–662
3. Hop JW, Rinkel GJ, Algra A, van Gijn J (1997) Case-
fatality rates and functional outcome after subarach-
noid hemorrhage: a systematic review. Stroke
28:660–664
4. Suarez JI, Tarr RW, Selman WR (2006) Aneurysmal
subarachnoid hemorrhage. N Engl J Med 354:387–396
5. Haley EC Jr, Kassell NF, Torner JC (1992) The inter-
national cooperative study on the timing of aneurysm
surgery. The North American experience. Stroke
23:205–214
6. Lovelock CE, Rinkel GJ, Rothwell PM (2010) Time
trends in outcome of subarachnoid hemorrhage:
population-based study and systematic review. Neurol-
ogy 74:1494–1501
7. Nieuwkamp DJ, Setz LE, Algra A, Linn FH, de Rooij
NK, Rinkel GJ (2009) Changes in case fatality of
aneurysmal subarachnoid haemorrhage over time,
according to age, sex, and region: a meta-analysis.
Lancet Neurol 8:635–642
547
8. Nishioka H (1966) Results of the treatment of intracra-
nial aneurysms by occlusion of the carotid artery in the
neck. J Neurosurg 25:660–704
9. Poppen JL, Fager CA (1960) Intracranial aneurysms.
Results of surgical treatment. J Neurosurg 17:283–296
10. Jha AN, Butler P, Lye RH, Fawcitt RA (1986) Carotid
ligation: what happens in the long term? J Neurol
Neurosurg Psychiatry 49:893–898
11. Drake CG, Peerless SJ, Ferguson GG (1994) Hunterian
proximal arterial occlusion for giant aneurysms of the
carotid circulation. J Neurosurg 81:656–665
12. Hopkins LN, Grand W (1979) Extracranial-intracranial
arterial bypass in the treatment of aneurysms of the
carotid and middle cerebral arteries. Neurosurgery
5:21–31
13. Fox AJ, Vinuela F, Pelz DM, Peerless SJ, Ferguson
GG, Drake CG et al (1987) Use of detachable
balloons for proximal artery occlusion in the treatment
of unclippable cerebral aneurysms. J Neurosurg
66:40–46
14. Higashida RT, Halbach VV, Dowd C, Barnwell SL,
Dormandy B, Bell J et al (1990) Endovascular detach-
able balloon embolization therapy of cavernous carotid
artery aneurysms: results in 87 cases. J Neurosurg
72:857–863
15. Elhammady MS, Wolfe SQ, Farhat H, Ali Aziz-Sultan-
M, Heros RC (2010) Carotid artery sacrifice for
unclippable and uncoilable aneurysms: endovascular
occlusion vs common carotid artery ligation. Neuro-
surgery 67:1431–1436, discussion 1437
16. Serbinenko FA (1974) Balloon catheterization and
occlusion of major cerebral vessels. J Neurosurg
41:125–145
17. Roski RA, Spetzler RF, Nulsen FE (1981) Late com-
plications of carotid ligation in the treatment of intra-
cranial aneurysms. J Neurosurg 54:583–587
18. Fujiwara S, Fujii K, Fukui M (1993) De novo aneu-
rysm formation and aneurysm growth following thera-
peutic carotid occlusion for intracranial internal carotid
artery (ica) aneurysms. Acta Neurochir (Wien)
120:20–25
19. Larson JJ, Tew JM Jr, Tomsick TA, van Loveren HR
(1995) Treatment of aneurysms of the internal carotid
artery by intravascular balloon occlusion: long-term
follow-up of 58 patients. Neurosurgery 36:26–30,
discussion 30
20. Matouk CC, Kaderali Z, terBrugge KG, Willinsky RA
(2012) Long-term clinical and imaging follow-up of
complex intracranial aneurysms treated by
endovascular parent vessel occlusion. AJNR Am J
Neuroradiol 33:1991–1997
21. Arambepola PK, McEvoy SD, Bulsara KR (2010) De
novo aneurysm formation after carotid artery occlusion
for cerebral aneurysms. Skull Base 20:405–408
22. Javedan SP, Deshmukh VR, Spetzler RF, Zabramski
JM (2001) The role of cerebral revascularization in
patients with intracranial aneurysms. Neurosurg Clin
N Am 12:541–555, viii
548
23. Sughrue ME, Saloner D, Rayz VL, Lawton MT (2011)
Giant intracranial aneurysms: evolution of manage-
ment in a contemporary surgical series. Neurosurgery
69:1261–1270, discussion 1270–1261
24. Lawton MT, Spetzler RF (1999) Surgical strategies for
giant intracranial aneurysms. Acta Neurochir Suppl
72:141–156
25. Leake CB, Brinjikji W, Kallmes DF, Cloft HJ (2011)
Increasing treatment of ruptured cerebral aneurysms at
high-volume centers in the united states. J Neurosurg
115:1179–1183
26. Barker FG 2nd, Amin-Hanjani S, Butler WE, Ogilvy
CS, Carter BS (2003) In-hospital mortality and mor-
bidity after surgical treatment of unruptured intracra-
nial aneurysms in the united states, 1996–2000: the
effect of hospital and surgeon volume. Neurosurgery
52:995–1007, discussion 1007–1009
27. Rabinstein AA, Lanzino G, Wijdicks EF (2010)
Multidisciplinary management and emerging therapeu-
tic strategies in aneurysmal subarachnoid
haemorrhage. Lancet Neurol 9:504–519
28. Bederson JB, Connolly ES Jr, Batjer HH, Dacey RG,
Dion JE, Diringer MN et al (2009) Guidelines for the
management of aneurysmal subarachnoid hemorrhage:
a statement for healthcare professionals from a special
writing group of the stroke council, american heart
association. Stroke 40:994–1025
29. Jane JA, Kassell NF, Torner JC, Winn HR (1985) The
natural history of aneurysms and arteriovenous
malformations. J Neurosurg 62:321–323
30. Etminan N, Beseoglu K, Barrow DL, Bederson J,
Brown RD Jr, Connolly ES Jr et al (2014) Multidis-
ciplinary consensus on assessment of unruptured intra-
cranial aneurysms: proposal of an international
research group. Stroke J Cereb Circ 45:1523–1530
31. Koivisto T, Vanninen R, Hurskainen H, Saari T,
Hernesniemi J, Vapalahti M (2000) Outcomes of
early endovascular versus surgical treatment of rup-
tured cerebral aneurysms. A prospective randomized
study. Stroke J Cereb Circ 31:2369–2377
32. McDougall CG, Spetzler RF, Zabramski JM, Partovi S,
Hills NK, Nakaji P et al (2012) The barrow ruptured
aneurysm trial. J Neurosurg 116:135–144
33. Molyneux A, Kerr R, Stratton I, Sandercock P,
Clarke M, Shrimpton J et al (2002) International sub-
arachnoid aneurysm trial (isat) of neurosurgical clip-
ping versus endovascular coiling in 2143 patients with
ruptured intracranial aneurysms: a randomised trial.
Lancet 360:1267–1274
34. Lanzino G, Murad MH, d’Urso PI, Rabinstein AA
(2013) Coil embolization versus clipping for ruptured
intracranial aneurysms: a meta-analysis of prospective
controlled published studies. AJNR Am J Neuroradiol
34:1764–1768
35. Mitchell P, Kerr R, Mendelow AD, Molyneux A
(2008) Could late rebleeding overturn the superiority
of cranial aneurysm coil embolization over clip ligation
S.F. Morales-Valero et al.
seen in the international subarachnoid aneurysm trial? J
Neurosurg 108:437–442
36. Lanzino G, Kanaan Y, Perrini P, Dayoub H, Fraser K
(2005) Emerging concepts in the treatment of intracra-
nial aneurysms: stents, coated coils, and liquid embolic
agents. Neurosurgery 57:449–459, discussion
449–459
37. D’Urso PI, Lanzino G, Cloft HJ, Kallmes DF (2011)
Flow diversion for intracranial aneurysms: a review.
Stroke J Cereb Circ 42:2363–2368
38. Hacein-Bey L, Provenzale JM (2011) Current imaging
assessment and treatment of intracranial aneurysms.
AJR Am J Roentgenol 196:32–44
39. Connolly ES Jr, Rabinstein AA, Carhuapoma JR,
Derdeyn CP, Dion J, Higashida RT et al (2012)
Guidelines for the management of aneurysmal sub-
arachnoid hemorrhage: a guideline for healthcare pro-
fessionals from the American Heart Association/
American Stroke Association. Stroke J Cereb Circ
43:1711–1737
40. Fifi JT, Meyers PM, Lavine SD, Cox V, Silverberg L,
Mangla S et al (2009) Complications of modern
diagnostic cerebral angiography in an
academic medical center. J Vasc Interv Radiol: JVIR
20:442–447
41. van Rooij WJ, Sprengers ME, de Gast AN, Peluso JP,
Sluzewski M (2008) 3d rotational angiography: the
new gold standard in the detection of additional intra-
cranial aneurysms. AJNR Am J Neuroradiol
29:976–979
42. Westerlaan HE, van Dijk JM, Jansen-van der Weide
MC, de Groot JC, Groen RJ, Mooij JJ et al (2011)
Intracranial aneurysms in patients with subarachnoid
hemorrhage: Ct angiography as a primary examination
tool for diagnosis–systematic review and meta-
analysis. Radiology 258:134–145
43. McDonald JS, Kallmes DF, Lanzino G, Cloft HJ
(2013) Use of ct angiography and digital
subtraction angiography in patients with
ruptured cerebral aneurysm: evaluation of a large
multihospital data base. AJNR Am J Neuroradiol
34:1774–1777
44. Sailer AM, Wagemans BA, Nelemans PJ, de Graaf R,
van Zwam WH (2014) Diagnosing intracranial aneu-
rysms with mr angiography: systematic review and
meta-analysis. Stroke J Cereb Circ 45:119–126
45. Todd MM, Hindman BJ, Clarke WR, Torner JC (2005)
Mild intraoperative hypothermia during surgery for
intracranial aneurysm. N Engl J Med 352:135–145
46. Yasargil MG, Fox JL (1975) The microsurgical
approach to intracranial aneurysms. Surg Neurol
3:7–14
47. Davies JM, Lawton MT (2014) Advances in open
microsurgery for cerebral aneurysms. Neurosurgery
74(Suppl 1):S7–S16
48. Park J, Hamm IS (2004) Anterior interhemispheric
approach for distal anterior cerebral artery aneurysm
25 Surgical Treatment of Aneurysms
surgery: preoperative analysis of the venous anatomy
can help to avoid venous infarction. Acta Neurochir
146:973–977, discussion 977
49. Molyneux AJ, Kerr RS, Birks J, Ramzi N, Yarnold J,
Sneade M et al (2009) Risk of recurrent subarachnoid
haemorrhage, death, or dependence and standardised
mortality ratios after clipping or coiling of an intracra-
nial aneurysm in the international subarachnoid
549
aneurysm trial (isat): long-term follow-up. Lancet
Neurol 8:427–433
50. Schaafsma JD, Sprengers ME, van Rooij WJ,
Sluzewski M, Majoie CB, Wermer MJ et al (2009)
Long-term recurrent subarachnoid hemorrhage
after adequate coiling versus clipping of ruptured
intracranial aneurysms. Stroke J Cereb Circ
40:1758–1763
 Endovascular Treatment of Brain
Aneurysms 26
Luca Quilici and Edoardo Boccardi

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 552 Brain aneurysms are the most frequent cause of
Myths, Faith, and a Little Bit of Science . . . . . . . . . . . . 552 intracranial hemorrhage in young people. In
Stone Age (Balloon and Parent Artery
the last 40 years the treatment of brain aneu-
Occlusion) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 553 rysm has been revolutionized by the advent
Procedure (Tips and Tricks) . . . . . . . . . . . . . . . . . . . . . . . . . 553 and the growth of endovascular techniques.
Iron Age (Coiling and Balloon Remodeling) . . . . . 554 The introduction of detachable balloons by
Procedure (Tips and Tricks) . . . . . . . . . . . . . . . . . . . . . . . . . 557 Serbinenko in the mid-1970s to occlude rup-
Renaissance (Stent-Assisted Coiling) . . . . . . . . . . . . . . 563
tured or symptomatic aneurysms was the first
Procedure (Tips and Tricks) . . . . . . . . . . . . . . . . . . . . . . . . . 568 step of the neurovascular progress that goes on
today. In the early 1990s detachable coils with
Modern Times (Flow Diverters) . . . . . . . . . . . . . . . . . . . 571
Procedure Tips and Tricks . . . . . . . . . . . . . . . . . . . . . . . . . . . 577 the aid of the balloon remodeling technique
before and of self-expanding stents later were
The Promises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579
the following steps of this path. Nowadays
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579 these techniques are well established, safe,
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579 effective, and usually used all over the world,
helping to secure ruptured aneurysms and to
prevent bleeding by unruptured lesions. The
recent advent of flow-diverter stents represents
the latest improvement, pretending to obtain in
most of the cases not only the exclusion of the
aneurysm from the flowing blood but also its
complete disappearance.

Keywords
Aneurysm • Blister-like aneurysm • Brain hem-
orrhage • Fusiform • Partially thrombosed
aneurysm • Clipping • Detachable balloon •
Coil • GDC • Balloon remodeling • Stent •
Flow diverter • Flow disruptors • Resorbable
L. Quilici (*) • E. Boccardi
stent
Ospedale Niguarda – Ca’ Granda, Milan, Italy
e-mail: Luca.quilici@ospedaleniguarda.it; Edoardo.
boccardi@ospedaleniguarda.it

# Springer Science+Business Media New York 2016 551

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_6
552
Introduction
Myths, Faith, and a Little Bit of Science
What is an aneurysm? Providing an accurate and
complete definition of an aneurysm is not easy.
The official definition of an aneurysm is a focal
dilation of the wall of an artery [1]. However, this
is a very simplistic and unsatisfactory definition.
What should the cutoff size of the dilation of the
wall be, to define it an aneurysm? And how should
we differentiate an aneurysm, for example, from a
normal variation in caliber of a section of an
artery? Moreover, this definition of “aneurysm”
would exclude some vascular lesions, like blister
aneurysms, which are a particular type of dis-
sected lesion often invisible on angiographic
images.
Even if a perfect definition and classification of
aneurysms does not exist, sharing a common lan-
guage and universal definitions is fundamental to
communicate thoughts, ideas, and activities.
Cerebral aneurysms are certainly where the
field of neurovascular disease focuses its greatest
attention and interest, likely due to the fact that
they are pretty common and probably the most
common cause of brain hemorrhages in younger
people – frequently a dramatic, often fatal, event.
This is probably the reason why – over the past
few decades and especially since the concept of
preventive medicine began to assert itself – the
medical community has created the aura of the
imminent threat of a ticking time bomb around the
idea of a brain aneurysm, even when unruptured
and asymptomatic. For the majority of doctors,
specialists or not, every aneurysm is a possible
source of hemorrhage, but the fears of the medical
community should come to terms with reality.
Every day in the world, tens of thousands of
neuroradiological exams are performed with
various indications and – as a result of these
exams – thousands of new, totally asymptomatic,
aneurysms are incidentally identified. The more
diagnostic techniques progress, the more in-depth
and detailed tests are carried out, with new,
smaller and smaller, aneurysms being diagnosed.
Recent estimates suggest that, globally, roughly
L. Quilici and E. Boccardi
one person every 12–15 is the bearer of a brain
aneurysm. This observation would lead to an
obvious consequence, to estimate that 6–8 % of
the world’s population, over the course of their
lives, would be affected by a brain hemorrhage
sooner or later. However, the scientific biblio-
graphic data show that the rate of brain hemor-
rhage from a ruptured aneurysm is constant: 5–10/
100,000 individuals per year in Western countries
(in Japan and Finland, 20/100,000) [2–6]. It is
obvious that only a very small percentage of the
aneurysms that are diagnosed each day will be
responsible for a brain hemorrhage. So the syllo-
gism, unruptured aneurysm = time bomb, is most
likely a myth of modern medicine, with no scien-
tific basis. Hundreds of millions of people cur-
rently live their lives as bearers of a brain
aneurysm, but only a few thousand of them will
be affected by a brain hemorrhage from that diag-
nosed aneurysm during their lifetimes.
The purpose of the treatment of an aneurysm is
the prevention of bleeding. In the most common
cases the presentation is already a rupture with
bleeding, so the goal is to prevent further bleed-
ings. Historically neurosurgery has found a very
good solution to exclude aneurysms from the cir-
culation, by using clips which would compress the
aneurysmal sac and remodel the vessel to a normal
anatomy.
Nowadays, endovascular treatment is in many
places the therapeutic option of choice in most
cases of brain aneurysms. The evolution of the
endovascular techniques, since its birth 40 years
ago, has been overwhelming. History, more than
science, has shown that the endovascular
approach, compared to the neurosurgical
approach, is almost always safer, simpler, and
cheaper [7]. Probably the intrinsic characteristics
of a more simple, physiologically respectful, min-
imally invasive approach have led to the gradual
expansion and success of endovascular therapy
over the past four decades. The evolution of mate-
rials and techniques has provided a further boost
to the remarkable progress of the method. Even if
in the late 1970s it was only possible to occlude
the parent vessel of an aneurysm with homemade
balloons, nowadays, with the flow diverters, it is
26 Endovascular Treatment of Brain Aneurysms
possible to occlude the aneurysm, and obtain the
definitive disappearance of the lesion.
Stone Age (Balloon and Parent Artery
Occlusion)
The beginning of the “endovascular age” is dat-
able to the mid-1970s [8, 9]. Until that time,
before the discovery of CT and MRI, catheter
angiography was only one of the few diagnostic
tools useful in identifying vascular lesions or
other space-occupying lesions that conditioned
displacement of the brain vessels. At that time
the treatment of a brain aneurysm was exclusively
a surgeon affair. The surgeon, in the case of a
ruptured aneurysm or a symptomatic one, decided
whether, how, and when to place one or more
clips, on the basis of angiographic images and
the direct exploration of the lesion. In the most
complicated cases, when clipping was impossible
for the site, or for the size of the sac, the aneurysm
would be wrapped, a technique with limited effi-
cacy. Then it came the time of detachable bal-
loons. These devices allowed the exclusion of
arteriovenous fistulas, especially the carotid-
cavernous type, the occlusion of vessels, and in
some cases the filling of aneurysms. The idea of
the detachable balloon catheter was conceived by
a Russian neurosurgeon, Fedor Serbinenko. He
developed the idea and produced the balloons in
his small workshop. The balloons were made in
latex, of various sizes, and were mounted on cath-
eters days prior to the procedure. His idea came
from observing a child playing with a balloon
inflated with helium during the May Day parade
on Red Square in 1959. In 1969, Serbinenko
successfully performed the first endovascular
treatment of a carotid-cavernous fistula. Over the
following decade, he executed 162 endovascular
treatments of arterial occlusion therapy, mostly in
inoperable brain aneurysms. He published his
experience in a Russian scientific journal in 1971
[10]. Notoriety for him and his invention arrived
in 1974 with the publication of his work in a
Western scientific journal (Journal of Neurosur-
gery) [11] (Fig. 1).
553
The idea and the technique of detachable bal-
loons became widespread and had a significant
impact on intervention strategy for direct arterio-
venous fistulas and brain aneurysms. Starting
from there, therapeutic endovascular artery occlu-
sion has become routine all over the world, in all
cases of inoperable aneurysms [12, 13]. Even
today, the choice of occluding the parent vessel
remains a valuable and safe option in cases of
extremely complex lesions, such as giant aneu-
rysms of the internal carotid or dissected aneu-
rysms (Fig. 2).
Procedure (Tips and Tricks)
The procedure of vessel occlusion has to be pre-
viously confirmed by an occlusion test. The test
allows the interventionist to know whether it is
possible to occlude the target vessel. A rough test
could be done with a manual compression at the
neck of the vessel one plans to occlude while
carrying out angiographic runs of the other ves-
sels. To obtain more complete information on
collaterals and safety, it is necessary to perform a
balloon test occlusion (BTO).
The BTO is done with the patient awake and
under systemic anticoagulation with heparin
(80 IU/kg). Two femoral sheaths must be placed:
one for the guiding catheter and the other one for
angiographic control from the other vessels dur-
ing the occlusion test. A complete angiographic
study of the cerebral circulation must always be
obtained before starting the procedure. The
angiographer has to carefully evaluate the func-
tion of the communicating arteries and the pres-
ence of anastomosis, particularly among the
branches of the external carotid artery and internal
carotid or vertebral artery. The second step is to
perform BTO. The BTO is realized by placing an
inflatable balloon into the target vessel under fluo-
roscopic guidance at the point where one wants to
occlude the artery. The point of occlusion should
be chosen very carefully, in light of the previous
angiographic study, to avoid occluding the artery
in the wrong location, i.e., a location where the
occlusion may leave a persistent filling of the
554
Fig. 1 Fedor Serbinenko, the Russian neurosurgeon that conceived the idea of detachable balloons (a) and developed the
technique of their use. Historical picture of latex detachable balloons (b). The mechanism of detachment of balloons (c)
aneurysm sac through anastomosis. The balloon is
inflated up to the point of occluding the lumen of
the vessel; the occlusion is maintained for at least
15–30 min during which time, through angio-
graphic runs; the intra- and extracranial collater-
alization is evaluated. Neuropsychological tests
are performed to assess the possible occurrence
of a neurological deficit. It is always possible to
perform stress tests, provoking a drug-controlled
hypotension, in such a way as to highlight sub-
liminal neurological deficit.
The balloon test occlusion has two results: an
angiographic result and a clinical one. The angio-
graphic test is passed only when the vein
opacification of the territory downstream from
the occluded artery is contemporary to the rest of
the brain up to, but not more than, two seconds
later. The clinical test is passed if the occlusion
does not provoke any neurological deficits. If the
occlusion test is not tolerated, the balloon is
deflated, the procedure is discontinued, and one
may consider the realization of a surgical bypass.
If the test occlusion is instead tolerated, one can
L. Quilici and E. Boccardi
finally occlude the artery, by detaching the bal-
loon (in case of a detachable balloon) or by
substituting the balloon with coils or a different
occlusion device.
Arterial occlusion is simple and very effective,
but it is certainly not riskless [14, 15]. Detachable
balloons are not available anymore, mostly due to
commercial reasons. For this reason in recent
years, therapeutic occlusion is almost always
performed using coils (Figs. 3, 4, and 5).
Iron Age (Coiling and Balloon
Remodeling)
In the early 1990s a new device changed forever
the treatment of cerebral aneurysms: the electrical
detachable coil invented by Guido Guglielmi
(Guglielmi detachable coil). Coils (metallic fila-
ments that would curl up on themselves) were
already being used for filling aneurysms, but the
problem was that by pushing them through a
microcatheter it was almost impossible to control
26 Endovascular Treatment of Brain Aneurysms
Fig. 2 Preparing the detachable balloon (Goldbal2; Balt)
(a–e). A GBV2 mounted on the needle of a 2 cc syringe
and a Magic 1.2 Fr catheter (Balt) (a). Inflation test of the
balloon. The syringe is prepared with a solution of ½ saline
their correct final position, with the risk of losing
them in distal arteries (Fig. 6).
Guglielmi achieved the goal of controlling the
detachment of the coils from the pusher wire. One
could finally detach the coil only once there was
the evidence of a correct positioning. The detach-
ment is realized with an electrolytic mechanism.
The aneurysm is then filled through the release of
several consecutive coils, obtaining a small ball of
thread. Initially the idea had been to cause the
thrombosis of the aneurysm sac by coagulating
the blood through an electric current necessary to
detach the coil inside the aneurysm. It was soon
evident that that approach would not work and
555
and ½ contrast medium (b). Testing the continence of the
GBV (c). The GBV is inflated and mounted on the tip of the
Magic catheter with the help of a Stylet wire (d, e)
that it was necessary to fill the sac with filaments
rather than causing thrombosis [16, 17].
The first GDCs made their appearance on the
world stage in 1991 and very quickly demon-
strated their effectiveness. The first coils had a
filament diameter of 0.010 in., and subsequently
coils of 0.015 in. were produced, making
more rapid filling of large aneurysms
possible; the 0.015 in. versions were later
replaced by those measuring 0.018 in. Platinum
was chosen because it is a metal that does not
undergo electrolysis and because it is spontane-
ously radiopaque, as well as being thrombogenic
and inert.
556
Fig. 3 Images obtained in a 50-year-old with a giant
cavernous aneurysm symptomatic for headache.
Endovascular treatment by vessel occlusion. Frontal view
of the left internal carotid injection (ICA) (a). The occlu-
sion of left ICA by 2 gold valve balloon (Balt) (b). Frontal
views of the right ICA and the left vertebral (VA) injections
The detachment of the coil takes place by the
closing of an electric circuit powered by a battery,
in which the positive pole is coupled to the tail of
the coil, placed outside the microcatheter, and the
negative pole is coupled to a needle inserted under
the skin of the patient. The passage of a
low-voltage current causes the electrolysis of the
micro-welding causing the detachment of the plat-
inum filament. The newest detachment systems
do not need needle insertion anymore.
GDCs were initially used for the treatment of
ruptured aneurysms which neurosurgeons consid-
ered technically impossible – or at least very dif-
ficult – to address. It could be because of their
location (e.g., the tip of the basilar) or of their
characteristics (Fig. 7). The selected patients
were mostly elderly or in poor clinical condition
[18]. The range of measurements and spatial con-
formations of the initially available GDCs,
L. Quilici and E. Boccardi
showing the collateralizations through communicating
arteries and through pial collaterals (c, d). T2-weighted
brain MR images of the aneurysm before (e) and after
2 years from the treatment (f). Note the complete shrinking
of the aneurysm
defined as the diameter of a single loop of the
coil and the length of the entire filament, was
very limited. Nevertheless, the idea had an imme-
diate extraordinary success. Endovascular treat-
ment with coils showed clear advantages
compared to neurosurgical clipping: in many “dif-
ficult” cases the treatment was relatively easy, fast
and, especially, less aggressive than a neurosurgi-
cal intervention, at the same time ensuring a more-
than-satisfactory result in excluding the aneu-
rysm. At the beginning of the coiling experience,
there was no scientific proof of the truth of these
observations, but the efficacy of the method was
simply evident [19–22]. Over time different com-
panies provided a more comprehensive range of
sizes of coils with different degrees of stiffness
(normal, soft, ultrasoft, etc.) and with different
spatial conformations of the loops (2D, 3D, com-
plex, etc.).
26 Endovascular Treatment of Brain Aneurysms
Fig. 4 Images obtained in a 68-year-old man suffering
from right CN III palsy caused by a partially thrombosed
cavernous aneurysm. A combined surgical and
endovascular treatment: bypass and artery occlusion. Fron-
tal view of right ICA injection before the treatments (a).
Frontal view of right common carotid (CCA) injection
Procedure (Tips and Tricks)
The procedure is performed under general
anesthesia to ensure immobility of the
patient and to allow the catheterization to be as
safe as possible. Systemic anticoagulation is
557
after the creation of a high-flow bypass between right
CCA and ipsilateral MCA (b). Note the initial slow
downing of the ICA. Occlusion of right ICA with balloons
(*, c). Frontal view of bypass injection with excellent
revascularization of the right hemisphere and reflow in
the supraclinoid ICA and in the ophthalmic artery (d)
usually required, and some authors would
consider also antiplatelet therapy. A good
practice in neurointerventions is that all catheters
– such as the guiding catheter and the
microcatheter – must be constantly perfused
using heparinized saline.
558
Fig. 5 Endovascular treatment by artery occlusion with
coils after repeated subarachnoid bleedings in a 63-year-
old lady. Oblique view of the left VA injection (a) showing
an intracranial dissection of the VA. Note the absence of
It is highly recommended to start the procedure
by performing an accurate angiographic study
with 3D reconstructions of the aneurysm, in
order to define the working projections that should
be used during the intervention. The projections
must ensure the best possible view of the aneu-
rysmal neck, namely, the plane passing between
the aneurysm and parent vessel. The working
projection should also take account of the pres-
ence of other vessels, especially if they originate
in the vicinity of the sac. The accurate study of 3D
and 2D images is crucial in allowing the operator
to build a mental image of the aneurysm and its
surroundings. In addition, the 3D helps in choos-
ing both the measures of the first coil and the
curve that the tip of the microcatheter should
have. The tip of the microcatheter might come
pre-curved in the package or it might be prepared
using steam, depending on the characteristics of
each single aneurysm. It is also very important to
bend the tip of the microwire.
The choice of the size and shape of the coils for
the treatment is entirely operator dependent: the
goal being to place coils in order to obtain a
L. Quilici and E. Boccardi
branches rising from the dissected tract. Oblique views of
left and right VAs showing the complete occlusion of the
left artery by coils and good supply of the posterior circu-
lation by the right one (b, c)
complete and compact filling of the aneurysm.
The common practice is to start with a coil of a
diameter corresponding to the maximal diameter
of the aneurysm and then fill the interstices with
progressively smaller coils. It is advisable to use a
Y connector with rotating hemostatic valve for
both the guiding catheter and the microcatheter
to ensure maximum stability of the system. The
microcatheter with the microwire is tracked inside
the guiding catheter and under fluoroscopic vision
pushed close to the aneurysmal neck.
The entrance of the wire and entrance of the
catheter into the aneurysm are two key moments,
especially in the case of ruptured aneurysms. The
maneuver should therefore be performed with
great care, avoiding touching the walls of the
aneurysm if possible. Once the aneurysm is
catheterized, the microwire should be withdrawn
and replaced with the first coil. The release of the
coil must be very gentle, a smooth movement,
continuous, progressive, and slow. In order to
allow the adaptation of the coil to the walls of
the aneurysm, creating a so-called basket. The
release of the first coil is crucial: if well realized,
26 Endovascular Treatment of Brain Aneurysms
Fig. 6 Guido Guglielmi (a, b) and his revolutionary coils
it allows easier filling of the basket by the subse-
quent coils. The loop protrusion outside the aneu-
rysmal neck in the parent artery is an occurrence
that should obviously be avoided due to the
chance that this will stimulate platelet aggregation
with subsequent embolism risk. After the release
of the first coil, coils with a progressively smaller
diameter are usually chosen. The filling of the sac
should be continued until the coil release becomes
too difficult.
Along with the advantages of this technique,
its limitations have also quickly emerged. The
wider the aneurysmal neck is, the more difficult
the treatment of an aneurysm with coils is. The
most frequent issues with coiling have been ische-
mic complications determined by the formation of
559
blood clots, along catheters or on the coils placed
at the level of the aneurysmal neck. The risk of
thromboembolic events is directly related to the
length of the procedure. Simple and fast
procedures have major advantages in terms of
reducing complication rates. The ischemic
complications are best taken care of with a good
technical approach, clean sheets and catheters,
and heparin injections. Only rarely one has to try
to dissolve some major clot in cerebral
vessels. The use of fibrinolytics is forbidden,
because of high chances of having a rebleed
from the aneurysm. Glycoprotein IIb/IIIa inhibi-
tors may be of help in some cases, but often the
clot especially if limited in size will dissolve
spontaneously.
560
Fig. 7 Images from one of the oldest cases (first in Italy)
of endovascular treatment with GDC coils in 1993. A
34-year-old lady with a basilar tip ruptured aneurysm.
Hemorrhages caused by the rupture of aneu-
rysms during catheterization of the sac or during
the release of coils are less frequent. In this
unlikely event, the only thing to do is to deploy
coils inside the aneurysm as quickly as possible
until the bleeding stops (Figs. 8 and 9).
Another complication of this type of treatment
is stretching of the coil: by pushing and pulling
repeatedly the platinum filament may stretch to
the point of rupture. At that point placing the coil
into the aneurysm or withdrawing it could be
impossible. Coil stretching was more frequent in
the past. Now, with the introduction of the stretch-
resistant mechanism, constituted by a further fila-
ment which links the first and the last loop of the
coil, stretching has become a more rare phenom-
enon. Stretching a coil is still possible though,
particularly in cases of long and difficult position-
ing with continuous advances and withdrawals of
the coil. If the event occurs stretched coils may be
retrieved using a retrieval system (goose neck or
similar). If unsuccessful one can stretch even the
proximal part of the coil down into the aorta, so
that the downward blood stream will keep it in
place.
Angiographic follow-up of the tens of thou-
sands of aneurysms treated with coils all over
L. Quilici and E. Boccardi
Frontal views of left vertebral injection (a–c). Before the
treatment, after the detachment of the first coil, and in the
5-year follow-up
the world has highlighted another important lim-
itation of this type of treatment: the tendency of
coiled aneurysms toward recanalization. Recana-
lization is variable from aneurysm to aneurysm,
partly depending on the coil density at the end of
treatment. Recanalization is more frequent in
large and giant aneurysms and in wide-neck aneu-
rysmal necks, but it often occurs even in those
aneurysms which appear to be perfectly filled at
the end of treatment. The first cause of the phe-
nomenon has to be found in a spontaneous settling
and compaction of the coils toward the bottom of
the sac, under the effects of blood flow and blood
pressure. The degree of recanalization of the sac,
as we have said, is variable and is usually
highlighted in angiographic controls already at
6–12 months after treatment, and then, in most
cases, it stabilizes over a longer period of 2–5
years. The main consequence of this observation
has been that a high proportion of treated aneu-
rysms have undergone a second, a third, or addi-
tional endovascular treatments in pursuit of the
best possible compaction of the coils inside the
sac, increasing, for each treatment, the overall risk
of clinical and technical complications [22].
In 1994, a French neuroradiologist, Jacques
Moret, devised and put into practice a strategy
26 Endovascular Treatment of Brain Aneurysms
Fig. 8 (continued)
for limiting the problem of wide necks. The idea
was to place a catheter with an undetachable com-
pliant balloon in the parent vessel in front of the
aneurysmal neck and inflate it during the release
of the coils through a second catheter previously
561
positioned in the aneurysmal sac. The inflated
balloon allows an easier deployment of coils
preventing coil prolapse in the parent vessel.
Moreover, the presence of the inflated balloon
allows to place more coils in the sac, but most
562
Fig. 8 Endovascular treatment of a ruptured aneurysm of
the AComm. Images obtained in a 52-year-old lady who
suffers of a subarachnoid hemorrhage. Preliminary 3D
rotational angiography (a) and frontal view of the left
internal carotid injection showing the work projection.
AP views subtracted and un-subtracted showing the
importantly, in cases of a rupture of the aneurysm,
the balloon can immediately stop the bleeding.
Moret named this advance in endovascular
techniques the remodeling technique. This
improvement quickly proved its effectiveness
and allowed the treatment of aneurysms until
then considered extremely difficult [23].
Nowadays single-lumen balloon and double-
lumen balloon catheters are available. The balloon
is inflated and deflated by a small (1–2 ml) syringe
injection. The balloon is maintained inflated for
the time necessary for the deployment of each
coil, but usually for no longer than 5 min, in
order to avoid ischemic complications. The
L. Quilici and E. Boccardi
consecutive steps of the procedure (b–e). The positioning
of the microcatheter in the aneurysmal sac (b). Progressive
filling of the lesion with consecutive deployment and
detachment of four coils (c, d). Final result after removal
of the catheter (e)
procedure is obviously best performed in a condi-
tion of systemic heparinization in the majority of
cases.
Endovascular treatment of aneurysms of the
bifurcation became possible thanks to the contem-
porary positioning of two balloons: one balloon
for each of the branches of the bifurcation, in
order to protect both of them.
Thanks to the remodeling technique the treat-
ment of aneurysms with coils had a further exten-
sion of its indications and, during the 1990s,
became quikly the treatment of choice for all
aneurysm [24, 25]. All over the world surgery
was progressively replaced by the endovascular
26 Endovascular Treatment of Brain Aneurysms
Fig. 9 Intraprocedural rupture of an AComm aneurysm in
a 49-year-old lady. The tip of catheter punches the aneu-
rysm at the level of the neck (a, b). Immediate leakage of
contrast medium (b, c) in the subarachnoid spaces during
approach, and in some countries neurovascular
surgery has progressively declined almost to the
point of disappearance.
In 2002, The Lancet published the results of the
ISAT (International Subarachnoid Aneurysm
Trial) which compared the outcomes of the two
types of intervention in a population of over 2000
patients with subarachnoid hemorrhage from
ruptured aneurysm suitable for both endovascular
treatment or surgical clipping. ISAT compared
the rate of mortality and the degree of
disability at 2 months and at 1 year after
treatment. The final results showed a reduction
of 7.4 % in the absolute risk of death or depen-
dency at 1 year from the endovascular treatment
[26, 27]. The trial results certified what experience
and history had already shown: endovascular
treatment with coils is a safe and effective way
to treat aneurysms, and in most cases it is safer
than neurosurgical clipping. For ruptured aneu-
rysms it is the treatment of choice, thanks to its
simplicity and to its low degree of invasiveness
(Figs. 10 and 11).
563
the left ICA injection. Fast, consecutive deployment of
three coils with the arrest of bleeding (d, e) and good
exclusion of the aneurysm
Renaissance (Stent-Assisted Coiling)
Treatment with coils and the balloon remodeling
technique aroused great enthusiasm and interest,
and many interventionists tried new ways to
advance the techniques. From the mid-1990s
onward, some of them experimented sporadically
with intracranial stenting procedures – the idea
being that a stent might produce the cure of an
aneurysm by changing the flow inside the sac and
by being a scaffold for endothelial regrowth. They
used coronary balloon-mounted stents, but the
maneuvers were extremely difficult and cumber-
some due to the high rigidity of the systems
[28, 29]. The idea of an intracranial stent subse-
quently considered the possibility to better fill the
difficult wide-neck aneurysms, being a support to
the coils, in an even more effective way than did
the remodeling technique. The need of a stent
dedicated to the specificity of intracranial vessels
became apparent: the new stent should be softer
and easily trackable along curves and tortuosities
564
Fig. 10 The “balloon remodeling technique” of a rup-
tured aneurysm in an 81-year-old lady. Frontal view of
the right internal carotid injection before the treatment of
a wide-neck supraclinoid aneurysm (a). Road mapping
image of the first coil deployment protected by the balloon
and should be self-expandable, in order to avoid
possible ruptures of vessels by balloon inflation
(Fig. 12).
The first prototype of a self-expandable intra-
cranial stent, the Smart, appeared in 2001: it was
developed by Dr Peter Kim Nelson and his team at
Smart Therapeutics (San Leandro, California),
and renamed Neuroform in 2002 when the patent
was acquired by Boston Scientific and approved
by the US FDA [30, 31]. The Smart system was
composed of three parts, a self-expandable stent, a
microcatheter, and a stabilizer over the wire. The
L. Quilici and E. Boccardi
inflated (Scepter; MicroVention). Note the little adjustment
(tiny white stripes) of the loops caused by the inflation of
the balloon (b). The cast of coils and the position of both
catheter and balloon at the end of the filling (c). Final
angiogram of the right internal carotid (d)
stent, preloaded in the microcatheter, was an
open-cell stent, formed by radiolucent filaments
made of nitinol. The diameters of the devices
ranged from 3 to 4.5 mm, while the lengths ran
from 15 to 20 mm. At the two extremities of the
stent, four radiopaque markers ensured visibility
and controllability. The delivery catheter was 3 Fr
in outer diameter. The stabilizer was a second
microcatheter of smaller diameter (0.25 in.) with
a radiopaque marker at the distal end. The whole
system was navigated by an exchange 0.01400
wire, needed to reach the landing zone. Both the
26 Endovascular Treatment of Brain Aneurysms
Fig. 11 Images obtained in a 65-year-old lady with a
ruptured basilar tip aneurysm. The endovascular treatment
with coils and double balloon. Frontal views of left
catheter and the stabilizer were equipped with “Y”
valve connectors. The system was assembled on
the sterile surgical field by inserting the stabilizer
in the proximal end of the microcatheter
containing a preloaded stent. At the time of the
release of the stent, the distal end of the stabilizer
was placed next to the proximal end of the stent.
565
vertebral injections (a–d). The arrowheads indicate the
catheter (Echelon 10, Covidien-Ev3) and two balloons
(Transform, Striker), one in every PCA
Upon reaching the area of release, the
microcatheter was withdrawn, keeping the stent
in position thanks to the stabilizer; so doing, the
stent was released at the target point, across the
neck of the lesion. The Neuroform (Smart) was
followed by other competing devices: Leo (Balt,
Montmorency, France), Enterprise (Cordis,
566
Fig. 12 Oblique view of left internal carotid angiograms
in a 53-year-old lady suffering from headache caused by a
large supraclinoid aneurysm (a). An un-subtracted image
immediately after the deployment of a stent graft (arrow-
head); the balloon is still inside (b). Left ICA angiogram
Neurovascular J&J), Solitaire (Ev3, Covidien,
Paris, France), and LVIS (MicroVention, Tustin,
California, USA), characterized by different
design philosophies [32–35] (Fig. 13).
Intracranial stents used for assisted coiling are
tubular structures in cells, made of a metal alloy
with shape memory, typically nitinol or cobalt-
chromium. The stents are released into the parent
vessel of an aneurysm, across the aneurysm sac,
usually before filling the aneurysm with coils. The
self-expandable stents do not need a balloon to
L. Quilici and E. Boccardi
showing the initial exclusion of the aneurysm (c). Note the
positions of the guiding catheter and the tip of the exchange
wire, respectively, in the petrosal tract of ICA and in a
distal branch of MCA to guarantee enough support. Final
left ICA angiogram (d)
open, and hence need an intrinsic radial force,
produced in part by the material of which they
consist and in part by the geometry with which the
cells that constitute them are joined. The memory
metal alloys, at a predetermined temperature (usu-
ally 37  C), tend to shape themselves, expanding
and bending in space, allowing the stents to
achieve a preprogrammed diameter once
unconstrained by the catheter. It is mainly the
different types of construction – open cell
(Smart, Neuroform I, II, and III), braided stent
26 Endovascular Treatment of Brain Aneurysms
Fig. 13 Stent-assisted coiling of an aneurysm of the right
superior cerebellar artery (SCA) (a–c). Images obtained in
a 36-year-old man suffering from headache. The right
vertebral angiogram showing the origins of the right SCA
and the ipsilateral PCA, respectively, from the bottom of
the aneurysm and from the basilar artery, behind the sac
(Leo and LVIS), closed cell (Enterprise), or folded
sheet (Solitaire) – which influence the character-
istics of different stents: their radial force, their
capacity in adapting to vessel tortuosity, their
opening and closing resistances, and their ability
to retrieve the stent. Every stent has its unique
combination of these characteristics. The Solitaire
adds the specificity of being fully retrievable.
The rationale of using a stent in the treatment
of aneurysms is tied to three potential beneficial
effects:
567
(a). The presence of the stent (Neuroform, Boston S.)
avoids the protrusion of loops in the basilar artery and in
both PCAs. Coils have been placed away from the bottom
of the aneurysm in order to keep regularly patent the SCA
(b, c)
1. The stent should be a stable support frame on
which to create a compact cast of coil, resulting
in a reduced incidence of prolapsing coil loops
in the parent vessel and of recanalization of the
sac in the follow-up.
2. The stent should change the hemodynamic
blood flow in the vessel in which it is released,
especially at the level of the aneurysmal neck.
This change is related to the geometric modi-
fication that the stent imposes on the artery -
typically a distension and a straightening - as
568
well as to the presence of the stent meshes
which cut into and alter the flow along the
artery walls.
3. The stent should become integrated into the
wall of the arterial vessel, covered by a layer
of fibroblasts and an endothelium, to further
help to achieve the complete healing of the
disease.
The disadvantages are mainly related to an
increase in the complexity and technical difficulties
of the procedures, resulting in longer durations of
interventions. The increased complexity is due to
the number of steps to be performed (first stenting,
then coiling, or vice versa), as well as the need to
use microcatheters for releasing stents which have
larger caliber than those used for coils. This
increases rigidity and decreases navigability of
the system. With the first generations of stents,
precisely because of the reduced navigability of
the proper microcatheters, it was necessary to per-
form an exchange with a “normal” catheter and a
300 cm long 0.01400 wire at the beginning of the
procedure. Another disadvantage is that stent
implantation also requires the establishment of an
effective antiplatelet therapy to prevent in-stent
platelet aggregation, responsible of increased risk
for ischemic embolic events or for the occlusion of
the parent vessel. However, the antiplatelet therapy
carries of course its own hemorrhagic risks.
Procedure (Tips and Tricks)
The introduction of self-expanding intracranial
stents has led operators to redesign the treatment
of aneurysms in two separate phases: the place-
ment of the stent and the filling of the sac with
coils. Usually the positioning of the stent precedes
the release of coils inside the aneurysm. In this
case, the microcatheter is placed inside the aneu-
rysm after the deployment of the stent by passing
the catheter through the meshes of the stent. In
other cases, the catheter can be placed into the
aneurysm before the deployment of the stent
(jailing technique). The technique is especially
useful for very large wide-neck aneurysms, but it
has been used also in simpler cases. Once a stent is
L. Quilici and E. Boccardi
placed, the filling of the aneurysm sac with coils is
usually easier and faster and the coil packing in
the aneurysm can be denser. It might be tricky
though, because coil filaments may become
stuck in the meshes, especially in long and diffi-
cult procedures.
The advent of stent-assisted coiling techniques
has allowed a more effective treatment of wide-
neck aneurysms as well as of those lesions located
at artery bifurcations: the basilar tip, the bifurca-
tion of the middle cerebral artery (MCA), the
termination of the carotid artery, and the anterior
communicating artery (AComm). As with the
remodeling technique, and even more so, in
these cases the positioning of a stent in the most
appropriate branch of the bifurcation determines
the constitution of a new artificial aneurysmal
neck most favorable to the coiling. Thanks to the
experience gained with balloons and with the
remodeling technique, very soon it was devised,
in cases of bifurcation aneurysm, to put two stents,
crossed or side by side.
Operators came up with different spatial con-
figurations of stents. The most commonly used is
the “Y stenting”, widely utilized for basilar tip
aneurysms. A stent is placed from the posterior
cerebral artery (PCA) to the basilar artery; the
second stent is placed symmetrically from the
other PCA to the basilar artery. The second device
is released by passing through the meshes of the
first stent at the level of the aneurysmal neck. The
proximal ends of both stents are placed one inside
the other in the basilar trunk (Fig. 14).
One more possible configuration is the X
stenting: used for aneurysms of the AComm. A
stent is released from the A2 tract of the anterior
cerebral artery (ACA) to the A1 tract of the con-
tralateral ACA. The second device is positioned
symmetrically, opening it through the mesh of the
first stent. T, 1/2-T, and Z stenting and more are
very rarely used.
The procedure of intracranial stenting is
performed, as are all intracranial procedures,
under general anesthesia in order to minimize
the movements of the patient. The positioning of
a stent requires a pharmacological preparation of
the patient with antiplatelet drugs. The prepara-
tion typically requires at least 4–5 days of intake
26 Endovascular Treatment of Brain Aneurysms
Fig. 14 Oblique views of successive treatment steps in a
Y stenting and coiling of a basilar aneurysm in a 46-year-
old man (a–d). Un-subtracted images of progressive
to be effective. The drugs used most commonly in
many labs are clopidogrel (75 mg daily) and
acetylsalicylic acid (ASA, 81–325 mg daily).
The procedure, as already stated, should be
always be preceded by a careful angiographic
study and a three-dimensional representation of
the aneurysm and of the parent vessel with its
569
deployment of a Neuroform stent (Boston S.) in the left
PCA (b), an Enterprise stent (Codman, J&J) in the right
PCA (c), and coils (d)
branches. The angiographic and 3D studies
allow the choice of working projections for the
placement of the stent and release of coils. The
projections for the two phases of treatment often
do not coincide. For the choice of working pro-
jection for coiling, apply the suggestions given in
the previous section. The working projections for
570
the release of the stent should depict in the best
possible way the vessel receiving the stent and
also its perforating branches with respect to the
aneurysm. This will allow you to identify the best
area of opening, landing, and deployment of the
stent. The preliminary angiographic and 3D
recostructions will also be helpful in choosing
the most appropriate sizes of stent and coil. The
measurements (diameter and length) of the stent
will have to take into account the diameters of the
receiving vessel, but one must also consider the
changes in length that each stent undergoes
depending on its opening diameter. It is important
to remember that for braided stents (Leo, LVIS,
Leo Baby), the smaller the diameter of the receiv-
ing vessel is when compared with the diameter of
the unconstrained stent, the greater the elongation
of the stent within the vessel will be. Open- and
closed-cell stents do not change in length, but we
shall find out soon that this concept will be very
important for another type of device, the flow
diverters.
The guiding catheter must have a diameter of at
least 6 Fr. The delivering microcatheter inner
diameter is 0.027" for an “old generation” stent,
and can now be 0.021" or 0.017". in the cases of
the LVIS Jr. and Leo Baby. Both microcatheters
can navigate with a 0.014 wire. As for all
neurovascular treatment, the guiding catheter
and the microcatheter should be perfused with
heparinized saline (4000–5000 units per liter). It
is advisable to use the silicone on-off Y valves, to
obtain an optimal management of the tensions and
pressures in the releasing phase of the stent. It is
better to push the delivering catheter into the
target vessel distal to the initial point of planned
stent release. This will give more space to maneu-
ver in the opening of the stent and in the apposing
to the walls of the artery. In the majority of stent
models, the release of the stent is achieved by
withdrawing the microcatheter while maintaining
the position of the stent pusher.
The technique of stent-assisted coiling for the
treatment of cerebral aneurysms has definitely
been another milestone along the way toward
affirmation of endovascular techniques. It has
made the treatment of very large wide-neck aneu-
rysms possible and has enabled a reduction in the
L. Quilici and E. Boccardi
incidence and extent of recanalization of aneu-
rysms compared with the treatment of aneurysms
with bare coils. On the other hand, the new tech-
nique has brought with it an increase both in the
incidence of periprocedural ischemic complica-
tions and in their severity, because of the presence
of more catheters and material in cerebral arteries
for a longer time. In addition, thrombotic occlu-
sions of the artery occurred more frequently, due
to ineffective antiplatelet therapy [36, 37].
However, the appearance of intracranial stents
has also represented a new opportunity for treat-
ment of a particular group of aneurysms: those
fusiform, partially thrombosed. This type of aneu-
rysm is very often localized at the level of the
vertebrobasilar circulation and involves long seg-
ments of these arteries. These lesions are rarely
the cause of bleeding in the brain. In most cases
they behave like a neoplasm: they grow slowly
over years to an advanced state, exerting a pro-
gressively worsening mass effect on the brainstem
and cranial nerves. Not infrequently, the diagnosis
of these aneurysms is by the onset of an obstruc-
tive hydrocephalus due to compression of the
cerebral aqueduct. Before the introduction of
stents the only treatment option available, if pos-
sible, was vessel occlusion by balloon or coils.
Some neurointerventionists started to try the use
of self-expandable stents in this kind of aneu-
rysms. The rationale for the use of stents is not
so much to help the optimal filling of the aneu-
rysm sac by the coils, but rather to be a true
intravascular endoprosthesis. The stent should
divert the flow in the lumen of the diseased vessel
and constitute the skeleton on which to favor the
creation of a new wall inside of the vessel itself, so
that over time the natural diseased wall of the
vessel is progressively excluded from circulation.
With this idea in mind, and having no other effec-
tive treatment options, over the last decade there
have been many treatments of these aneurysms
using, in most cases, the Leo stent (Balt). The
braided construction makes the Leo meshes very
dense, so as to have some effect on blood flow. To
cover the segment of pathological arteries, often
involving both vertebral and the basilar arteries,
Leo stents are released one inside the other, par-
tially overlapping, to form long chains
26 Endovascular Treatment of Brain Aneurysms
(telescoping technique). Unfortunately, this tech-
nique has rarely been successful in these aneu-
rysms, and the results of arrested growth of these
aneurysms or even their healing are anecdotal
[38, 39]. However, this technique has provided
the ideological and technical support on which
to base the latest evolution of the technique of
endovascular treatment of aneurysms by flow
diversion (Fig. 15).
Modern Times (Flow Diverters)
In 2007 the neurovascular world changed again,
as dramatically as it changed in 1991 with the
advent of GDCs. In 2007, over a short period of
time, two new intracranial devices became avail-
able, both the result of a new idea, a new theory in
the treatment of aneurysms: the Silk (Balt, Mont-
morency, France) and the Pipeline
(Ev3-Covidien, Paris, France). The first two flow
diverters (FDs), built by hand by weaving 48 thin
strands of metal alloy – nitinol for the Silk and
cobalt-chromium for the Pipeline – to form tubes
with very high density meshes and very low
porosity. Released in front of the neck of the
aneurysm, flow diverters redirect the flow in the
parent vessel, away from the aneurysm, causing
its exclusion from circulation. Over the following
weeks and months, this causes progressive and
spontaneous intra-aneurysmal thrombosis and
then the formation of a clot that subsequently
retracts, leading to deflation of the sac until the
aneurysm has totally shrunk.
The impact of these devices on the
endovascular treatment of aneurysms has been
truly revolutionary. With the flow diverters, it is
no longer necessary, in the great majority of cases,
to care or worry about the aneurysm sac. The
possibility of rupturing the aneurysm during
endovascular treatment no longer exists or nearly
so. The possibility of coils dislodging during or
after endovascular treatment no longer exists.
Aneurysm recanalization no longer exists or
nearly so.
It’s a totally different vision of endovascular
treatment of aneurysms – no more unsuccessful
filling of a sac. We have finally realized the self-
571
healing of the sac by changing the hemodynamic
environment around it. This tool allows us to
obtain the definitive cure of an aneurysm by its
disappearance, for the first time in the history of
endovascular treatment of aneurysms. The devel-
opment of flow diverters has also changed the
method of evaluating the success or failure of
treatment: with these devices we no longer talk
merely about percentages of filling or non-filling
of the aneurysm; mostly we discuss whether or not
the aneurysm has disappeared.
As with the advent of coils in 1991, the first flow
diverters have been implanted for the treatment of
unruptured aneurysms previously considered tech-
nically impossible and extremely difficult or char-
acterized by a high probability of relapse-
recanalization with traditional treatments: internal
carotid aneurysms of the petrosal segment, cavern-
ous aneurysms, or para-ophthalmic aneurysms,
large and giant, aneurysms without neck or with a
very large neck, aneurysms from which cerebral
arteries emerged, or recurrent aneurysms after pre-
vious endovascular treatments (Fig. 16). In most
cases it was sufficient to implant one FD for every
aneurysm, but in lesions without neck or with a
very large neck often more than one or two FDs
have been implanted. They have been deployed in
a partially overlapping fashion, one inside the other
to form a chain or a telescopic column.
With a small percentage of technical failures in
the delivery of devices, and with a rate of intra-
and periprocedural complications comparable to
those of the classical methods, the flow-diverter
devices showed a high or very high (up to more
than 90 %) rate of exclusion from circulation of
these aneurysms in the medium term (6 months to
1 year). Not only in the statistics that have taken
account of these issues over an identical follow-up
period, more than half (66 %) of these aneurysms
treated with flow diverters have disappeared,
while another good proportion of these have
decreased in size, mitigating or resolving the
mass effect on the brain and on cranial nerves
and allowing the bony structures eroded by the
presence of the aneurysm to grow back – all these
by keeping the parent vessel and the vessels that
emerged from the sac properly open [40–42]
(Fig. 17).
572
Fig. 15 A blister aneurysm treated with telescopic stents
technique in a 43-year-old lady after two subarachnoid
hemorrhages. The 3D rotational angiography and the left
ICA angiogram document the absence of collateral
branches rising from the neck of the aneurysm (a). Final
angiogram after the endovascular treatment (b). Note the
L. Quilici and E. Boccardi
deployment of the stents one inside the other (c).
Six-month follow-up angiogram with complete exclusion
of the aneurysm (d). The un-subtracted image documents
the straightening of the supraclinoid tract of ICA and of the
stents (e)
26 Endovascular Treatment of Brain Aneurysms
Fig. 16 Pretreatment (a, c) and follow-up (b, d) images
obtained in a 53-year-old lady with a large left VA aneu-
rysm symptomatic for CN VI palsy, treated using a flow-
diverter stent (PED, Covidien-Ev3). Left VA injection and
the 3D rotational angiograms showing the aneurysm (a).
The posterior inferior cerebellar artery (PICA) arises from
the lateral wall of the aneurysm. Left VA angiogram
Fig. 17 A giant cavernous right ICA aneurysm symptom-
atic for headache and CN III palsy treated with a Silk stent
(a, b, d). Images obtained in the 6-month follow-up (c, e).
573
obtained 6 months after endovascular treatment (b). Note
the normal patency of PICA that rises directly from left
VA. Axial T2-weighted MR images showing pretreatment
aneurysm impression on the pons (c) and complete shrink-
ing and healing of the aneurysm in the follow-up images at
1 year (d)
It is difficult to identify both in the angiogram and in the CT
scan the previous presence of the aneurysm
574
Fig. 18 Artery arrangement after endovascular treatment
of a right MCA large aneurysm with a stent flow diverter.
Pre-treatment (a) 6-month follow-up images (b) and an
enlarged image of follow-up (c). In the follow-up images
the superior branch of MCA and the ACA are thinner than
In the light of these incredible results, soon
flow-diverter implantation came to be performed
successfully even in “normal” aneurysms and in
“distal” aneurysms, placed on small arteries, the
AComm, the pericallosal artery, branches of the
MCA (Fig. 18), or distal tracts of the PCA,
confirming the excellent results already shown.
They proved their efficacy even in small aneu-
rysms, while certainly becoming the definitive
solution in cases of recanalization of those aneu-
rysms already treated with coils or with stent and
coils [43].
Soon FDs were implanted in dissected aneu-
rysms, like blister lesions of the carotid siphon,
responsible for recent subarachnoid hemorrhages.
Blister aneurysms are lesions that are very diffi-
cult to treat by clipping or by coiling because their
walls are extremely fragile, like a tiny sheet of
paper. In these very rare cases, usually one FD is
able to prevent the rebleeding and help to create a
new vessel wall and protect the patient [44].
L. Quilici and E. Boccardi
in the pretreatment, and the origin of ACA seems to be a
little detached from the ICA (ghost junction). After the
treatment most of perforators of MCA seem to rise from
the superior branch of MCA. All the artery arrangement
occurred asymptomatic
The FD has even been used for the treatment of
giant, fusiform, partially thrombosed aneurysms
of the basilar artery, but again in this type of
pathology, the successes are anecdotal [45, 46]
(Fig. 19).
As was the case for the Guglielmi detachable
coils, flow diverters have become, over the course
of just a few years, an essential tool in the
endovascular treatment of unruptured aneurysms
for nearly all the neuro-interventionists in the
world.
In the wake of this incredible success, compet-
itors have proposed flow-diverter devices with the
same philosophy, but with different technical
characteristics: Fred (MicroVention), Surpass
(Stryker), P64 (Phenox), and Derivo (Acandis).
The power of the FD is very strong. When
implanted, their hemodynamic influence is
exerted not only on the aneurysm but all over the
surrounding arteries. The FDs may decrease the
flow in the arterial branches in front of which they
26 Endovascular Treatment of Brain Aneurysms
Fig. 19 Images obtained in a 16-year-old girl with acute
onset of headache and right ischemic hemiplegia caused by
a dissected aneurysm of MCA involving the M1 segment
and the perforators branches. Axial T2 FLAIR images
showing the aneurysm (intramural hematoma) (a) and a
left lenticular ischemia (b). Frontal view of left ICA
575
angiograms before and immediately after the deployment
of a Pipeline stent (c, d). Note the dissected stenosis at the
neck of the aneurysm and the absence of the lateral perfo-
rators of M1. Left ICA angiogram at 1-year follow-up (e)
showing the normalization of diameter and course of MCA
and its branches
576
Fig. 20 Vascular plasticity after endovascular treatment
of a left vertebrobasilar junction aneurysm with a stent
FD. Oblique views of left VA angiograms obtained before
the stent deployment, at 6-month, at 1-year, and at 2-year
follow-up (a–d). The aneurysm is located on the larger arm
of a fenestration. The anteroinferior cerebellar (AICA)
are positioned, leading over time to a progressive
reduction in the caliber of these arteries also. The
reduction in size of the vessel can be up to the
point of angiographic disappearance, especially if
the artery has a small caliber (perforator arteries)
or has a small distribution area. The reduction of
size and flow occurs slowly and progressively; it
takes weeks or months and the changes are almost
always asymptomatic. During this time the other
arteries, not subjected to the effect of the stent,
increase in caliber and flow and take charge of the
vascular territories lost from the branches covered
by the stent. In the end, the flow diverters reshape
the local anatomy, changing the territorial distri-
bution of the different branches (Figs. 18 and 20).
The dark side of FDs is mainly represented by
the emergence of a new kind of complications,
never occurred before. One new kind of compli-
cation is the aneurysms rupture few weeks after
FD deployment. This type of hemorrhage is often
fatal or severely disabling. This complication
rarely happens and is more frequent with large or
giant aneurysms (in about 2–3 % of aneurysms
larger than 15 mm) in which a quick and massive
L. Quilici and E. Boccardi
artery before the treatment origins from the thinner fenes-
tration arm (a). The follow-up images show the progres-
sive cure of the aneurysm (star) and the changes of the
course and diameter (b, c, d) of the AICA artery (arrow-
head). All these modifications occurred totally
asymptomatic
intra-aneurysmal thrombosis of the aneurysm sac
excluded from circulation after stent placement
can cause rupture of the sac itself. In cases of
aneurysms larger than 15 mm, it is possible to
reduce the risks of this complication by adding
coiling of the sac, usually with a jailing technique.
The idea is that the coiling will make the throm-
bosis more gradual, slow, and fragmented by the
presence of the filaments. The second expedient is
also the administration of steroids for a few weeks
of post-treatment in order to moderate the inflam-
mation caused by the intra-aneurysmal
thrombosis [47].
Another new kind of complication appeared
with the development of flow-diverter experience:
intraparenchymal hemorrhage in locations far
from the treated aneurysm. This is a less frequent,
but certainly more insidious, complication that
usually develops within a few days after treat-
ment, with moderate clinical-neurological conse-
quences, rarely fatal, sometimes even
asymptomatic. The etiology is still unknown
although there has been a higher incidence in
cases where more than one FD has been released.
26 Endovascular Treatment of Brain Aneurysms
At first it was thought that the bleeding had the
same pathogenesis as that which occurs with tra-
ditional treatment (stents and coils), caused by the
perforations of arteries produced by wires during
the treatments. Yet in the case of FD treatments,
these hemorrhages often are far away from where
the microwires had been during the intervention.
One possible explanation is that the FD itself is the
problem. Once it is released, it changes the cere-
bral hemodynamic by diverting a large amount of
blood flow to brain areas which were not “pre-
pared” to receive blood in this amount. This
excessive perfusion caused by the FD would
lead to a kind of hemorrhagic infarction. Other
interventionists have speculated that, during the
release and the opening of the FD, an emboliza-
tion occurs via a shower of angiographically
invisible air microbubbles which had been previ-
ously trapped in the small meshes of the stent.
This shower of microbubbles could cause the
occlusion of very small vessels and an ischemia
with subsequent hemorrhagic infarction. Still
much about this complication remains
unknown [48].
Procedure Tips and Tricks
Endovascular treatment with flow diverters is a
simple procedure and in most cases quicker than
other procedures using conventional systems. All
needed is a coaxial system that consists of a 6 Fr
guiding catheter and a 0.02700 microcatheter with a
0.01400 microwire. The procedure requires phar-
macological preparation with dual antiplatelet
therapy. Specific technical maneuvers are rarely
necessary mainy in cases where the access road to
the parent vessel of the aneurysm is complicated
or tortuous. In these instances, it is useful or
necessary to switch to a triaxial system interpos-
ing an distal access catheter between the guiding
catheter and the microcatheter. The procedure, as
usual, must always be preceded by an accurate
angiographic study and a three-dimensional
reconstruction of the aneurysm and the parent
vessel. Even more than for conventional stent
procedures, the 3D study is needed to choose the
working projections and especially the
577
measurements of the FD. The working projections
should allow the best areas for opening and
releasing the device along the parent vessel, just
as in the procedures with conventional stents.
However, differently from these, they also allow
us to perfectly locate the origin of the arterial
branches that it is possible to cover or not with
the stent.
The 3D study is also useful for the choice of the
size (diameter and length) of the FD, a choice that
is certainly crucial for the success of the procedure,
much more crucial than for procedures with con-
ventional stents. A device with an overestimated
diameter with respect to the receiving vessel will
not open properly. The cell sizes of an oversized
device will remain too large, limiting or canceling
their effect of flow diversion. Moreover, not
reaching its nominal diameter, the device will be
much longer than expected in the vessel. Vice
versa, an undersized FD with respect to the receiv-
ing vessel will tend to open up more than its nom-
inal diameter; the device will not appose correctly
to the vessel walls, making the treatment ineffec-
tive; and finally it might easily displace. This will
also result in a considerable shortening of the stent.
These features greatly affect the phases of opening,
release, and implantation of the stent, which are not
linear as with conventional stents.
The deployment should always be done in no
hurry, taking time, slowly and carefully. The
opening and the start of the release of the FD
should begin on a section of the artery down-
stream of the desired point of landing of the
device. Once the distal end is opened and fixed
at the walls of the artery in the desired point, the
release of the remaining portion of the device
must be done in consecutive and alternating
steps. The point is that the FD has to be pushed
and shortened in order to be well open. Only then
the flow diversion effect will be maximal. There-
fore, it is necessary to carefully pull and push the
device so that subsequent portions of it will
appose correctly the vessel walls. A good deploy-
ment of an FD takes much longer than the deploy-
ment of a traditional stent: 5, 10, or even 15 min
versus 30 seconds. Notably, this phase, for an FD,
is much more critical than with a traditional stent.
If a portion of the FD is not well deployed, or is
 578

Fig. 21 Treatment of a basilar tip aneurysm with an endosaccular device WEB. Frontal and lateral views of left vertebral angiogram and 3D rotational angiography before the
treatment (a). Frontal and lateral views both un-subtracted (b) and subtracted (c) of left VA angiograms (b) after the deployment of device
L. Quilici and E. Boccardi
26 Endovascular Treatment of Brain Aneurysms
not completely opened, it is better to resheath it or
retrieve it and restart the procedure.
As coils represent the best solution for the
treatment of ruptured aneurysms because of their
simplicity, safety, and effectiveness in most cases,
the FD, in our opinion, equally represents the best
solution for unruptured aneurysms. As previously
stated, in most cases they represent a definitive
cure of the aneurysm.
The Promises
Exploiting FD technology, a new type of device
has been presented in recent years: intra-
aneurysmal flow disruptors, such as the WEB
(Sequent Medical) and the Luna ev3. These
devices are small bags that behave like a plug,
which has to be positioned inside a saccular aneu-
rysm. Their structure, made by filaments like an
FD, slows down the flowing of blood inside the
aneurysm and provokes the formation of a clot
with subsequent thrombosis of the aneurysm.
Flow disruptors follow the old philosophy that
the target of the treatment is the aneurysmal sac
and their goal is to thrombose the sac by staying
inside it. They are a promising technology, espe-
cially for ruptured aneurysms, and in the near
future they might be able to substitute coils in
emergency cases [49, 50].
On the FD side hopefully the next generation
of flow diverters will not need any antiplatelet
therapy. Or perhaps we will come up with a
resorbable flow diverter which could be easily
positioned for any aneurysm, ruptured or
unruptured [51]. The dream of course would be
the invention of a pill, which you take in the
morning and by night the aneurysm is gone
(Fig. 21).
Summary
Brain aneurysms are the most frequent cause of
intracranial hemorrhage in young people. In the
last 40 years the treatment of brain aneurysm has
been revolutionized by the advent and the growth
of endovascular techniques. The introduction of
579
detachable balloons by Serbinenko in the
mid-1970s to occlude ruptured or symptomatic
aneurysms was the first step of the neurovascular
progress that goes on today. In the early 1990s
detachable coils with the aid of the balloon
remodeling technique before and of self-
expanding stents later were the following steps
of this path. Nowadays these techniques are well
established, safe, effective, and usually used all
over the world, helping to secure ruptured aneu-
rysms and to prevent bleeding by unruptured
lesions. The recent advent of flow diverter stents
represents the latest improvement, pretending to
obtain in most of the cases not only the exclusion
of the aneurysm from the flowing blood but also
its complete disappearance.
References
1. Osborn AG, Salzman KL, Barkowich AJ, Katzman GL
et al (2010) Diagnostic imaging brain. Amirsys, Salt
Lake City
2. International Study of Unruptured Intracranial Aneu-
rysms Investigators (1998) Unruptured intracranial
aneurysms – risk of rupture and risks of surgical inter-
vention. N Engl J Med 339:1725–1733
3. Menghini VV, Brown RD Jr, Sicks JD et al (1998)
Incidence and prevalence of intracranial aneurysms
and hemorrhage in Olmsted County, Minnesota, 1965
to 1995. Neurology 51:405–411
4. Ronkainen A, Hernesesiemi J, Ryynanen M (1993)
Familial subarachnoid hemorrhage in east Finland
1977–1990. Neurosurgery 33:787–797
5. Juvela S, Porras M, Poussa K (2000) Natural history of
unruptured intracranial aneurysms: probability of and
risk factors for aneurysm rupture. J Neurosurg
93:379–387
6. Winn HR, Jane JA Sr, Taylor J et al (2002) Prevalence
of asymptomatic incidental aneurysms: review of 4568
arteriograms. J Neurosurg 96:43–49
7. Lad SP, Babu R, Rhee MS, Franklin RL et al (2013)
Long-term economic impact of coiling vs clipping for
unruptured intracranial aneurysms. J Neurosurg
72(6):1000–1011
8. Djindjian R, Picard L, Manelfe C, Merland JJ, Theron J
(1978) Development de la neuroradiologie
therapeutique. Neuroradiology 16:381–384
9. Wolpert SM (1974) In re: Serbinenko FA. Balloon
catheterization and occlusion of major cerebral vessels.
J Neurosurg 41:1974
10. Serbinenko FA (1971) Balloon occlusion of cavernous
portion of the carotid artery as a method of treating
carotid-cavernous fistulae. Zh Vopr Neirokhir 6:3–9
580
11. Serbinenko FA (1974) Balloon catheterization and
occlusion of major cerebral vessels. J Neurosurg
41(2):125–145
12. Debrun G, Lacour P, Caron J-P et al (1978) Detachable
balloon and calibrated-leak balloon techniques in the
treatment of cerebral vascular lesions. J Neurosurg
49:635–649
13. Fox AJ, Vinuela F, Pelz DM, Sydney I et al (1987) Use
of detachable balloons for proximal artery occlusion in
the treatment of unclippable cerebral aneurysms.
J Neurosurg 66:40–46
14. Langford KH, Vitek JJ, Zeiger E (1983) Migration of
detachable mini-balloon from the ICA causing occlu-
sion of the MCA. J Neurosurg 58:430–434
15. Swann KW, Heros RC, Debrun G et al (1986) Inadver-
tent middle cerebral artery embolism by a detachable
balloon: management by embolectomy. J Neurosurg
64(2):309–312
16. Guglielmi G, Vinuela F, Sepetka I, Macellari V (1991)
Electrothrombosis of saccular aneurysms via
endovascular approach. Part 1: electrochemical basis,
technique, and experimental results. J Neurosurg
75:1–7
17. Guglielmi G, Vinuela F, Dion J, Duckwiller G (1991)
Electrothrombosis of saccular aneurysms via
endovascular approach. Part 2: preliminary clinical
experience. J Neurosurg 75:8–14
18. Guglielmi G, Vinuela F, Dion J, Duckwiller G, Lylyk P
(1992) Endovascular treatment of posterior circulation
aneurysms by electrothrombosis using electrically
detachable coils. J Neurosurg 77:515–524
19. Moret J, Pierot L, Boulin A, Castaings L (1996)
Endovascular treatment of anterior communicating
artery aneurysms using Guglielmi detachable coils.
Neuroradiology 38:800–805
20. Houdart E (1996) Traitement par spires (coils) à
detechement controlé électrique de 315 aneurysmes
intracraniens. Acad Natl Med 180:1173–1183
21. Cognard C, Pierot L, Boulin A et al (1997) Intracranial
aneurysms: endovascular treatment with mechanical
detachable spirals in 60 aneurysms. J Radiol
202:783–792
22. Brilstra EH, Rinkel GJE et al (1999) Treatment of
intracranial aneurysms by embolization with coils a
systematic review. Stroke 30:470–476
23. Moret J, Cognard C, Weill A et al (1997) The
“remodeling technique” in the treatment of wide
neck intracranial aneurysms. Angiographic results
and clinical follow up in 56 cases. Interv Neuroradiol
3:21–35
24. Aletich VA, Debrun GM, Mukesh M (2000) The
remodeling technique of balloon-assisted Guglielmi
detachable coil placement in wide-necked aneurysms:
experience at the University of Illinois at Chicago.
J Neurosurg 93:388–396
25. Cottier JP, Pasco A, Gallas S, Gabrillargues J
et al (2001) Utility of balloon-assisted Guglielmi
detachable coiling in the treatment of 49 cerebral
L. Quilici and E. Boccardi
aneurysms: a retrospective, multicenter study. Am J
Neuroradiol 22:345–351
26. ISUIA (2003) International study of unruptured intra-
cranial aneurysms: natural history, clinical outcome,
and risks of surgical and endovascular treatment.
Lancet 362(12):103–110
27. Molyneux AJ, Kerr RS, Clarke M et al (2002) Interna-
tional subarachnoid aneurysm trial (ISAT) of neurosur-
gical clipping versus endovascular coiling in 2143
patients with ruptured intracranial aneurysms:
a randomised trial. Lancet 360:1267–1274
28. Wakhloo AK, Lanzino G (1998) Stents for intracranial
aneurysms: the beginning of a new endovascular era?
Neurosurgery 43(2):377–379
29. Lanzino G, Wakhloo AK (1999) Efficacy and current
limitations of intravascular stents for intracranial inter-
nal carotid, vertebral, and basilar artery aneurysms.
J Neurosurg 91(4):538–546
30. Henkes H, Bose A, Felber S (2002) Endovascular coil
occlusion of intracranial aneurysms assisted by a novel
self-expandable nitinol microstent (neuroform). Interv
Neuroradiol 8:107–119
31. David F, Felipe A, Vivek D, Cameron GM (2005)
Usefulness of the Neuroform stent for the treatment
of cerebral aneurysms: results at initial (3–6 mo)
follow-up. Neurosurgery 56:1191–1202
32. Mocco J, Snyder KV, Albuquerque FC et al (2009)
Treatment of intracranial aneurysms with the Enterprise
stent: a multicenter registry. J Neurosurg 110:35–39
33. Pumar JM, Blanco M, Vasquez F (2005) Preliminary
experience with Leo self-expanding stent for the treat-
ment of intracranial aneurysms. Am J Neuroradiol
26:2573–2577
34. Liebig T, Henkes H, Reinartz J (2006) A novel self-
expanding fully retrievable intracranial stent (SOLO):
experience in nine procedures of stent-assisted aneu-
rysm coil occlusion. Neuroradiology 48:471–478
35. Turner RD, Turk A, Chaudry I (2013) Low-profile
visible intraluminal support device: immediate out-
come of the first three US cases. J Neurointerv Surg
5:157–160
36. Geyik S, Yavuz K, Yurttutan N (2013) Stent-assisted
coiling in endovascular treatment of 500 consecutive
cerebral aneurysms with long-term follow-up. Am J
Neuroradiol 34:2157–2162
37. Nishido H, Piotin M, Bartolini B (2014) Analysis of
complications and recurrences of aneurysm coiling
with special emphasis on the stent-assisted technique.
Am J Neuroradiol 35(2):339–344
38. Pumar JP, Lete I, Pardo MI (2008) Leo stent
monotherapy for the endovascular reconstruction of
fusiform aneurysms of the middle cerebral artery. Am
J Neuroradiol 29:1775–1776
39. Juszkat R, Nowak S, Mol S (2007) Leo stent for
endovascular treatment of broad-necked and fusiform
intracranial aneurysms. Interv Neuroradiol 13:255–269
40. Saatci I, Yavuz K, Ozer C (2012) Treatment of intra-
cranial aneurysms using the pipeline flow-diverter
26 Endovascular Treatment of Brain Aneurysms
embolization device: a single center experience with
long-term follow-up results. Am J Neuroradiol
33:1436–1446
41. Piano M, Valvassori L, Quilici L (2013) Midterm and
long-term follow-up of cerebral aneurysms treated with
flow diverter devices: a single center experience.
J Neurosurg 118:408–416
42. Lanzino G (2013) Flow diversion for intracranial aneu-
rysms. J Neurosurg 118:405–407
43. Brinjikji W, Murad MH, Lanzino G (2013)
Endovascular treatment of intracranial aneurysms with
flow diverters: a meta-analysis. Stroke 44:442–447
44. Çinar C, Oran I, Bozkaya H et al (2013) Endovascular
treatment of ruptured blister-like aneurysms with spe-
cial reference to the flow-diverting strategy. Neurora-
diology 55:441–447
45. van Oel LI, van Rooij WJ, Sluzewski M et al (2013)
Reconstructive endovascular treatment of fusiform and
dissecting basilar trunk aneurysms with flow diverters,
stents and coils. Am J Neuroradiol 34:589–595
46. Siddiqui AH, Abla AA, Kan P et al (2012) Panacea or
problem: flow diversion the treatment of symptomatic
581
large or giant fusiform vertebrobasilar aneurysms.
J Neurosurg 116:1258–1266
47. Siddiqui A, Kan P, Abla AA et al (2013) Complication
after treatment with pipeline embolization for giant
intracranial aneurysms with or without coil emboliza-
tion. Neurosurgery 71(2):509–513
48. Tomas C, Benaissa A, Herbreteau D et al (2014)
Delayed ipsilateral parenchymal hemorrhage follow-
ing treatment of intracranial aneurysms with flow
diverter. Neuroradiology 56:155–161
49. Pierot L, Liebig T, Sychra V et al (2012) Intrasaccular
flow-disruption treatment of intracranial aneurysms:
preliminary results of a multi center clinical study.
Am J Neuroradiol 33:1232–1238
50. Lubicz B, Klisch J, Gauvrit Y et al (2014)
Web-endovascular treatment of wide-neck bifurcation
aneurysms: short- and mid-term results in a European
study. Am J Neuroradiol 35:432–438
51. Wang K, Yuan S, Zhang X (2013) Biodegradable flow-
diverting device for the treatment of intracranial
aneurysm: short term results of a rabbit experiment.
Neuroradiology 55:621–628
 Postop Evaluation of Aneurysms
(Including Vasospasm) 27
Michael Mayich, Brian P. Walcott, Christopher J. Stapleton,
and Daniel Thomas Ginat

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583 Both endovascular and microsurgical
approaches are used to treat cerebral aneu-
Overview of Treatment Options . . . . . . . . . . . . . . . . . . . 584
rysms. Endovascular techniques include coil
Imaging Modalities and Expected Posttreatment embolization, stenting, or a combination of
Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585 these. Microsurgical procedures include clip-
Vasospasm and Posttreatment Complications . . . 589 ping, without or with vessel bypass. Radiolog-
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 599 ical imaging, particularly with digital
subtraction angiography, CTA, and MRA,
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 599
plays a critical role in the posttreatment assess-
ment of patient with cerebral aneurysms. The
indications for various imaging modalities
after treatment of cerebral aneurysms, potential
artifacts related to imaging, and strategies to
mitigate these, expected imaging findings fol-
lowing treatment, as well as the imaging find-
ings for associated complications are discussed
and illustrated in this chapter.

Keywords
Endovascular • Microsurgical • Cerebral aneu-
rysms • Coil embolization • Stenting • Digital
M. Mayich
subtraction angiography • CTA • MRA • Arti-
Department of Radiology, University of Chicago, Chicago,
IL, USA facts • Complications
e-mail: Michael.Mayich@uchospitals.edu
B.P. Walcott • C.J. Stapleton
Department of Neurosurgery, Massachusetts General Introduction
Hospital, Boston, MA, USA
e-mail: Walcott.Brian@mgh.harvard.edu;
Cerebral aneurysms are rather common in the
Stapleton.Christopher@mgh.harvard.edu
general adult population, with a prevalence of
D.T. Ginat (*)
1–6 %, and aneurysm rupture accounts for the
Department of Radiology, University of Chicago, Chicago,
IL, USA majority of the estimated 30,000 cases of
e-mail: ginatd01@gmail.com nontraumatic subarachnoid hemorrhage per year
# Springer Science+Business Media New York 2016 583
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_44
584
in the United States [1]. The lesions can result in
considerable morbidity and mortality, particularly
if there is rupture into the subarachnoid space.
Consequently, large numbers of cerebral aneu-
rysms undergo treatment on a yearly basis
throughout the world. The two main treatment
options include open surgical approaches, which
involve clipping the base of the aneurysm with or
without accompanying bypass, versus
endovascular techniques, which includes coil
embolization, with or without accompanying
stenting, and the use of more recent flow-diverting
devices. Diagnostic imaging plays a crucial role in
the posttreatment evaluation of cerebral aneu-
rysms. Various imaging modalities are available,
including computer tomography (CT)/computer
tomography angiography (CTA), magnetic reso-
nance imaging (MRI)/magnetic resonance angi-
ography (MRA), and conventional catheter
angiography. The treatment option, imaging
modalities and associated artifact and technical
advances, and expected and complicated
posttreatment findings related to aneurysms, with
an emphasis on vasospasm, are reviewed in the
following sections.
Overview of Treatment Options
The optimal treatment of aneurysms has been a
source of intense investigation which is evolving
as our understanding of the benefits and draw-
backs of endovascular versus microsurgical ther-
apeutic approaches continues to deepen. While
endovascular approaches including coiling or
flow diversion offer the possibility of avoiding a
craniotomy, there are a unique set of consider-
ations to take into account when considering this
approach. The International Subarachnoid Aneu-
rysm Trial (ISAT) released in 2005 was a land-
mark study which prospectively compared
endovascular coiling versus neurosurgical clip-
ping in aneurysms amenable to either therapy
[2]. The trial demonstrated a risk reduction of
7.4 % for death or dependence at 1 year with an
early survival advantage maintained for up to
7 years in patients who underwent coiling versus
M. Mayich et al.
clipping and led to an increased acceptance of coil
embolization clinically. There remains a well-
established role for both therapies; however, an
understanding of their features and potential com-
plications is essential for those involved in the
management of patients with intracranial aneu-
rysm. Certainly, whether either type of interven-
tion is performed, medical therapy also plays an
important role in the management of cerebral
aneurysms and includes smoking cessation and
blood pressure control.
Microsurgical clipping is an open neurosurgical
procedure that excludes a saccular aneurysm from
the parent circulation and offers definitive treat-
ment. Aneurysm clips basically consist of blade
and spring components and are almost universally
constructed of titanium in the modern era. Thus, the
titanium aneurysm clips are MRI compatible. The
clips are placed under direct visualization with the
aid of the operating microscope. A variety of sur-
gical approaches (craniotomy techniques) can be
utilized, which permit direct line of sight to the area
of interest. Examples include mini-pterional, lateral
supraorbital, and orbital-pterional craniotomies.
Additional clinoidectomy, or removal of portions
of the clinoid process, may be necessary to more
readily access the ophthalmic segment ICA aneu-
rysms. In addition to bone removal, extensive
Sylvian fissure subarachnoid dissection is often
necessary to allow for the separation of the frontal
and temporal lobes in order to access circle of
Willis aneurysms. By creating a working space
with exposure of the circle of Willis, the aneurysm
clips can be delivered via delicate handheld appli-
cators to the neck of an aneurysm. The main goal of
aneurysm clipping is to minimize the risk of rup-
ture. Another important goal of any surgery is to
preserve blood flow through the parent artery, since
any stenosis or occlusion of the parent artery circu-
lation could result in cerebral ischemia. In order to
preserve adequate cerebral perfusion, direct extra-
cranial to intracranial arterial bypass is indicated
for the treatment of complex aneurysms and con-
sists of anastomosing relatively large-caliber ves-
sels or grafts, most commonly the superficial
temporal artery (STA) to the middle cerebral artery
(MCA) [3].
27 Postop Evaluation of Aneurysms (Including Vasospasm)
Endovascular therapy for cerebral aneurysms
is a minimally invasive alternative to microsurgi-
cal clipping. Endovascular techniques include coil
embolization of the aneurysm sac for narrow-neck
saccular aneurysm and stent-assisted coil emboli-
zation for wide-neck saccular aneurysms or fusi-
form aneurysms. The coils form a “mass” within
the lumen of the aneurysm. The formation of
thrombus associated with the coils is responsible
for occluding the aneurysm. When stents are used,
these span the neck of the aneurysm in the parent
vessel in order to prevent prolapse of the coils and
maintain patency of the parent vessel. Nitinol,
platinum, nickel, iridium, and tungsten have
been the primary metals used in construction of
coils and stents, which are MRI-compatible.
Matrix coils have also been developed, which
consist of thin platinum coils covered with a
bioabsorbable, polymeric material, such as
polyglycolic acid/lactide, and can accelerate aneu-
rysm fibrosis and neointima formation without
parent artery stenosis. Occasionally, liquid embo-
lization agents such as Onyx may be used in
conjunction with coils for the treatment of large
cerebral aneurysms.
More recently, flow-diverting stents, such as
the Pipeline Embolization Device and SILK
Flow Diverter, have been introduced for the treat-
ment of aneurysms without coil embolization and
are mainly indicated for large, wide-necked inter-
nal carotid artery (ICA) aneurysms. These devices
consist of a porous tubular tight mesh that can be
positioned across the aneurysm neck while
maintaining patency of the patent vessel. The
strategy behind the use of flow-diverting stents is
to channel blood flow away from the aneurysm,
which consequently thrombose. The devices are
composed of platinum and nickel-cobalt-chro-
mium alloy and are thus MRI-compatible. These
flow-diverting devices provide successful and
durable occlusion of wide-neck intracranial aneu-
rysms without severe complications [4].
As a last resort, parent vessels may be occluded
via an open surgical or endovascular approach,
particularly for posterior circulation fusiform
aneurysms or dissecting aneurysms. Typically a
balloon occlusion test is performed before such an
585
intervention in order to assess adequate collateral
circulation and risk of developing a stroke if the
vessel is sacrificed. Vessel occlusion can be
performed through the use of coils or clips.
Ultimately, the overall success of aneurysm
treatment can be gauged by the degree of aneu-
rysm obliteration, patency of parent and perforat-
ing arterial vasculature, and the durability of the
aneurysm obliteration. Recanalization, recur-
rence, and incomplete obliteration are complica-
tions known to occur infrequently following any
aneurysm treatment, and their incidence may be
related to the treatment modality [5]. ISAT and
other studies have demonstrated the superiority of
clipping compared with coiling providing long-
term protection from re-hemorrhage [2]. In ISAT,
two patients that had documented complete angio-
graphic occlusion of their aneurysm following
coiling suffered from re-hemorrhage events. It
should be noted that aneurysm growth after clip-
ping has also been reported in several cases
[6]. Regardless of whether an aneurysm is clipped
or coiled, follow-up imaging is mandated to
ensure the durability of treatment. The various
imaging modalities available for accomplishing
this task and the potential complications that
may be encountered after aneurysm treatment
are reviewed in the subsequent sections.
Imaging Modalities and Expected
Posttreatment Findings
The main purpose of diagnostic imaging follow-
ing aneurysm clipping and endovascular therapy
is to confirm adequate exclusion of the aneurysm.
In general, CTA is superior to MRA for the eval-
uation of aneurysms after surgical clipping, while
MRA is the preferred noninvasive imaging
modality following coil embolization [10]. Diag-
nostic digital subtraction angiography is generally
reserved as second-line imaging modality for
addressing issues that cannot be adequately
answered on the noninvasive cross-sectional
imaging exams.
Axial source images, coronal and sagittal
multiplanar reformats, and interactive 3D volume
586
Fig. 1 Aneurysm clip. The CT image shows the spring
(arrow) and blade (arrowhead) components of a titanium
aneurysm clip
renderings are used to evaluate the treated aneu-
rysm and the rest of the cerebral circulation. CTA
can often adequately depict small aneurysm rem-
nants, as well as demonstrate patency, stenosis, or
vasospasm in the adjacent parent vessels. Cer-
tainly, metal artifacts can be a limiting factor for
the use of CTA following aneurysm clipping.
However, the presence of artifacts is much less
with titanium aneurysm clips compared with
cobalt clips, and the components of the clips can
even be delineated in detail (Fig. 1). In addition,
the orientation of the clips can affect artifacts in
the peri-clip region. For example, clips position-
ing perpendicular relative to the parent vessel
results in fewer artifacts than those parallel to the
parent vessel. In addition, clips that lie oblique to
the scanning plane produce fewer artifacts.
Another challenge pertaining to postoperative
via CTA is the proximity of osseous structures,
such as the skull base or clinoid processes, to
segments of the cerebral circulation. Indeed,
beam-hardening artifact from the dense bone can
further obscure the treated aneurysm in addition to
the effects of the clips. This issue can be mitigated
M. Mayich et al.
by using CTA post-processing bone subtraction
techniques. However, this approach requires an
additional scan and increased radiation dose.
The advent of dual-energy CT (DECT) imag-
ing in conjunction with Metal Artifact Reduction
Software (MARS) has been shown to improve
image quality of aneurysm, parent vessel, and
adjacent parenchyma when compared to single-
energy conventional CT examinations and offers a
potential future application for CT in surveillance
of treated lesions [8]. Additionally, DECT offers
the capability of constructing virtual unenhanced
images, offering the capability of monitoring rel-
ative contributions of hemorrhage versus contrast
to hyperattenuation in the subarachnoid space
related to pre- or post-procedure aneurysm rup-
ture, extravasated contrast from the procedure
itself, or both [9]. Furthermore, DECT can be
useful for bone subtraction without an additional
scan. Extracranial-intracranial bypass performed
in conjunction with cerebral aneurysm clipping is
most commonly performed between the superfi-
cial temporal artery and MCA or between the
occipital artery and posterior cerebral artery
(PCA). Patency of the anastomoses between
bypass vessels can be assessed on CTA and
MRA (Fig. 2). The STA often increases in caliber
after successful revascularization.
Following endovascular coil embolization,
metallic streak artifact related to platinum alloy
coils can considerably obscure the aneurysm and
the parent vessel as well as the adjacent paren-
chyma, limiting the usefulness of CTA. On the
other hand, several studies have demonstrated the
sensitivity of MRA for residual aneurysm after
coil embolization or stent-assisted coil emboliza-
tion to be between 90 % and 100 % [9], while
others have determined that MRA may be at least
as sensitive as DSA if performed carefully (Fig. 3)
[10]. TOF-MRA and contrast-enhanced MRA
have similar accuracy for the evaluation of coiled
aneurysms. Nevertheless, some institution proto-
cols perform both techniques sequentially,
employing CE-MRA data to supplement assess-
ment by TOF-MRA. However, with regard to
post-stent-assisted coiling, the excess susceptibil-
ity artifacts related to the stent material on time-of-
27 Postop Evaluation of Aneurysms (Including Vasospasm) 587

Fig. 2 Aneurysm clipping with extracranial-intracranial corresponding time-of-flight MTA (c) show a patent right
bypass. Axial CT image (a) shows a right paraclinoid extracranial bypass graft and anastomosis with the right
aneurysm clip (arrow). 3D CTA image (b) and MCA (arrowhead)

Fig. 3 CT versus MRA.

Non-contrast CT (a) after
coiling demonstrates
extensive metallic streak
artifact that obscures
adjacent structures. The
corresponding time-of-
flight 3-T MRA (b)
demonstrates minimal
susceptibility artifact (*)
allowing visualization of
the patent parent vessel
(arrow) adjacent to the
coil mass
588
Fig. 4 The utility of
contrast MRA. (a) 3D
TOF-MRA post-stenting of
basilar fusiform aneurysm
at 3-T basilar artery
obscured by susceptibility
artifact (white arrow).
Contrast-enhanced MRA at
the same level (b)
demonstrates both
opacification of the vessel
lumen (arrow) and a small
focus of residual flow in the
aneurysm (arrowhead)
flight MRA can obscure the aneurysm neck and
parent vessel. Thus, in such cases, follow-up with
contrast-enhanced MRA may be more beneficial
(Fig. 4). In the case of embolization devices, the
used CTA for follow-up is often preferable due to
even greater artifact of MRA. Indeed, the lumen
of the embolization devices and status post the
treated aneurysms can be defined in detail on
CTA (Fig. 5).
There are several variables which may be con-
sidered when optimizing imaging parameters. For
example, increasing receiver bandwidth can min-
imize susceptibility artifact related to metal
implants [11]. By minimizing the amount of time
during which echo sampling occurs (TE),
susceptibility-related distortion of the image can
be decreased, although this occurs to the detriment
of signal-to-noise ratio. While protocols may vary,
as a general principle, doubling the readout band-
width yields good results. Shortening TE time
may also decrease susceptibility artifact related
to the coil mass and/or associated hardware
[12]. Gonner et al. found that decreasing TE
from 6 to 2.4 ms significantly reduced the volume
of susceptibility artifact associated with the coil
mass and additionally allowed improved visuali-
zation of the adjacent artery in 36 % of cases [13].
Although imaging at 3-T provides greater
signal-to-noise ratio than at lower field strengths,
the degree of susceptibility artifact associated
with the coil mass is also increased. An in vitro
study by Walker et al. showed a significant
increase in the amount of susceptibility artifact
and coil mass volume overestimation at 3-T
M. Mayich et al.
versus 1.5-T [12]. They also showed that the
volume of artifact depended on packing density
of the aneurysm, with a more densely packed
aneurysm resulting in significant differences in
volume overestimation. They found that the use
of a short TE protocol improved visualization,
especially on 3D TOF images, concluding that
with all variables being equal, 1.5-T was superior
to 3-T when 3D TOF images were used. However,
a prospective clinical study by Kaufmann
et al. showed no significant difference in sensitiv-
ity for detecting an aneurysm remnant between
1.5-T and 3-T [14]. Ultimately, an accurate inter-
pretation of the posttreatment MRA requires a
familiarity with the pretreatment and immediate
post-embolization conventional angiograms. A
review of the images from the initial treatment
provides the reader with a 2D projection of the
3D anatomy, demonstrating the orientation of the
coil mass and aneurysm neck with respect to
the parent vessel.
Digital subtraction angiography (DSA) is the
modality of choice for treated aneurysm follow-up
in cases when examination with MRA is
contraindicated or noninvasive diagnostic imag-
ing is inconclusive. DSA allows for exceptional
spatial resolution, multiplanar projection, and the
added dimension of dynamic temporal assessment
of the contrast bolus, eliminating issues such as
venous contamination encountered in cross-
sectional techniques. The procedure is invasive,
however, carrying an associated risk of neurologic
complication in 2.6 %, stroke with permanent
disability in 0.14 %, and death in 0.06 %
27 Postop Evaluation of Aneurysms (Including Vasospasm) 589

Fig. 5 Pipeline device.

Pretreatment coronal CTA
(a) shows a large right
superior hypophyseal
aneurysm (*). Fluoroscopic
image (b) obtained shows
the Pipeline device
(arrowheads) in position
within the right ICA. Digital
subtraction angiography (c)
at the completion of
Pipeline Embolization
device insertion shows
persistent filling of the
aneurysm sac (*). Follow-
up CTA obtained 9 months
later shows near-complete
exclusion of the aneurysm
due to thrombosis of the sac
with minimal residual
filling of the neck (arrow)
and Pipeline device in
position (arrowheads).
Curved planar reformatted
CTA image (e) shows the
full length of the patent
flow-diverting stent (arrow)

[7, 15]. 2.6 % is a commonly quoted value,

although it is notable that this large series showed
a decreasing rate of complications in examina-
tions performed more recently in the series.
Meta-analyses also demonstrate that the risk of
angiographic complication is significantly lower
in patients with SAH, cerebral aneurysm, or AVM
than in patients with TIA/stroke [16]. In general,
CTA tended to overestimate aneurysm volume and
MRA consistently underestimated volume, while
DSA can both over- and underestimate aneurysm
volume, providing a more accurate value on aver-
age than either of the other modalities [17].
Vasospasm and Posttreatment
Complications
Aneurysmal subarachnoid hemorrhage is associ-
ated with a high mortality rate and is attributable
to the initial hemorrhage event or aneurysm
re-hemorrhage prior to surgical or endovascular
obliteration. In survivors, the days that follow
aneurysm obliteration are a critical period where
patients are monitored intensely for the develop-
ment of cerebral vasospasm. It is this secondary
insult of vasospasm that impairs cerebral blood
590
flow and is the major cause of delayed ischemia
and stroke [18]. Although the pathophysiology of
vasospasm is incompletely understood, it is gen-
erally accepted that the rapid institution of thera-
pies, including systemic hypertension and intra-
arterial vasodilator therapy, can be used in symp-
tomatic patients to prevent permanent deficits
[19]. Therefore, the period following aneurysmal
subarachnoid, whether the aneurysm is treated with
clipping or coiling, represents a critical period
where clinical care is heavily influenced by the
information provided by neuroimaging. The defi-
nition of cerebral vasospasm is new onset of intra-
cranial arterial narrowing and it typically described
qualitatively (mild, moderate, or severe) as subjec-
tively interpreted on catheter or CT angiography
[20]. Elevated transcranial Doppler velocities of
>200 cm/s not thought to be a result of hyperemia
are also classified as cases of vasospasm. Symp-
tomatic vasospasm is a subgroup of vasospasm
where clinical neurological symptoms are referable
to a region of radiographic or sonographic vaso-
spasm in the absence of alternative explanations.
The purpose of any imaging surveillance for
vasospasm is to identify changes in vessel caliber
and/or blood flow prior to neurological deteriora-
tion. Not all patients with subarachnoid hemor-
rhage from ruptured cerebral aneurysms go on to
develop cerebral vasospasm. Radiographic grad-
ing scales, such as the Fisher scale, have been
developed to describe a relationship between
quantity or pattern of blood and the risk of future
vasospasm development. However, patients with
similar hemorrhage patterns and quantitative
amount of subarachnoid blood may have a drasti-
cally different clinical course (with some patients
experiencing no vasospasm at all). This empha-
sizes that in addition to the subarachnoid hemor-
rhage itself, there are patient-specific factors that
predispose an individual to cerebral vasospasm.
Development of radiographic biomarkers and
studies assessing the efficacy of surveillance
imaging will be important to improve manage-
ment in these critical patients. However, there is
no evidence from randomized controlled trials
that guide the detection or management of vaso-
spasm following aneurysmal subarachnoid hem-
orrhage. Practice patterns vary widely from
M. Mayich et al.
institution to institution, where various means of
radiographic and sonographic surveillance aid to
augment serial clinical examinations. In patients
with Glasgow Coma Scale scores less than
9 (comatose), these radiographic surveillance
methods are the main method of vasospasm detec-
tion, as it is challenging to detect subtle differ-
ences in neurological exam in a poor neurological
state.
Doppler ultrasound is a noninvasive modality
that can be used to measure cerebral blood flow
[21]. The general concept of the modality is that
blood velocity increases as vessel diameter
decreases, as is seen in the setting of vasospasm.
However, hyperemia can also result in elevated
velocity recordings, confounding interpretation.
A variety of measurement techniques have been
used to help differentiate between these two con-
ditions (vasospasm and hyperemia), such as the
calculation of the Lindegaard ratio, which com-
pares the blood flow in the more proximal carotid
artery with the MCA [22]. Identification of these
differences in blood flow is potentially useful in
determining focal areas of vasospasm as com-
pared to global changes in blood flow that are
seen with hyperemic states or diffuse vasospasm.
Drawbacks of transcranial Doppler ultrasound to
detect vasospasm are inherent in its limited sensi-
tivity and specificity. Some reports caution that
there is a lack of evidence and lack of accuracy of
transcranial Doppler, criticizing its use in routine
clinical practice [23]. Nevertheless, because of its
noninvasive nature, lack of radiation exposure,
and ease of daily bedside acquisition, it is rou-
tinely used at several high volume neurovascular
institutions.
CTA is also useful for the detection of cerebral
vasospasm following subarachnoid hemorrhage
post-clipping, particularly when there is severe
spasm located in proximal artery locations, such
as the A1 segment of the anterior cerebral artery
(ACA) or the M1 segment of the MCA (Fig. 6). In
addition, CT perfusion can be performed concur-
rently to provide insight into the extent of cerebral
ischemia resulting from vasospasm. However,
CTA may be less accurate for detecting mild and
moderate spasm in distal vessel locations
[24]. The use of CTA in patients with renal
27 Postop Evaluation of Aneurysms (Including Vasospasm) 591

Fig. 6 Cerebral vasospasm

depicted on CTA and CT
perfusion. Axial
non-contrast CT (a) shows
extensive subarachnoid
hemorrhage related to a
ruptured basilar artery
aneurysm. Axial CTA MIP
image (b) obtained several
days after aneurysm coiling
shows severe diffuse
narrowing of the proximal
circle of Willis branches,
consistent with vasospasm.
The corresponding CBV (c)
and MTT (d) maps show
perfusion deficits,
particularly severe in the
left parietal lobe (arrows)

impairment is limited secondary to the necessity
of iodinated contrast. CTA is also associated with
a radiation dose that can be significant in critical
patients undergoing serial imaging in the vaso-
spasm “window.” Future study in the form of
randomized trials is needed to define the efficacy
of either symptom-prompted CTA or routine sur-
veillance CTA following aneurysmal subarach-
noid hemorrhage and its effect on outcome.
Ultimately, catheter-based angiography has
been considered to be the historical gold standard
to diagnose vasospasm [25]. Advantages of
catheter-based angiography include its high spec-
ificity and sensitivity. It also allows for the imme-
diate administration of intra-arterial therapy
should severe vasospasm be detected (Fig. 7).
The main risks of the diagnostic procedure
include stroke (estimated at 1 % or less), local
access site complications, the need for iodinated
contrast, and radiation exposure [16].
In addition to direct imaging of the cerebral
vasculature, there are several imaging modalities
that may be able to detect aberrations in cerebral
perfusion irrespective of changes in large vessel
caliber. This is an important consideration
because vasospasm of the microcirculation may
not routinely be detected on either CT angiogra-
phy or digital subtraction angiography. The most
commonly used dynamic imaging technique for
this purpose is CT perfusion, likely as a result of
its widespread availability and ease of acquisition.
Other dynamic imaging techniques, such as posi-
tron emission tomography, have also been used to
identify regions of impaired regional blood flow
or regional cerebral metabolic rate of oxygen.
Emphasis on imaging of brain perfusion may be
able to better evaluate for vasospasm, particularly
since “large vessel” constriction may only occur
in a portion of affected cases.
Periprocedural aneurysm rupture is a
documented complication of coil embolization
occurring in 4.6 % of all cases and 20.4 % of
balloon-assisted cases in one large series (Fig. 8)
[26]. While a distressing complication, the
592
Fig. 7 Digital subtraction angiography of cerebral vaso-
spasm. Five days postoperatively, the patient was
suspected to have developed left MCA vasospasm based
on elevated transcranial Doppler velocities in that distribu-
tion. Initial lateral projection DSA image (a) shows severe
Fig. 8 Intraprocedural aneurysm rupture: DSA demon-
strates extravasation of contrast from a partially coiled
ACoA aneurysm (arrow)
M. Mayich et al.
stenosis of the MCA (arrow). The patient was treated with
intra-arterial administration of milrinone and nicardipine,
with nearly resolved spasm of the MCA (arrow), as dem-
onstrated on follow-up lateral DSA projection (b)
majority of patients demonstrated an absence of
clinical sequelae (74 %) related to the event. Early
rebleeding is also a well-documented complica-
tion following coil embolization [27]. The ISAT
study found rates of rebleeding among ruptured
aneurysms treated with coiling of 2.5 % in the first
year and 0.2 % in subsequent years. The CARAT
trial demonstrated rebleeding risk in the first
3 years to be 2.2 %, 0.2 %, and 0 %, respectively
[28]. Hemorrhage at any point following surgical
clipping is considered to be a rare event with
accumulative estimated risk of between 1 % and
2 % [3]. Unruptured aneurysms treated with coil
embolization demonstrate an exceptionally low
rate of subsequent bleeding, estimated at 0.2 %
annually in one meta-analysis [29].
Thromboembolic complications are a well-
described complication of coil embolization and
range from silent incidental findings on subse-
quent imaging to devastating findings. Silent
thromboembolic events are common, presenting
as multifocal diffusion restriction on DWI
sequence images in the postoperative period.
27 Postop Evaluation of Aneurysms (Including Vasospasm)
Fig. 9 Silent
thromboembolic events. At
6 days post-coiling of a
ruptured 6 mm ACoA
aneurysm, there is
multifocal diffusion
restriction depicted on DWI
(a) and ADC (b)
Lesions are typically small, often multifocal, and
typically localize to the vascular territory of the
vessel being treated (Fig. 9). One of the earlier
descriptions by Rordorf et al. showed multifocal
restricted diffusion in 61 % of patients who were
imaged within 48 h of uncomplicated coiling
[30]. A recent larger series of 342 patients by
Kang et al. showed an incidence of
DWI-positive lesions of 54.5 %, while symptoms
were only detected in 4.1 % of patients, resolving
in approximately half. The presence of 6 lesions
was, however, correlated with the presence of
symptoms – 78.6 % of symptomatic patients had
6 lesions implying the burden of DWI could
potentially be employed as a surrogate marker
for symptomatic ischemia [31]. Occasionally,
larger territorial infarcts result when a branch
associated with an aneurysm is sacrificed or as a
result of parent vessel is purposely embolized
(Fig. 10). Stent-assisted coiling is also associated
with silent thromboembolism, although limited
literature at present precludes assessment of
whether the risk is significantly elevated in this
population. Altay et al. found DWI lesions in
30 % of 46 patients with unruptured aneurysms
treated with stent-assisted coiling [32]. Hahne-
mann et al. recently reported on a series of
75 patients treated with stent assistance in which
48 (64 %) were found to develop ischemic lesions
[33]. Stent-assisted coiling techniques are associ-
ated with an increased risk of thrombosis
593
compared to coil embolization alone. In addition
to early thrombosis, stent-assisted coiling results
in the risk of both delayed in-stent stenosis and
delayed thrombosis. A recent meta-analysis deter-
mined a rate of delayed stenosis of 5.3 % (range of
0–20.6 %), which correlates with a prospective
database assessing the Enterprise stent that found
rates of delayed stenosis and thrombosis of 3 %
each [34]. Rates of in-stent stenosis and thrombo-
sis compare favorably with rates documented fol-
lowing stenting of intracranial vessels for
atherosclerosis during the SAMMPRIS trial of
28.3 % and 3.9 %, respectively [35].
Clinically significant thromboembolism is a
much less common complication, although this
is the most common cause of periprocedural mor-
bidity associated with coil embolization of intra-
cranial aneurysms, reported to occur in 1–17 % of
cases, including 4.7 % of cases in one large retro-
spective series [36, 37]. During as many as 11 %
of procedures, distal emboli or intraluminal
thrombus related to the coil mass/aneurysm may
be demonstrated (Fig. 11) [38]. If encountered
during the coiling procedure, therapy including
infusion of a glycoprotein IIB-IIIA inhibitor
such as abciximab (ReoPro™) may be employed
to acutely lyse thrombus, although thromboem-
bolic complications occurring after the comple-
tion of the procedure pose a more difficult
therapeutic conundrum, especially if the treated
aneurysm was ruptured at the time of presentation.
594 M. Mayich et al.

Fig. 10 Infarct due to

parent vessel sacrifice with
coil embolization.
Pretreatment 3D surface-
rendered CTA image (a)
shows a dissecting
aneurysm of the basilar
artery with proximal
stenosis. Post-embolization
3D surface-rendered CTA
image (b) shows coils
within the lesion and absent
opacification of the distal
basilar artery (arrowhead).
Axial DWI (c) and ADC
map (d) shows associated
restricted diffusion in the
pons, which correspond to
acute infarcts in the
perforator artery territory

Risk factors for symptomatic thromboembolism thromboembolic complications. The increased

in one series of 159 procedures included mass
effect by the treated lesion, residual flow within
the coil mass at completion of the procedure,
prolapsed coils, and mean aneurysm diameter
larger than 12 mm [37]. Remodeling of the aneu-
rysm neck is also a risk factor for thromboembolic
complication. Use of a stent construct to remodel
the aneurysm neck resulted in acute or subacute
in-stent thrombosis in 4.5 % of cases in a series of
161 patients, and 2.3 % experienced delayed ste-
nosis or occlusion [39]. These results were in
keeping with a recent meta-analysis which found
a 4.6 % risk of intraprocedural and 4.3 % post-
procedural risk of stent-related thromboembolism
[40]. Balloon remodeling of the neck was also
associated with an increased risk of
risk of thromboembolism with neck remodeling
is at least partially attributable to selection bias
wherein larger, more complex aneurysms requir-
ing longer procedure times may impart an inher-
ently higher risk of thromboembolism.
Distant coil migration is a rare intraprocedural
or delayed complication of aneurysm emboliza-
tion (Fig. 12) [41]. This phenomenon can occur
despite the use of balloon remodeling or stent-
assisted techniques and can lead to territorial
infarction due to vessel occlusion. Vascular imag-
ing can be used to localize the migrated coil, and
MRI is useful for confirming and delineating the
extent of any associated cerebral infarction.
Aneurysm expansion or overpacking of aneu-
rysms from liquid agent or coil embolization can
27 Postop Evaluation of Aneurysms (Including Vasospasm)
Fig. 11 Intraprocedural thrombus formation related to a
prolapsed coil loop. Immediate administration of IV hepa-
rin halted clot propagation and decreased clot burden.
Antiplatelet therapy was optimized postoperatively, and
follow-up examination at 10 days failed to demonstrate
clot at the coil mass or occlusion of distal vessels.
Intraprocedural thrombus formation may complicate as
many as 11 % of procedures
Fig. 12 Coil migration. Sagittal MIP CTA obtained after
stent-assisted coiling of an ACOM aneurysm (arrowhead)
shows a coil (arrow) located in a more distal pericallosal
branch
occasionally lead to new symptoms due to the
increased mass effect [42]. The most commonly
involved structures are the optic and abducens
nerves. MRI is useful for evaluating the causative
595
Fig. 13 Cranial nerve compression after embolization.
The patient presented with diplopia following coil and
Onyx embolization of a giant left cavernous carotid aneu-
rysm. The coronal T2-WI image shows that the low signal
intensity embolic material exerts considerable mass effect
upon the cavernous sinus (arrow)
factor and affected cranial nerves (Fig. 13). Symp-
toms can persist or sometimes improve spontane-
ously over time.
A late complication of coil embolization is the
risk of aneurysm recurrence due to coil compac-
tion or aneurysm expansion, which manifests as a
discrete region of increased signal intensity on
MRA in contrast to the signal void from the coil
material (Fig. 14). Potential false positives include
hyperintense thrombus being mistaken for flow-
or contrast-related enhancement or enhancement
peripheral to the coil mass on CE-MRA which is a
normal finding most likely attributable to periph-
eral thrombus, inflammation, or healing. A recent
meta-analyses determined recurrence occurs in an
estimated 20.8 % of patients, necessitating
re-treatment in 10 % of cases [43]. This rate stands
in contrast to the exceptional stability of clipped
aneurysms; one study demonstrated an annual
recurrence rate of only 0.5 % at 4.4 years
[44]. Recurrence has been reported to be more
common following treatment of ruptured (17 %)
versus unruptured (7 %) aneurysms [45]. Other
factors that have been suggested to correlate with
596 M. Mayich et al.

Fig. 14 Recurrent
aneurysm after coil
embolization. Initial post-
embolization lateral DSA
image (a) shows no filling
of a pericallosal artery
aneurysm. Follow-up
lateral DSA image (b)
shows new partial contrast
opacification (arrow) of the
aneurism sac associated
with coiled compaction
(arrowhead). Time-of-flight
MRA images obtained at
1.5-T with a TE of 2.9 ms
(c) and 3-T with a TE of
3.4 ms (d) show
corresponding flow-related
enhancement within a
portion of the aneurysm sac
(arrows) and coil
compaction (arrowheads).
There is similar image
quality with regard to the
coiled aneurysm at 1.5-T
and 3-T

higher rate of recurrence include large aneurysm
size, presence of intraluminal thrombus, low
packing density, initial incomplete occlusion, lon-
ger follow-up, and wide neck, although specific
contributions by each or any of these risk factors
are debatable on the basis of differences in study
design [46]. The frequency at which recurrence
occurs has led to the standard practice of follow-
up imaging, although there is an absence of con-
sensus guidelines regarding the specific fre-
quency. One series of 126 patients found that
100 % of recurrences occurred within 6 months
[47], although the LOTUS I study showed 4 of
111 treated lesions reopened 5–11 years following
coiling, one of which necessitated re-treatment
[48]. A study assessing the incidence of subarach-
noid hemorrhage following successfully occluded
lesions on a 6-month follow-up exam demon-
strated a cumulative incidence of recurrent sub-
arachnoid hemorrhage of 0.4 % within 8 years of
coiling in a series of 283 patients [49]. These
studies together demonstrate an extremely low
incidence of aneurysm recurrence and exceed-
ingly low risk of subarachnoid hemorrhage in
patients whose treated aneurysms remain
occluded on examination at 6 months.
Both the ISAC and CARAT studies suggest a
low risk of late rebleeding, an observation corrob-
orated by Sluzewskia and Van Rooij who found
four instances in their series of 392 patients
followed for a mean of 51.5 months, with the
latest occurring at 40 months [50]. This compares
favorably with series of clipping, where hemor-
rhage following treatment is considered to be a
rare event with an estimated risk of between 1 %
and 2 % [3]. However, with regard to flow-
diverting devices, both early and delayed aneu-
rysm ruptures have been reported in the literature.
The underlying mechanism for this phenomenon
may include changes in the wall, shear stress at the
aneurysm dome, unfavorable hemodynamic
changes, and autolysis of the aneurysm wall sec-
ondary to red thrombus formation with accompa-
nying inflammation. The hemorrhage associated
27 Postop Evaluation of Aneurysms (Including Vasospasm)
Fig. 15 Aneurysm recurrence after clipping. Lateral dig-
ital subtraction angiogram shows aneurysm recurrence at
the origin of the posterior communicating artery (PComm),
adjacent to the clip (arrow)
with the presence of flow-diverting stents can be
distant from the treated aneurysm and can be
intraparenchymal.
Aneurysm recurrence after clipping is much
less common than after coiling but can neverthe-
less lead to devastating subarachnoid hemorrhage
and warrants evaluation via CTA or DSA
(Fig. 15). On the other hand, procedure-related
surgical complications are relatively common,
with an overall incidence of approximately
20 %, although this estimate can vary depending
on aneurysm location and surgeon experience.
Aneurysms of the basilar artery and the anterior
choroidal artery are more difficult to access via
craniotomy and thus have higher rates of compli-
cations than for clipping of MCA aneurysms, for
example. Procedure-related complications mainly
consist retraction-related injury, unintended cere-
bral infarction, wound infections, CSF leakage,
hydrocephalus, and cranial nerve injury. CT of
the brain is routinely performed in the early post-
operative period to evaluate many of these com-
plications, while MRI can be obtained to further
evaluate complications that are not explained by
CT. Indeed, CT can readily depict acute hemor-
rhage and edema associated with retraction injury
597
Fig. 16 Postoperative intraparenchymal hemorrhage and
edema. Coronal CT image shows acute hemorrhage and
vasogenic edema in the left temporal lobe (arrows) after
recent left MCA and ICA aneurysm clipping, ruptured
intraoperatively
and is important for excluding the presence of
herniation and midline shift, which may require
surgical decompression (Fig. 16). Diagnostic
imaging can be useful to confirm suspected
unintended infarction. While hypoattenuation
and swelling can be identified on CT (Fig. 17),
CT is relatively insensitive for the detection of
acute infarcts. Rather, MRI, particularly DWI
sequences, can be useful for further investigation.
Although superficial wound infections may be
clinically apparent, diagnostic imaging can be
useful for evaluating the extent of the infection,
particularly if intracranial involvement is
suspected. MRI with contrast is particularly useful
to rule out the presence of a pyogenic abscess,
which appears as a rim enhancing fluid collection
with restricted diffusion. Analogous to coil
masses, aneurysm clips can impinge upon the
cisternal segments of the cranial nerves, which
can be assessed using MRI cisternography-type
sequences, such as CISS or FIESTA.
The major complications associated with
extracranial-intracranial bypass performed with
aneurysm clipping include ischemic injury and
hyperperfusion syndrome [7]. Cerebral ischemia
can result from thromboembolism, hemodynamic
ischemia, or watershed shift phenomenon, which
598
Fig. 17 Perioperative
infarct. Axial CT images (a,
b) obtained after left PCOM
aneurysm clipping show
hypoattenuation in the left
medial temporal lobe
(arrow). MRI was obtained
shortly afterward and the
DWI (c) and ADC map (d)
confirms the presence of an
acute infarct in the left
medial temporal lobe
(arrows)
is a dynamic change in cerebral hemodynamics
caused by bypass flow. MRI findings can confirm
acute cerebral infarction by showing restricted
diffusion. Vascular imaging may show a filling
defect in the graft, anastomosis, or supplied vas-
cular territory. Perfusion imaging can reveal def-
icits in cerebral blood flow. Alternatively,
extracranial-intracranial bypass can cause an
increase in cerebral blood flow that can lead to
hyperperfusion syndrome after direct
extracranial-intracranial bypass can appear as
high signal intensity of the donor STA and dilated
branches of the MCA around the site of the anas-
tomosis, which is most apparent with 3T versus
1.5T MRI and approximately 1 week after sur-
gery. Perfusion imaging can help differentiate
hyperperfusion from ischemia, which is important
because these conditions have contradictory
management.
M. Mayich et al.
Finally, while aneurysm recurrence is a pri-
mary indication for surveillance imaging, it is
also important to consider the possibility of
de novo aneurysms or enlargement of
concurrent aneurysms that were not treated
on follow-up cerebrovascular imaging.
Tsutsumi et al. reported an annual risk of 0.89 %
of de novo aneurysm formation when they studied
a series of 112 patients treated for 140 clipped
aneurysms. This translates to a cumulative risk
of 10 % at 9 years, prompting the recommenda-
tion that follow-up is warranted at 9 years even in
the context of completely clipped aneurysms
(Fig. 18) [5]. Alternatively, the incidence of de
novo aneurysm formation is estimated to be
1.5 %, and approximately 20 % of patients with
treated aneurysms have additional cerebral aneu-
rysms, of which a minority of these demonstrates
growth on 5-year follow-up imaging [51].
27 Postop Evaluation of Aneurysms (Including Vasospasm)
Fig. 18 Aneurysm follow-up. The patient has a history of
right paraclinoid aneurysm, status post right pterional cra-
niotomy for clipping. Follow-up CTA images show
Nevertheless, these lesions infrequently require
treatment. These observations have led to the sug-
gestion of a simplified algorithm for aneurysm
follow-up that all patients should be followed with
MRA at 6 months and if satisfactorily occluded are
considered cured. Those not satisfactorily occluded
could be re-treated at that point or undergo contin-
ued imaging follow-up. Continued follow-up may
be warranted in higher-risk patients, including
those with giant aneurysms, multiple aneurysms,
positive family history of aneurysm, heritable dis-
ease associated with aneurysm formation, or with
proven growth of additional aneurysms to assess
both the treated lesion and surveillance for devel-
opment of any additional lesions.
Summary
Imaging is indispensable for the evaluation of
aneurysms after both endovascular and microsur-
gical treatment. Multiple imaging modalities are
available, each with advantages and disadvan-
tages. In general CT is routinely obtained after
aneurysm clipping to evaluate for major compli-
cations, such as hemorrhage, stroke, and hydro-
cephalus, while MRI is performed if CT does not
reveal anything to explain the patient’s symptoms.
Furthermore, CTA is preferable for imaging resid-
ual or recurrent aneurysms after aneurysm clip-
ping, while MRA is suitable for imaging after
599
stability of an additional left cavernous carotid artery aneu-
rysm (arrowheads) but increase in size of a partially
thrombosed fusiform basilary artery aneurysm (arrows)
aneurysm coiling. Ultimately, conventional cathe-
ter angiography remains the gold standard imaging
modality if there are any questions that cannot be
answered with the noninvasive imaging tech-
niques. It is important to recognize the expected
and complicated posttreatment imaging findings.
References
1. Bederson JB, Connolly ES Jr, Batjer HH et al (2009)
American Heart Association. Guidelines for the man-
agement of aneurismal subarachnoid hemorrhage: a
statement for healthcare professionals from a special
writing group of the Stroke Council, American Heart
Association. Stroke 40(3):994–1025
2. Molyneux AJ, Kerr RSC, Yu L-M, Clarke M,
Sneade M, Yarnold JA et al (2005) International sub-
arachnoid aneurysm trial (ISAT) of neurosurgical clip-
ping versus endovascular coiling in 2143 patients with
ruptured intracranial aneurysms: a randomised com-
parison of effects on survival, dependency, seizures,
rebleeding, subgroups, and aneurysm occlusion. Lan-
cet 366(9488):809–817
3. Ginat DT, Smith ER, Robertson RL, Scott RM,
Schaefer PW (2013) Imaging after direct and indirect
extracranial bypass surgery. AJR Am J Roentgenol
201(1):W124–W132
4. Deutschmann HA, Wehrschuetz M, Augustin M,
Niederkorn K, Klein GE (2012) Long-term follow-up
after treatment of intracranial aneurysms with the Pipe-
line embolization device: results from a single center.
AJNR Am J Neuroradiol 33(3):481–486. doi:10.3174/
ajnr.A2790
5. Tsutsumi K, Ueki K, Morita A, Usui M, Kirino T
(2001) Risk of aneurysm recurrence in patients with
600
clipped cerebral aneurysms: results of long-term
follow-up angiography. Stroke 32(5):1191–1194
6. David CA, Vishteh AG, Spetzler RF, Lemole M, Law-
ton MT, Partovi S (1999) Late angiographic follow-up
review of surgically treated aneurysms. J Neurosurg
91(3):396–401
7. Wallace RC, Karis JP, Partovi S et al (2007) Noninva-
sive imaging of treated cerebral aneurysms. Part I. MR
angiographic follow-up of coiled aneurysms. AJNR
Am J Neuroradiol 28:1001–1008
8. Shinohara Y, Sakamoto M, Iwata N, Kishimoto J,
Kuya K, Fujii S, Kaminou T, Watanabe T, Ogawa T
(2014) Usefulness of monochromatic imaging with
metal artifact reduction software for computed tomog-
raphy angiography after intracranial aneurysm coil
embolization. Acta Radiol 55:1015–1023.
doi:10.1177/028418511351049
9. Phan CM, Yoo AJ, Hirsch JA et al (2012) Differentia-
tion of hemorrhage from iodinated contrast in different
intracranial compartments using dual-energy head
CT. AJNR Am J Neuroradiol 33:1088–1094
10. Kaufmanna TJ, Huston J III, Cloft HJ, Mandrekar J,
Gray L, Bernstein MA, Atkinson JL, Kallmes DF
(2010) A prospective trial of 3T and 1.5T time-of-flight
and contrast-enhanced MR angiography in the
follow-up of coiled intracranial aneurysms. AJNR
31:912–918
11. Olsen RV, Munk PL, Lee MJ et al (2000) Metal artifact
reduction sequence: early clinical applications. Radio-
graphics 20(3):699–712
12. Walker MT, Tsai J, Parish T et al (2005) MR angio-
graphic evaluation of platinum coil packs at 1.5T and
3T: an in vitro assessment of artifact production: tech-
nical note. Am J Neuroradiol 26:848–853
13. Gonner F, Heid O, Remonda L, et al. (1998) MR
angiography with ultrashort echo time in cerebral aneu-
rysms treated with Guglielmi detachable coils. AJNR
Am J Neuroradiol 19:1324–1328
14. Kaufmann TJ, Huston J III, Cloft HJ et al (2010) A
prospective trial of 3T and 1.5T time-of-flight and
contrast-enhanced MR angiography in the follow-up
of coiled intracranial aneurysms. Am J Neuroradiol
31:912–918
15. Leffers AM, Wagner A (2000) Neurologic complica-
tions of cerebral angiography. A retrospective study of
complication rate and patient risk factors. Acta Radiol
41:204–210
16. Cloft HJ, Joseph GJ, Dion JE (1999) Risk of cerebral
angiography in patients with subarachnoid hemor-
rhage, cerebral aneurysm, and arteriovenous malfor-
mation. A meta-analysis. Stroke 30(2):317–320
17. Hanley M, Zenzen WJ, Brown MD, Gaughen JR,
EvansHanley AJ et al (2008) Comparing the accuracy
of digital subtraction angiography, CT angiography
and MR angiography at estimating the volume of cere-
bral aneurysms. Interv Neuroradiol 14(2):173–177
18. Kassell N, Sasaki T, Colohan A, Nazar G (1985) Cere-
bral vasospasm following aneurysmal subarachnoid
hemorrhage. Stroke 16(4):562–572
M. Mayich et al.
19. Dankbaar JW, Slooter A, Rinkel G, Schaaf IC (2010)
Effect of different components of triple-H therapy on
cerebral perfusion in patients with aneurysmal sub-
arachnoid hemorrhage: a systematic review. Crit Care
14(1):R23
20. Wilson DA, Nakaji P, Abla AA, Uschold TD, Fusco
DJ, Oppenlander ME et al (2012) A simple and quan-
titative method to predict symptomatic vasospasm after
subarachnoid hemorrhage based on computed tomog-
raphy: beyond the Fisher scale. Neurosurgery
71(4):869–875
21. Aaslid R, Markwalder T-M, Nornes H (1982) Nonin-
vasive transcranial Doppler ultrasound recording of
flow velocity in basal cerebral arteries. J Neurosurg
57(6):769–774
22. Lindegaard K-F, Nornes H, Bakke SJ, Sorteberg W,
Nakstad P (1988) Cerebral vasospasm after subarach-
noid hemorrhage investigated by means of transcranial
Doppler ultrasound. In: Proceedings of the 8th
European Congress of Neurosurgery, 6–11 September
1987. Springer, Barcelona
23. Lysakowski C, Walder B, Costanza MC, Tramèr MR
(2001) Transcranial Doppler versus angiography in
patients with vasospasm due to a ruptured cerebral aneu-
rysm a systematic review. Stroke 32(10):2292–2298
24. Anderson GB, Ashforth R, Steinke DE, Findlay JM
(2000) CT angiography for the detection of cerebral
vasospasm in patients with acute subarachnoid hemor-
rhage. Am J Neuroradiol 21(6):1011–1015
25. Chappell ET, Moure FC, Good MC (2003) Compari-
son of computed tomographic angiography with digital
subtraction angiography in the diagnosis of
cerebral aneurysms: a meta-analysis. Neurosurgery
52(3):624–631
26. van Rooij WJ, Sluzewski M, Beute GN, Nijssen PC
(2006) Procedural complications of coiling of ruptured
intracranial aneurysms: incidence and risk factors in a
consecutive series of 681 patients. Am J Neuroradiol
27:1498–1501
27. Sluzewski M, van Rooij WJ (2005) Early rebleeding
after coiling of ruptured cerebral aneurysms: incidence,
morbidity, and risk factors. Am J Neuroradiol
26:1739–1743
28. Johnston SC, Dowd CF, Higashida RT, Lawton MT,
Duckwiler GR, Gress DR, CARAT Investigators
(2008) Predictors of rehemorrhage after treatment of
ruptured intracranial aneurysms: the Cerebral Aneu-
rysm Rerupture After Treatment (CARAT) study.
Stroke 39(1):120–125
29. Naggara ON, White PM, Guilbert F et al (2010)
Endovascular treatment of intracranial unruptured
aneurysms: systematic review and meta-analysis of
the literature on safety and efficacy. Radiology
256:887–897
30. Rordorf G, Bellon RJ, Budzik RE Jr et al (2001) Silent
thromboembolic events associated with the treatment
of unruptured cerebral aneurysms by use of Guglielmi
detachable coils: prospective study applying diffusion-
weighted imaging. Am J Neuroradiol 22:5–10
27 Postop Evaluation of Aneurysms (Including Vasospasm)
31. Kang D-H et al (2013) MR-DWI–positive lesions and
symptomatic ischemic complications after coiling of
unruptured intracranial aneurysms. Stroke 44:789–791
32. Altay T, Kang HI, Woo HH et al (2011) Thromboem-
bolic events associated with endovascular treatment of
cerebral aneurysms. J Neurointerv Surg 3:147–150
33. Hahnemann ML, Ringelstein A, Sandalcioglu IE,
Goericke S, Moenninghoff C, Wanke I, Forsting M,
Sure U, Schlamann M (2014) Silent embolism after
stent-assisted coiling of cerebral aneurysms: diffusion-
weighted MRI study of 75 cases. J Neurointerv Surg
6:461–465
34. Mocco J, Fargen KM, Albuquerque FC, Bendok BR
et al (2011) Delayed thrombosis or stenosis following
enterprise-assisted stent-coiling: is it safe? Midterm
results of the interstate collaboration of enterprise
stent coiling. Neurosurgery 69(4):908–914
35. Albuquerque FC, Levy EI, Turk AS, Niemann DB,
Aagaard-Kienitz B, Pride GL Jr et al (2008) Angio-
graphic patterns of Wingspan in-stent restenosis. Neu-
rosurgery 63(1):23–27
36. Song JK, Niimi Y, Fernandez PM, Brisman JL,
Buciuc R, Kupersmith MJ, Berenstein A (2004)
Thrombus formation during intracranial aneurysm
coil placement: treatment with intra-arterial abciximab.
Am J Neuroradiol 25:1147–1153
37. Derdeyn CP, Cross DT III, Moran CJ, Brown GW
et al (2002) Postprocedure ischemic events after treat-
ment of intracranial aneurysms with Guglielmi detach-
able coils. J Neurosurg 96(5):837–843
38. Cognard C, Weill A, Castaings L et al (1998) Intracra-
nial berry aneurysms: angiographic and clinical results
after endovascular treatment. Radiology
206:499–5108
39. Kanaan H, Jankowitz B, Akeu A, Kostov D et al (2010)
In-stent thrombosis and stenosis after neck-remodeling
device-assisted coil embolization of intracranial aneu-
rysms. Neurosurgery 67(6):1523–1533
40. McLaughlin N, McArthur D, Martin D (2013) Use of
stent-assisted coil embolization for the treatment of
wide-necked aneurysms: a systematic review. Surg
Neurol Int 4:43
41. Fiorella D, Kelly ME, Moskowitz S, Masaryk TJ
(2009) Delayed symptomatic coil migration after
601
initially successful balloon-assisted aneurysm coiling:
technical case report. Neurosurgery 64(2):E391–E392;
discussion E392
42. Bhatti MT, Peters KR, Firment C, Mericle RA (2004)
Delayed exacerbation of third nerve palsy due to aneu-
rysmal regrowth after endovascular coil embolization.
J Neuroophthalmol 24(1):3–10
43. Murayama Y, Nien YL, Duckwiler G, Gobin YP,
Jahan R, Frazee J et al (2003) Guglielmi detachable
coil embolization of cerebral aneurysms: 11 years’
experience. J Neurosurg 98(5):959–966
44. Ferns SP, Sprengers ME, van Rooij WJ et al (2009)
Coiling of intracranial aneurysms: a systematic review
on initial occlusion and reopening and retreatment
rates. Stroke 40:e523–e529
45. Cognard C, Weill A, Spelle L et al (1999) Long-term
angiographic follow-up of 169 intracranial berry aneu-
rysms occluded with detachable coils. Radiology
212:348–356
46. van Rooij WJ, Sluzewski M (2009) Opinion: imaging
follow-up after coiling of intracranial aneurysms. Am J
Neuroradiol 30:1646–1648
47. Sluzewski M, van Rooij WJ, Rinkel GJ et al (2003)
Endovascular treatment of ruptured intracranial aneu-
rysms with detachable coils: long-term clinical and
serial angiographic results. Radiology 227:720–724
48. Sprengers ME, Schaafsma J, van Rooij WJ et al (2008)
Stability of intracranial aneurysms adequately
occluded 6 months after coiling: a 3T MR angiography
multicenter long-term follow-up study. Am J
Neuroradiol 29(9):1768–1774
49. Schaafsma JD, Sprengers ME, van Rooij WJ
et al (2009) Long-term recurrent subarachnoid hemor-
rhage after adequate coiling versus clipping of ruptured
intracranial aneurysms. Stroke 40:1758–6
50. McLaughlin N, Bojanowski MW (2004) Early
surgery-related complications after aneurysm clip
placement: an analysis of causes and patient outcomes.
J Neurosurg 101(4):600–606
51. Sprengers ME, van Rooij WJ, Sluzewski M, Rinkel GJ,
Velthuis BK, de Kort GA, Majoie CB (2009) MR
angiography follow-up 5 years after coiling: frequency
of new aneurysms and enlargement of untreated aneu-
rysms. AJNR Am J Neuroradiol 30(2):303–307
 Part VI
Other CNS Vascular Diseases
 Brain Arteriovenous Malformations
28
Daniel Sahlein and Nathan Manning

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605 Brain arteriovenous malformations (bAVMs)
are relatively rare complex vascular lesions
bAVM Natural History: The Etiology
of Controversy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 606
involving a shunt between the artery and vein
Estimating Hemorrhage Risk in an Individual within the subpial space. This physiology can
bAVM: The Need for An result in hemorrhage, which can cause severe
Individualized Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608 neurological morbidity and death. There is tre-
Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 611 mendous controversy surrounding appropriate
The Brain Arteriovenous Malformation . . . . . . . . . . 612
treatment, particularly for unruptured bAVMs,
Embryology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 612 which stems at least in part from debate as to
Pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 614 the natural history of bAVMs as well as vari-
Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 622 ability from lesion to lesion. This chapter will
Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 631
review bAVM imaging and in doing so will
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 635 summarize the origins of the current debate
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 635 over bAVM treatment utilizing a detailed
review of the natural history literature. Further-
more, this chapter will discuss bAVM epide-
miology, embryology, and treatment.

Keywords
Brain arteriovenous malformation • Stroke •
Intracranial hemorrhage • Vascular disorders •
Embolization • Microsurgery • Gamma Knife
surgery • Angiography • CT angiography • MR
angiography

D. Sahlein (*)
Goodman Campbell Brain and Spine, Indianapolis, Introduction
IN, USA
e-mail: dsahlein@goodmancampbell.com
A brain arteriovenous malformation (bAVM) is
N. Manning
defined as a direct communication from an artery
Department of Neurological Surgery, Columbia University
Medical Centre, New York, NY, USA to a vein through an intervening “nidus” which is
e-mail: dr.nathan.manning@gmail.com located in the subpial meningeal space [1]. Brain
# Springer Science+Business Media New York 2016 605
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_21
606
arteriovenous malformations are relatively rare
vascular lesions but nevertheless represent as
high as 33.5 % of intracerebral hemorrhage in
patients under the age of 40 [2]. There remains
considerable controversy with regard to their
pathogenesis, natural history, and optimal man-
agement. Controversies notwithstanding, it
seems likely these lesions form during early
embryologic development. Arteriovenous
malformations are nonneoplastic; however, they
are highly dynamic as a result of molecular sig-
naling molecules intimately involved in
vasculogenesis and angiogenesis. Brain arteriove-
nous malformations have a complex natural his-
tory, the predictors of which have not entirely
been elucidated but are likely to be dependent
upon bAVM architecture which may be under-
stood through neuroimaging, and may evolve
over time. The diagnostic radiologist can there-
fore play a critical role in understanding bAVMs
and in contributing to an understanding of both
natural history and treatment-related risk.
Treatment is not without risk and is a source of
recent controversy, particularly with respect to the
management of unruptured bAVMs [3]. Neverthe-
less, treatment risk has been extensively studied,
and a prospectively tested pretreatment risk strat-
ification scale is available for microsurgical resec-
tion [4, 5] though not for endovascular
embolization [6] likely as a result of fundamental
differences in these two treatment modalities.
This chapter will review the fundamentals of
bAVM formation, natural history, and treatment
with an eye toward diagnostic neuroradiology.
bAVM Natural History: The Etiology
of Controversy
Few subjects in neurovascular disease are as con-
troversial or as current as the debate over optimal
bAVM management. Much of the debate sur-
rounding bAVMs is the result of the publication
of A Randomized trial of Unruptured Brain AVMs
(ARUBA) in the Lancet in 2014 [3]. The ARUBA
trial was a multicenter, randomized, non-blinded
trial that enrolled patients with unruptured
bAVMs into one of two arms: medical
D. Sahlein and N. Manning
management with an intervention (surgery, embo-
lization, and/or Gamma Knife) or medical man-
agement alone. Randomization began in April of
2007 and was stopped in April of 2013 when a
data and safety monitoring board recommended
halting randomization because of superiority of
the medical management group. Specifically,
after a mean follow-up of 33.3 months (with a
standard deviation of almost 20 months), the pri-
mary outcome of death or symptomatic stroke was
reached by 10.1 % of patients in the medical
management arm and 30.7 % in the intervention
group. Without getting too enmeshed in the details
of this study, or why the study may have been
biased toward the medical therapy arm, the very
existence of this study raises a broader question as
to why such a basic aspect of bAVM management
is still open to debate after decades of rigorous
study. Why was an NIH-sponsored study enroll-
ing unruptured bAVMs in a randomized trial com-
paring intervention versus medical therapy as late
as 2007? Nowhere is the answer to this question
clearer than in the variability and lack of clarity
provided by the bAVM natural history literature.
Determining the true natural history of bAVMs
is a daunting task to be sure. First, bAVMs are
fairly rare lesions, occurring with an incidence of
approximately 1 per 100,000, and, second, they
account for only around 4 % of all intracranial
hemorrhage [7]. However, the seemingly precise
numbers quoted in the previous sentence belie the
true controversy that exists in this field and mask
the variability observed from paper to paper with
regard to disease incidence, prevalence, and rup-
ture risk (see below for a thorough discussion of
the epidemiology). Very quickly does a review of
the bAVM literature become more a study in epis-
temology and less an exploration of a complex
vascular lesion. The lack of a consensus under-
standing of bAVM natural history is the direct
result of the complexity of studying a rare lesion
and the variability in the disease from patient to
patient that manifests as inconsistency between
papers. This lack of clarity convolutes counseling
of individual patients in ways that will be
discussed in detail below.
The number of citations that a paper receives
gives a general sense of its relative value to the
28 Brain Arteriovenous Malformations
biomedical community [8]. The five most cited
bAVM natural history papers in order of publica-
tion year are Fults et al. [9] published in 1984,
Crawford et al. [10] from 1986, Brown et al. [11]
from 1988, Ondra et al. [12] from 1990, and Mast
et al. [13] from 1997. While the Ondra study
found a remarkably stable per annum risk of hem-
orrhage independent of clinical presentation and
time from presentation of 3.9–4.0 % [12], there is
a tremendous variability in hemorrhage rate per
annum from study to study ranging from 0 % to
32.9 % [13] depending on the type of presentation
and time elapsed. Furthermore, the associated
mortality resulting from a hemorrhage is remark-
ably variable ranging from 0 % [13] to 29 % [11]
per hemorrhage. The incidence of significant neu-
rological morbidity is similarly variable, ranging
from 0.4 % per annum in one subset of the Fults
cohort [9] to 3.5–4.7 % per annum in patients
presenting with hemorrhage [9, 11]. This degree
of variability makes it difficult if not impossible to
accurately determine the natural history risk in an
individual patient. Many practitioners take some
amalgam of this data and suggest a 2–4 % per
annum risk of hemorrhage and discuss the poten-
tial associated morbidity and mortality at least
partially based on bAVM location [9–13].
However, a closer examination of these papers
as “natural history” studies would suggest that
attempts at creating an informal average or formal
meta-analysis would not necessarily result in a
more accurate calculation of natural history
despite a theoretical increase in precision, a subtle
but critical distinction. In each of the above
referenced natural history papers, a cohort of
patients was treated immediately and therefore
excluded from the natural history observation,
leading to serious questions about the generaliz-
ability of the findings. A not-insignificant 36.6 %
[9], 36.7 % [10], and 36 % [12] of patients were
treated immediately and excluded from three of
the five studies. Furthermore, those patients
treated were statistically significantly different
from those who were included in the natural his-
tory analysis. Patients in the nonsurgical arms
tended to have larger and deep bAVMs [9] in
one study; tended to be left-sided, cross the mid-
line, larger, deeper, and more posterior bAVMs
607
[10] in another; and tended to be larger, deeper,
and closer to eloquent cortex in a third study
[12]. The Brown et al. paper states that patients
treated immediately were excluded from natural
history analysis but gives no data on the number
of patients or differences between the treated and
followed cohorts. The Mast et al. paper included
all patients until they were treated, creating huge
ranges in the duration of follow-up in the study of
0.1–96.2 months in the hemorrhage arm and
0.4–90.9 months in the non-hemorrhage arm. In
other words, some patients were contributing
almost 1000 the patient-year data in the study
than others. Furthermore, there is no discussion of
the differences in the lesions or patient demo-
graphics in the patients treated within days of
presentation and those who were followed for
months or years, and this was not a randomized
study. In summary, each paper for which data is
given excludes approximately 40 % of the cohort
from the natural history analysis, and excluded
bAVMs are significantly different from those
that were included. It is from these data that prac-
titioners are estimating risk and determining
whether treatment is necessary, i.e., calculating a
risk/benefit ratio for intervention. Without even
delving into a discussion of treatment risk, it is
not surprising that well-informed clinicians come
to very different conclusions about appropriate
clinical management [14, 15].
For the sake of general approximation, two
recent papers would suggest that the average of a
2–4 % per annum risk of hemorrhage is within a
reasonable range. The prospective, randomized
ARUBA study [3] found a 2.2 % per annum risk
of spontaneous rupture, and a recent meta-
analysis which harmonized data from multiple
large databases to include over 6000 patient-
years of follow-up found an annual 2.3 % risk of
hemorrhage with a 95 % confidence interval of
2.0–2.7 % [16]. The latter study however elimi-
nated patients once treated using a similar meth-
odology to the Mast paper [13] and like Mast
et al. gives no data about the differences between
patients treated early and those followed.
However, even if the 2–4 % per annum risk of
hemorrhage is accurate over large numbers of
patients, variability between studies regarding
608
both hemorrhage risk and morbidity/mortality
raises a critical question. Can these averages
truly be applied to an individual patient in the
context of this variability? In other words, is this
“average” approach appropriate for making treat-
ment decisions in individual cases when disease
variability renders such averages likely to be
highly inaccurate for an individual patient? Mod-
ern prospective data would suggest significant
disability in approximately 33 % of patients fol-
lowing an initial hemorrhage and in 45 % of
patients following a second hemorrhage [17],
and older studies suggest mortality of up to 29 %
[11] per hemorrhage, considerable numbers for a
disease that presents on average in patients in their
early 30s.
Estimating Hemorrhage Risk
in an Individual bAVM: The Need for An
Individualized Approach
Estimating hemorrhage risk in an individual
bAVM is particularly critical in calculating a
risk/benefit ratio for therapy. A significant litera-
ture has attempted to address this issue [9–11, 13,
18–40], and two meta-analyses have studied the
aggregate data more recently [41, 42]. Many fea-
tures of bAVMs have been found to be statistically
significantly associated with hemorrhage presen-
tation. However, there is no proffered hypothesis
for why many of these bAVM features might
increase rupture risk [43]. For example, bAVM
deep location [10, 32, 35, 36, 41], bAVM nidal
size [10, 18, 21, 23, 26, 28, 35–39, 41, 42], deep/
only deep venous drainage [13, 18, 20–22, 27–30,
36, 37, 40–42, 44], ventricular/periventricular
location [19, 22], posterior fossa location [9, 28],
and arterial nonborderzone location [26] have all
been statistically significantly associated with
rupture risk without any compelling explanation
as to causality. In the absence of a clear rationale
as to why these factors are associated with hem-
orrhage, clinicians cannot confidently use them to
predict hemorrhage prospectively [43].
However, one study lays the foundation for a
model of bAVM rupture risk based on parameters
with clear physiologic causality [43, 45]. In a
D. Sahlein and N. Manning
retrospective single-center study, 121 patients
with 122 bAVMs at NYU Langone Medical Cen-
ter from 1996 to 2006 were entered into the anal-
ysis. All patients received full diagnostic
angiography with superselective microcathe-
terization [43]. Detailed angioarchitectural data
included bAVM location, depth, nidus size (mea-
sured in three planes), bAVM flow physiology
(nidus predominant, fistula predominant, or
mixed – see below for further details), arterial
supply (organized by major feeding territory and
categorized as cortical or perforator), associated
aneurysms (number and location characterized as
nidal, flow-related, proximal [circle of Willis],
venous), pial collaterals (described by whether
the collaterals originated from the same arterial
distribution [i.e., middle cerebral artery (MCA) to
middle cerebral artery] or different [i.e., anterior
cerebral artery (ACA) to middle cerebral artery]),
bAVM border morphology (compact versus
diffuse – see below for a detailed explanation),
the presence of moyamoya-type changes, venous
characteristics (number of draining veins, superfi-
cial versus deep drainage, the presence of venous
angiopathy), and bAVM eloquence as defined by
Spetzler and Martin [5]. Indisputable intracranial
hemorrhage was confirmed by computed tomog-
raphy scan in all patients [43].
Interestingly, bAVMs with a single draining
vein were more likely to present with hemorrhage
than those with two or more draining veins
[43]. This is not the first study to find this rela-
tionship between hemorrhage risk and a single
draining vein [19, 22, 34, 38, 39]. However, it is
the first to test the physiologic implications of this
finding by analyzing different venous outflow
conditions. The study hypothesized that bAVMs
with a single vein were more often associated with
hemorrhage because of restricted outflow imped-
ance. This hypothesis implies that bAVMs with
multiple draining veins would be protected from
hemorrhage by lower outflow impedance. Vari-
ability in the venous drainage of patients with
two or more draining veins provided an opportu-
nity to test this hypothesis. Specifically, the study
looked at bAVMs with two or more draining veins
with significant venous stenosis and those with-
out. Interestingly, those bAVMs with multiple
28 Brain Arteriovenous Malformations
draining veins and a stenosis had significant asso-
ciation with hemorrhage with an odds ratio very
similar to those with one draining vein, thereby
validating the hypothesis [43].
The study also found that small nidal size, deep
location, exclusively deep venous drainage, and
the absence of pial-to-pial collaterals were signif-
icantly associated with hemorrhage and venous
outflow stenosis and the presence of any aneu-
rysm showed a trend toward significance in the
univariate analysis. Only those potentially physi-
ologic parameters were added to the multivariate
analysis (the number of draining veins (one versus
Fig. 1 A 29-year-old female presented to the ER with
severe headaches. She was found to have a small
paramedian right parietal intracerebral hemorrhage. Sagit-
tal CTA (a) shows an enlarged paramedian vein (white
arrowheads). One can almost make out a nidus (black
arrowhead). AP (b) and lateral (c) angiography show an
enlarged anterior cerebral pericallosal artery (white
609
two or more), the presence of stenosis, and the
presence of any aneurysm). Not surprisingly, all
were statistically significant in the multivariate
logistic regression analysis with adjusted odds
ratios ranging from 2.4 to 6.6 (one draining vein,
6.6; venous stenosis, 2.6; the presence of aneu-
rysm, 2.4). Refer to Figs. 1 and 2 as examples of
bAVMs with one draining vein versus multiple
draining veins and a stenosis.
Furthermore, those variables without clear
physiologic implication that were significantly
associated with hemorrhage in the univariate anal-
ysis were assessed as confounders using
arrows) and superior parietal artery (yellow arrowhead)
supplying a subcentimeter bAVM (black arrowhead).
There is a single paramedian superficial parietal draining
vein which drains into the superior sagittal sinus. (blue
arrowheads). The patient was treated with open surgical
resection
610
Fig. 2 A 46-year-old male was evaluated with
noncontrast head CT following a seizure. Transaxial
noncontrast head CT (a) shows a right angular cortical
and subcortical hyperdensity (white arrowheads). Two
sagittal precontrast T1-WI images (b, c) show the lesion
spanning the cortices of the superior and middle temporal
gyri (arrowhead) and involving the angular gyrus. Right
ICA angiograms in the lateral projection in the early
agreement analysis. In each case, the less clearly
physiologic parameter was highly associated with
one of the more plausibly physiologic parameters,
confirming that the less physiologic parameters
were acting as confounders. Small nidal size, a
characteristic without clear physiologic implica-
tion, for example, showed 80 % agreement with
one draining vein. Likewise, the presence of only
deep venous drainage showed an 84 % agreement
with one draining vein. The absence of pial-to-pial
collaterals from different distributions had a 71 %
agreement with one draining vein. Deep location
had a 77 % agreement with one draining vein.
These results confirmed that various bAVM char-
acteristics with less plausible physiological
D. Sahlein and N. Manning
arterial and mid arterial phases (d and e, respectively).
The mid arterial phase depicts an enlarged angular branch
of the right MCA (black arrowhead) terminating within the
nidus (white arrowheads). Venous drainage is into two
dominant veins: the veins of Labbe and Trolard. There is
a venous ectasia on Labbe (white arrow) and stenosis
(black arrow) associated with the vein of Trolard
implications likely serve as confounders in
assessing hemorrhagic risk, thereby validating the
selection of parameters with potential physiologi-
cal implications for the multivariate analysis [43].
Not only does this study explain many of the
literature findings of bAVM features associated
with hemorrhage without clear physiologic impli-
cation, it paves the way for a patient-specific
analysis of bAVMs from the standpoint of natural
history risk assessment [45]. The diagnostic radi-
ologist’s role in assessing for venous stenosis,
number of draining veins, and aneurysms also
becomes critical. Future studies will elaborate on
and refine this model. While this model uses ana-
tomic features as physiologic surrogates, with
28 Brain Arteriovenous Malformations
inherent limitations, new MRI modalities have
tremendous potential to directly interrogate
bAVM physiology [46, 47].
Epidemiology
The nature of the bAVM disease process, physi-
cian and institutional tendency, and difficulties in
case detection lead to significant bias in the avail-
able epidemiological data and result in persistent
uncertainty with regard to the epidemiology and
natural history of bAVMs. The first population-
based study to address the incidence and preva-
lence of bAVMs (in addition to intracranial vas-
cular malformations (IVMs)) as a whole was the
Olmsted County study [48]. Medical records from
Olmstead County, Minnesota, were reviewed for
all IVMs from 1965 to 1992. This study demon-
strated an incidence of 2.05 per 100,000 person-
years. The vast majority (85 %) of detected
bAVMs were symptomatic, and the mean age at
diagnosis was ~40 years. The authors noted an
increasing detection rate during the course of the
study that was presumably the result of improved
understanding of bAVMs and technological
advances such as improved medical imaging. In
a population-based study from the state of West-
ern Australia, a considerably lower incidence of
0.89/100,000 person-years was determined with a
prevalence of only 5.47/100,000 [49]. In this pop-
ulation, the mean age at diagnosis was ~34 years,
approximately 60 % of patients presented with
hemorrhage and 30 % presented with seizures.
Only 2.5 % of bAVMs in this study were inciden-
tal suggesting a degree of under-detection, a met-
ric that has implications for our understanding of
natural history. The authors of the Western
Australia study noted that the incidence of
bAVMs rose during the study period peaking at
1.46/100,000 person-years in 1989, exceeding the
population rise. However, the incidence decreased
from this point in spite of increasing availability
of CT and MRI. The authors conclude that this
trend argues for a low incidence of occult bAVM;
however, this remains speculative. Interestingly,
the authors also noted that headache, as a
presenting symptom, did not exceed the incidence
611
of headache in the general population, suggesting
that this is an unlikely attributable symptom of
bAVMs, a controversial conclusion.
The Scottish Intracranial Vascular Malforma-
tion Study (SIVMS) [50] was the first large
population-based, prospective study of bAVM
epidemiology. It used multiple sources of case
ascertainment over a prospective 24-month period
to identify “first-ever-in-life” diagnosis of any
intracranial vascular malformation in Scottish
adults. In this study, the incidence of bAVM was
1.12/100,000 person-years with a hemorrhagic
presentation incidence of 0.89/100,000 person-
years and a nonhemorrhagic presentation inci-
dence of 0.23/100,000 person-years. Two addi-
tional population-based studies from North
America, the New York Islands AVM Study and
data from the Northern Manhattan Stroke Study
(NOMASS), were published in 2003 and 2002,
respectively [44, 51]. The NOMASS found three
patients with bAVMs out of 207 patients with
first-ever intracranial hemorrhage, suggesting
that bAVMs account for 1.4 % of first-ever intra-
cranial hemorrhage in an unselected population
[44]. The estimated crude incidence for first-ever
bAVM hemorrhage was 0.55/100,000 person-
years. However, two of the three discovered
bAVMs were in the posterior fossa, disproportion-
ately high with respect to bAVM location, thereby
casting doubt as to how representative the data is
and, by extension, the generalizability of the
study. In spite of these concerns, the New York
Islands AVM Study produced similar results
[51]. The New York Islands AVM Study encoun-
tered 284 AVM patients from a total of
21,216,467 person-years of observation for an
incidence of 1.34/100,000 person-years. The
crude incidence for AVM-related hemorrhage of
0.51/100,000 person-years was remarkably simi-
lar to that from the NOMASS data set. The mean
age of bAVM rupture was ~31 years.
A Swedish population-based cohort study also
reported similar findings with a bAVM incidence
of 1.24/100,000 person-years [52]. Hemorrhage
was by far the most common presentation seen in
~70 % with a presentation incidence of 0.87/
100,000 person-years. Like others before, these
authors concluded that incidental bAVMs are rare.
612
The issue of incidental bAVMs has been
addressed, at least obtusely, by several population
MRI studies. A German Air Force recruit study in
which 2,536 males aged 17–35 years old
underwent brain MRI as part of medical screening
for flying service found bAVMs in a surprising
0.2 % [53]. While this is significantly higher than
previous clinical studies, it is within the range
reported in autopsy series, as summarized in a
review of the epidemiology of bAVMs by Berman
et al. [54]. However, several other MRI screening
studies have failed to identify any patients with
bAVM [55–57]. This may in part be due to the
younger age of participant screened in the German
study [53] in that older individuals are less likely
to harbor an unruptured or otherwise asymptom-
atic bAVM [10]. Furthermore, the imaging param-
eters appear to be more rigorous in the German
study.
There are considerable challenges associated
with estimating epidemiology in a rare and poten-
tially occult disease. Appropriate studies require
adequate radiological and/or pathological investi-
gation to reliably diagnose bAVMs as well as to
exclude other entities that may have a similar
presentation such as cavernous malformations or
dural arteriovenous fistulas (dAVF) [7]. While this
seems reasonably achievable, clinicians may
forgo aggressive investigation due to a myriad of
reasons including clinical urgency or futility.
Pathological investigations may lack sufficient
zeal given the minor role bAVMs may play on a
population basis in the theater of intracranial hem-
orrhage. Representative cohorts should preferably
include individuals diagnosed at large referral
centers as well as small peripheral hospitals in
addition to those who have died in the community.
To this end, many of the larger, population-based
trials would be expected to be fairly accurate.
While the challenge is significant, some comfort
can be taken by the relatively similar results
reported by large, population-based studies
[48, 58]. The epidemiology data prior to 2001 is
well reviewed by Al-Shahi and Warlow [7]. The
authors reported a bAVM incidence of 1/100,000
person-years and a prevalence of 18/100,000.
Rupture of a bAVM was estimated to account for
1–2 % of all strokes and 4 % of intracerebral
D. Sahlein and N. Manning
hemorrhage. This figure rises to as high as
33.5 % of hemorrhages in young adults (< age
40). Contrary to later reports, Al-Shahi and
Warlow estimated that 15 % of bAVMs are
asymptomatic at the time of diagnosis. However,
in line with later studies, they estimated that
approximately 60–70 % present with hemorrhage
and 20 % with seizures. A more recent study has
provided updated data [59]. Review of a 10-year
period of Kaiser Permanente Medical Care Pro-
gram in northern California medical records, radi-
ology reports, and administrative databases
yielded an overall bAVM detection rate of 1.42/
100,000 person-years [59].
The Brain Arteriovenous Malformation
Embryology
It is likely that bAVMs develop in utero, during
the vascular development of the central nervous
system. An understanding of bAVM anatomy and
physiology therefore requires a basic understand-
ing of the embryology of the neurovasculature.
The neural tissue and vascular tissue develop
in concert in the early embryo. Two processes are
responsible for the vascularization of the central
nervous system and indeed the developing fetus:
vasculogenesis, the formation of the primitive
endothelial cells [60], and angiogenesis, which
follows vasculogenesis and is responsible for
remodeling and expansion of the network
[61]. While there remains some debate based on
rare observations, bAVMs are generally consid-
ered to be congenital lesions that likely form, at
least as immature precursors, during angiogenesis
of the central nervous system [62]. It is impossible
to convey the current theories on bAVM patho-
physiology without at least an introduction to the
concepts of vasculogenesis and angiogenesis as
they pertain to the central nervous system. The
following sections will attempt to cover these
complicated subjects in an expedient fashion.
Numerous texts and reviews are available which
cover these processes in greater detail [1, 63–65].
Closure of the neural tube occurs at around 3–4
weeks gestation and signals the beginning of the
28 Brain Arteriovenous Malformations
embryonic stage. In regard to the morphogenesis
of the cerebral circulation, the embryonic stage
can be further classified into the prechoroidal,
choroidal, and brachial phases. During the
prechoroidal stage, blood vessels develop by
way of vasculogenesis from the meninx primitiva
on the neural tube surface. The meninx primitiva
is the mesenchymal covering from which the
dura, arachnoid, and pia mater are derived. Within
vasculogenesis, the critical interaction occurs
between the principal signaling molecules, vascu-
lar endothelial-derived growth factors (VEGF),
and their receptors (VEGFR). This is a relatively
large group of molecules (VEGF A-E), which
have recurring roles in vasculogenesis and angio-
genesis both in physiological and pathological
states. The interaction between VEGF-A and
VEGFR-2 promotes angioblasts to form
angiocysts, which ultimately migrate to form the
early vascular network of the meninx [60].
The neural tissue is supplied by simple diffu-
sion from the primitive vascular plexus on its
surface during the prechoroidal phase. As the
cranial neural tube develops into the five vesicles
(telencephalon, diencephalon, mesencephalon,
metencephalon, and myelencephalon), the neural
tissue becomes too abundant to be supported by
simple diffusion from the peripheral vascular
plexus. Support is added when the choroid
develops as an invagination of the meninx from
the roof of the neural tube into the central canal as
the choroid is quite metabolically active and
brings with it its vascular supply, thereby initiat-
ing the choroidal phase [63–65].
The brachial phase sees the emergence of the
adult-type arterial pattern by a series of regres-
sions and enlargements of preexisting transverse
vessels. As the developing central nervous system
exceeds the capacity of simple diffusion, vascular
channels form within the parenchyma to keep up
with the metabolic demands of the rapidly devel-
oping primitive brain. Initially there is vasodila-
tion directly mediated by increases in nitric oxide
(NO). Following vasodilation, there are signifi-
cant increases in vascular permeability again
mediated by nitric oxide, however, this time via
upregulation of VEGF family molecules. VEGF
acts directly on endothelial cells to effect cell-cell
613
adhesion molecules resulting in leakage of plasma
proteins. These plasma proteins facilitate the
breakdown of extracellular matrix components
and at the same time provide the necessary tem-
porary scaffolding allowing migration of endothe-
lial cells [66]. Pseudopodia are formed with a lead
cell that penetrates between glial elements
[67]. These early vascular channels are orientated
at right angles to the pial surface with little
branching.
A similar pattern occurs from the ependymal
surface of the primitive ventricles. For both the
more dominant ventriculopetal and the more mod-
est ventriculofugal cerebral vascular develop-
ment, VEGF is again a key factor driving the
development with high VEGF levels found in
the deep subventricular zone acting as a powerful
chemoattractant to vascularization [60]. The
migrating and dividing endothelial cells initially
form solid cords of elongated endothelial cells
directed toward high concentrations of VEGF,
particularly in the subventricular zone. Following
the formation of solid cords, there is canalization
with lumen formation. This type of angiogenesis
is referred to as “sprouting” and is the dominant
form in the developing CNS [68]. The definitive
intraparenchymal vascular pattern is not arrived at
immediately. Immature vessels, which are little
more than continuous vascular channels at this
stage, without recognizable arterial, capillary, or
venous components, undergo further selection
with dominant vessels undergoing further
branching, while other channels regress. To this
end, an interaction between the endothelial cells
and their supportive structures, namely, smooth
muscle cells (SMCs) and extracellular matrix
(ECM), is critical. Both the ECM and the SMCs
provide a source of growth factors and supportive
ligands, which lend permanency to the developing
vascular structures. During both vasculogenesis
and angiogenesis within the developing CNS,
the VEGF family of signaling molecules plays
the dominant role and in particular is critical
for endothelial cell maturation and replication
[60, 63–65].
In addition to these critical signaling pathways,
several other molecules have been identified
which play a role not only in endothelial cell
614
signaling but also in directing the supporting
SMCs and interactions with the ECM. ANG-1
interacts with the TIE2 receptor to encourage
recruitment and interaction between the endothe-
lial cell and the supporting environment. Equally
important is the ANG-2 molecule, which has an
antagonistic effect on the TIE2 receptor neces-
sary for appropriate vascular remodeling to form
the final organized functional arterial tree
[60]. Further signaling molecules include
endoglin and the related Alk1 receptors which
play a role in endothelial cell to cell interactions
during capillary formation and act on supporting
structures, triggering mesenchymal cells to
differentiate into pericytes and SMCs [69, 70].
There has been an increasing interest in these
molecules as their role in pathological processes
becomes more apparent, particularly given their
role in genetic forms of bAVM (Osler-Weber-
Rendu (also called hereditary hemorrhagic telan-
giectasia)) [66, 71, 72].
Pathophysiology
On gross inspection, bAVMs are composed of one
or more, often enlarged, afferent arterial feeders
with one or more, frequently dilated, efferent
veins. The core pathophysiology in bAVMs is
that of arteriovenous shunting. The degree of
shunting has been categorized as either
“nidiform” or “fistulous,” though these two types
can coexist within a single lesion (Fig. 3)
[1, 73, 74]. (It is important that the reader not
confuse this distinction with dural (Fig. 4) or pial
(Fig. 5) arteriovenous fistula, which in the former
reflects an abnormal connection between the
artery and vein within the dura and in the latter
refers to an abnormal single connection within the
pia, without a subpial component). Rather, this is a
description of the physiologic shunting between
the arterial and venous compartments of a bAVM.
The nidiform type demonstrates a larger network
of abnormal vascular channels leading from the
arterial compartment to the venous compartment.
These channels comprise the “nidus” which gives
the bAVM its mass-like appearance and which
adds a conduit that separates the arterial feeders
D. Sahlein and N. Manning
from the venous outflow. The bAVMs that dem-
onstrate a more direct physiologic shunt, and
therefore possess less of a vascular network
between the arterial and venous compartments,
instead possess a shunt in which the artery and
vein are more directly connected [75, 76]. In the
end, the nidiform and fistulous designations exist
not as absolutes but along a continuum. The fistu-
lous physiology is more rare than the nidiform
type, representing only around 2 % of patients in
some series [6]. Interestingly, this morphological
type is typically reported on the surface of the
central nervous system [1].
Brain arteriovenous malformations are found
in the supratentorial compartment in the over-
whelming majority (>90 %) and demonstrate a
predilection for the MCA territory [6]. Deep struc-
tures are involved in approximately 15 % of cases
[43]. In approximately 50 %, there is deep venous
drainage [77].
There are many ways of categorizing and
understanding bAVMs. Some authors have
suggested that there is a degree of predictability
in both the pattern of arterial supply and venous
drainage based on the location of the bAVM
[1]. This is useful to keep in mind when
interpreting the imaging appearance of lesions,
particularly in addressing arterial supply and
venous drainage. Cerebral bAVMs may be topo-
graphically divided into those that reach the cor-
tex, those restricted to a deep location, and those
involving the choroid [1]. Lesions involving the
cortex are further classified into four categories:
1. Cortical lesions are on or reach the cortex and
supplied exclusively by cortical arteries with
only superficial venous drainage. It should be
noted that the basal veins of Rosenthal are not,
anatomically speaking, deep cerebral veins as
they lie on the brain surface and may drain
cortical bAVM (i.e., temporal lobe lesions).
They are, however, considered deep veins
from the neurosurgical point of view given
their relatively inaccessible location and,
when involved in the surgical treatment of
bAVMs, have higher associated morbidity.
Cortical lesions have also been referred to as
sulcal bAVMs.
28 Brain Arteriovenous Malformations
Fig. 3 A 58-year-old male with a left posterior temporal/
temporo-occipital brain AVM discovered incidentally fol-
lowing a motor vehicle accident. T2-WI transaxial MRI (a)
shows the classic appearance of the nidus within the cortex
and subcortical white matter of the left temporo-occipital
region (white arrowheads). The decision was made to treat
the patient with embolization and surgery. (b) Left internal
carotid angiography, lateral projection, in the early arterial
phase demonstrates two dominant arterial feeders, both
grossly enlarged: the posterior temporal (white arrows)
and temporo-occipital (black arrows) divisions of the left
MCA supplying the nidus (white arrowheads). The capil-
lary (c) and venous (d) phases depict the bAVM venous
outflow. Venous outflow is predominantly into the left
sigmoid sinus (white arrows) as well as retrograde into
the vein of Labbe (black arrows) and out via Trolard
(yellow arrowheads). There is lesser outflow within corti-
cal occipital veins and into the posterior aspect of the
superior sagittal sinus (black arrowhead). (d) A superficial
vein (blue arrowheads) courses along the inferior and
posterior occipital lobe superficially and then heads ante-
riorly along the paramedian occipital lobe, draining into
the basal vein of Rosenthal (yellow arrowheads), vein of
615
Galen (red arrowhead), and straight sinus (white arrows), a
unique “deep” venous drainage for an otherwise superficial
bAVM. Lateral angiogram (e) following a single pedicle
embolization shows both fistulous (white arrowhead) and
nidiform (black arrowhead) components of the bAVM.
Note the en passage supply from the parieto-occipital
branch which goes on to supply the normal brain (yellow
arrowheads). Superselective microcatheter angiography
prior to embolization demonstrates discrete outflow into
separate veins from each microcatheter injection, a
multicompartmental physiology. (f) The catheter tip is
depicted by the black arrowhead. Venous drainage from
the bAVM (white arrowheads) is directed inferiorly into
the sigmoid sinus (white arrow). (g) This pedicle opacifies
a separate compartment which, in distinction, drains pos-
teriorly along the occipital pole (white arrowheads). The
catheter tip is depicted by a black arrowhead. The shadow
of the glue cast can be seen (yellow arrowheads). (h) An
unsubtracted lateral x-ray shows the glue cast following
three pedicle embolizations (yellow arrowheads). Lateral
angiogram in the arterial phase (i) shows significant
devascularization of the nidus (white arrowheads) follow-
ing three pedicles of embolization
616
Fig. 4 A 53-year-old male presented to the ER for wors-
ening episodes of dizziness. CTA was performed (a) which
demonstrated increased vascularity abutting the right
petrous ridge (white arrowhead) with a few enlarged
enhancing curvilinear structures adjacent to the
pontomedullary junction on the right (black arrowheads).
This could easily be confused for a bAVM given its almost-
nidiform appearance. (b) Catheter angiography (right
external carotid anterior trunk in the arterial phase) shows
a dural arteriovenous fistula along the right petrous ridge
with dominant supply from the petrosquamosal (yellow
arrowheads) and basal tentorial (blue arrowheads)
branches of the right middle meningeal artery. Dominant
D. Sahlein and N. Manning
venous outflow is into the petrosal vein (black arrowhead)
into the basal vein of Rosenthal (white arrow) which drains
out via the vein of Galen (black arrow) and straight sinus
(red arrow). Additionally, there is infratentorial drainage
into a superficial cerebellar hemispheric vein (white arrow-
heads). Reformatted CTA can assist with surgical planning
(c). The vascular fullness along the petrous ridge is appar-
ent (white arrowheads). The petrosal vein courses toward
the viewer in this projection (black arrowheads) into the
basal vein of Rosenthal (at the top of the image, yellow
arrows). The internal auditory canal (red arrow) is
included as a helpful osseous landmark for surgery
28 Brain Arteriovenous Malformations
Fig. 5 A 6-year-old girl presented with subarachnoid
hemorrhage and intraventricular hemorrhage. Transaxial
time-of-flight MRA (a) demonstrates unusual ectasias or
aneurysms in the region of the ACAs, one oriented towards
the right and the other to the left (white arrows). Left
internal carotid AP (b) and lateral (c) angiograms in the
early arterial phase demonstrate a rapid early opacification
from a small frontal branch of the ACA, arising from the
right ACA (white arrow shows the junction of the artery
and vein). Capillary phase AP (d) and lateral (e) angio-
grams demonstrate the early opacification of a paramedian
cortical vein with multiple ectasias (black arrow), draining
into the anterior aspect of the superior sagittal sinus (white
2. Cortico-subcortical bAVMs, also referred to as
gyral bAVM, are supplied by cortical arteries
and drained by cortical veins. However, they
may also drain into the deep venous system if
617
arrowhead). The individual feeding artery was
catheterized and an arteriogram was performed (f, g).
Note the absence of a nidus or subpial component. The
artery drains directly into the ectatic vein (junction marked
with a white arrow). The microcatheter was advanced
across the fistula and into the ectactic venous outflow. (h)
depicts the microcatheter tip within the first venous ectasia
(black arrowhead). The fistula was then coiled. Follow-up
internal carotid angiogram in the lateral projection (i) dem-
onstrates no residual shunting. The coil mass is positioned
within the ectactic primary draining vein, extending prox-
imally into the arterial supply (white arrowhead)
the transcerebral venous pathways remain open
(Figs. 6 and 7). Unusual cortical pathways into
the medial temporal cortical venous drainage
can also lead to deep venous drainage.
618
Fig. 6 A cortico-subcortical bAVM in a 33-year-old
patient found to have a bAVM during a work-up for head-
aches. (a) A sagittal T1-WI precontrast MRI depicts the
cortical and subcortical bAVM (white arrow). Lateral angi-
ography from a right ICA injection in the mid arterial (b)
and late arterial (c) phases depicts a parietal cortical and
subcortical bAVM (white arrow) with supply from the
ACA parietal divisions (black arrowhead) and MCA pari-
etal division (white arrowhead). The venous outflow is
depicted in (c). There is superficial venous drainage into
the superior sagittal sinus which might be expected in a
parasagittal bAVM such as this one (white arrows depict
3. Corticoventricular bAVMs extend from
the cortex to the ventricular surface (Figs. 7
and 8). These lesions often demonstrate the
classic pyramidal morphology of bAVM with
a broad cortical surface and apex pointing to or
indeed reaching the ventricular margin or
D. Sahlein and N. Manning
the venous outflow into the superior sagittal sinus (skinny
black arrows)). More unusual is the transcerebral pathway
that remains open in this case and leads to deep venous
drainage despite the relatively superficial nidus (white
arrowheads). Finally, it drains into the vein of Galen (red
arrowhead) and straight sinus (yellow arrowhead). These
latter two veins barely opacify because of the relative
paucity of opacified blood coming through the
transcerebral venous conduit relative to the large amount
of unopacified blood coming through both cerebral hemi-
spheres (it is still the late capillary phase on the right- the
side of the injection.) into the deep venous system
subventricular zone. These lesions will be
supplied by both cortical and perforating arter-
ies and have both superficial and deep
drainage.
4. Corticocallosal bAVMs share many similari-
ties with corticoventricular lesions, having
28 Brain Arteriovenous Malformations
Fig. 7 An 11-year-old boy was sent for an MRI as part of
an evaluation for with headaches. MRI was performed
which revealed a right parietal corticoventricular brain
arteriovenous malformation with superficial and deep
venous drainage and measuring 5.1  2.0  2.0 cm in
maximum transverse, craniocaudal, and AP dimensions,
respectively. (a) Transaxial T2-WI images demonstrate the
classic flow voids extending from the cortex to the
ependymal surface (white arrowheads). The large flow
void within the ventricle represents a deep vein (black
arrowhead). Right internal carotid angiography in the
early arterial (b) and later arterial (c) phases shows the
nidus extending from the cortex to the ependymal surface
both deep and superficial draining veins. The
arterial supply to the corpus callosum is
derived from cortical vessels; however, it may
appear to represent perforators, particularly
anteriorly and at the splenium [1, 78].
Deep lesions can be found in any structure of
either the supra- or infratentorial compartments
(Fig. 9). They are exclusively supplied by perfo-
rating arteries and drained via the deep venous
system. These lesions obviously pose the greatest
challenge to open surgical treatment. Choroidal
bAVMs are of particular importance to
619
(white arrowheads). Note the classic perforator supply
from the lateral lenticulostriate arteries (white arrow).
The late arterial phase clearly depicts both superficial
(black arrow) and deep (black arrowhead) drainage. In
addition, there is an unusual transcerebral vein which
drains from the deeper aspect of the lesion into the super-
ficial venous system (yellow arrowhead). That
transcerebral vein is visible on the MRI T2-WI sequence
(magnified view (d–j) going from the level of the ventricle
sequentially to the level of the vertex), a testimony to the
excellent soft tissue contrast that MRI affords. It would be
difficult, however, to see this finding prospectively in the
absence of the catheter angiogram (yellow arrowheads)
prospectively identify as their extra-axial location
makes them both surgically and endovascularly
accessible. This extra-axial location can be chal-
lenging to identify on cross-sectional imaging in
extensive lesions or in those lesions that do not
respect anatomical boundaries. The key to appre-
ciating this topographical subtype of bAVM is in
identifying supply by choroidal and
subependymal arteries arising from the circle of
Willis including the basilar tip. Cortical arterial
supply is not a feature of these lesions. Drainage is
typically via ventricular veins but may also
involve transcerebral veins. While these features
620
Fig. 8 A sagittal postcontrast image to the right of midline
demonstrating a classic pyramidal or wedge-shaped
corticoventricular bAVM in a 33-year-old, discovered dur-
ing a work-up for headache. Note the excellent soft tissue
contrast on MRI and the precise anatomic localization
are largely reliable, it must be remembered that
bAVMs are dynamic lesions capable of evolution
and adaptation. This is particularly true for venous
drainage which may shift significantly if previ-
ously established drainage undergoes thrombosis
or becomes stenotic [1].
Histology and Histopathology
Microscopic examination of bAVMs demon-
strates abnormal vessels from both the arterial
and venous compartments, with still further ves-
sels that do not easily fit into either venous or
arterial categories [62]. While it is commonly
stated that the bAVM nidus lacks brain paren-
chyma, gliotic neural tissue is often seen at the
margin and even within the lesions [61, 79]. There
is frequently evidence of microhemorrhage that
does not have a clinical event correlation and thus
implies subclinical extravasation [79]. Likely
related to this phenomenon, evidence of calcifica-
tion can be seen within bAVM vessel walls and
has even been reported within the adjacent paren-
chyma [80]. This calcification is more easily iden-
tified histologically [80], although may
occasionally be identified on imaging [81]. Clus-
ters of abnormal, muscularized arteries become
D. Sahlein and N. Manning
increasingly frequent as one approaches the site
of shunt [62]. One of the hallmarks of these abnor-
mal vessels is the marked variability in the wall
thickness, particularly involving the media. There
is both thickening and focal thinning of the vas-
cular wall with medial fibrosis and duplication
and degradation of the elastic lamina [62]. The
media itself may be partially or completely absent
in sections or alternatively demonstrate marked
structural disarray with separation into an inner
and outer coat by an aberrant elastic lamina [61,
62, 79, 80]. Veins often have an “arterialized”
appearance with lack of elastin and hypertrophy
of the smooth muscle layer [69]. However, in
sections there may be near complete loss of nor-
mal venous mural architecture, being apparently
replaced by the fibrous tissue [62, 79].
The vast majority of bAVMs are sporadic;
however, bAVMs associated with genetic syn-
dromes have lead to insight regarding the roles
that various genes play in vascular development
and in the development of bAVMs [1, 61]. It
comes as little surprise that many genes with a
role in either vasculogenesis or angiogenesis dem-
onstrate altered levels of expression of functional
proteins in bAVMs.
Studies into hereditary hemorrhagic telangiec-
tasia (HHT) (Rendu-Osler-Weber syndrome) 1 and
2 have provided some insights into the disrupted
pathways that may be involved in bAVM forma-
tion. Hereditary hemorrhagic telangiectasia is an
autosomal dominant condition of vascular dyspla-
sia involving multiple organ systems. The clinical
symptoms in HHT are caused by direct arteriove-
nous connections without an intervening capillary
bed. This results in multiple mucocutaneous telan-
giectasias particularly about the oral cavity, finger-
tips, and within the nasal cavity. The latter leads to
recurrent epistaxis, which is a common presenting
symptom. In addition to these superficial telangi-
ectasias, there are visceral AVMs involving, for
example, the lungs, liver, and gastrointestinal tract
in around one third of patients. Cerebral AVMs are
seen in around 10 % of patients; however, they are
probably under-recognized [82]. A mutation of
endoglin, a component of the transforming growth
factor (TGF) receptor complex, is responsible for
HHT1 [61, 82]. The gene responsible for HHT2
28 Brain Arteriovenous Malformations
Fig. 9 A 16-year-old with a history of a ruptured diffuse
thalamic bAVM. (a) depicts the intracerebral and intraven-
tricular acute hematoma. Note the hydrocephalus. The
patient was sent for catheter angiography and possible
embolization. AP Towne projection from a left vertebral
injection (b) shows a diffuse thalamic and ventricular nidus
(white arrows) with supply from right posterior cerebral
artery (PCA) perforators and the posteromedial choroidal
artery. The patient was sent for embolization (c) to
has also been identified: ALK1, which belongs to
the TGF receptor superfamily [61, 82]. Transgenic
mice carrying homozygous defects in endoglin die
in utero with defective vascular remodeling. Het-
erozygous mice demonstrate a phenotype consis-
tent with HHT1 including the development of
bAVMs [70]. Endoglin appears critical for normal
capillary remodeling and extension, being highly
expressed during capillary extension prior to the
commencement of blood flow. It has been
suggested that failure of this process leaves the
newly formed capillaries vulnerable to opportunis-
tic micro-shunts at the time of the first blood flow,
which, over time, develop into bAVMs [83].
621
temporize the lesion for Gamma Knife therapy. The
microcatheter tip is depicted by the black arrow. A
pseudoaneurysm (white arrowhead) was successfully
targeted. Two sequential transaxial slices from a
noncontrast head CT (d and e) acquired 4 years later
show the onyx (embolic) cast, the very hyperdense curvi-
linear material in the right thalamus and posterior body of
the right lateral ventricle. The patient has been stable
neurologically
While these genes play a known role in hered-
itary bAVMs, they may very well play a role in
sporadic bAVMs as well. In fact, a study from
2005 found an association between a common
polymorphism of the ALK1 and sporadic
bAVMs [84], suggesting that these genes may
play a larger role in sporadic bAVMs than previ-
ously believed. Gene expression studies have
identified numerous genes with either increased
or decreased levels of expression in bAVMs com-
pared with normal cerebral vasculature [85]. How-
ever, the challenge remains to identify potential
therapeutic targets. Interestingly, increased VEGF
and VEGF receptor expression has been reported
622
Fig. 10 A 16-year-old male who presented with acute
onset of severe headache found to have a 2 cm left insular
bAVM. Lateral angiogram in the early arterial (a) and mid
arterial (b) phases from a left ICA injection shows a 2 cm
left insular bAVM supplied by the middle cerebral artery
(MCA) (white arrowheads). Venous drainage is into the
vein of Labbe (white arrows). Repeat lateral angiogram
following surgical resection of the lesion in the mid arterial
phase (c) and late arterial phase (d) shows no nidal
opacification or early venous opacification and the lesion
was thought to be cured. At the age of 25, the patient
sustained a sudden onset of severe headache and loss of
consciousness. Noncontrast head CT showed a large
following partial embolization of bAVMs
[86, 87]. Given VEGF’s role in angiogenesis,
this may be central to the dynamic and recurrent
nature of bAVMs and is an attractive target for
therapeutic intervention in combination with
endovascular treatments. Figure 10 depicts a
recurrent bAVM 9 years following what was
thought to be curative microsurgical resection.
D. Sahlein and N. Manning
parenchymal, subarachnoid, and intraventricular hemor-
rhage (e). Repeat angiogram (f) (left internal carotid injec-
tion in the lateral projection in the early arterial phase)
demonstrated a 3.4 cm left insular bAVM. Venous drainage
(g) was via Labbe (white arrows) as well and numerous
smaller frontal cortical veins (black arrowheads) into the
superior sagittal sinus. This was an unusual case of a
bAVM that was thought to be cured following microsurgi-
cal resection, but which recurred and ruptured 9 years later.
Cases such as this one stress the need for interval imaging
follow-up in younger patients even in cases that are
thought to be cured
Imaging
Whether a bAVM is identified incidentally or due
to clinical sequelae such as seizure or focal neu-
rological deficit resulting from hemorrhage or
venous congestion, imaging studies establish the
diagnosis. Imaging studies in their totality are
critical to define simple bAVM characteristics
28 Brain Arteriovenous Malformations
such as location and size as well as establishing
more complex characteristics such as drainage
pattern, feeding arteries, and the presence of com-
plicating factors such as associated aneurysms.
Therefore, an understanding of the concepts
already presented and applied to various imaging
studies allows the radiologist to provide critical
information in regard to the characteristics of a
bAVM, thus providing support for clinical deci-
sion making as well as therapeutic planning.
Every opportunity should be taken to provide
more than a simple diagnosis of bAVM and help
better understand individual bAVM physiology
and procedural risk.
Computed Tomography
Noncontrast computed tomography (CT) is not
well suited to assessing the detailed angioarchi-
tectural components critical in making treatment
decisions for bAVMs, however, as it is often the
first neuroimaging patients have for practical rea-
sons, and therefore every radiologist should be
aware of the CT imaging features. In particular,
patients who initially present with hemorrhage
secondary to bAVM will most likely undergo CT
as their first diagnostic procedure, and radiologists
should always be cognizant to the possibility of
bAVM as the cause of hemorrhage. When intra-
cerebral hemorrhage is encountered in young
patients (<45 years old), has a lobar distribution,
or is associated with parenchymal calcifications or
obvious hyperdense serpiginous structures,
bAVM must be considered as a possible etiology
[88]. In the case of unruptured bAVMs, slightly
hyperdense serpiginous structures may be seen
and generally represent draining veins which at
their largest are typically quite a bit larger than the
largest feeding arteries. Parenchymal calcification
can be seen in up to 20 % of bAVM either relating to
previous hemorrhage or venous thrombosis
(Fig. 11) [80]. Gliosis as a consequence of vascular
shunting or recurrent microhemorrhage may be
directly identifiable or alternatively may be demon-
strated by volume loss and asymmetry. The latter
may be obvious or, alternatively, may come to the
radiologist’s attention by appreciating asymmetry
in the CSF spaces such as the sulci or Sylvian
fissure. Ex-vacuo dilatation of ventricular structures
623
may give a clue to the presence of regional paren-
chymal volume loss. Hydrocephalus may also be
present due to mass effect from the nidus itself, or
more commonly a prominent draining vein.
CT angiography (CTA) will often establish the
diagnosis of bAVM by demonstrating the com-
plex mesh-like network of the nidus and, if flow is
sufficient, enlarged arterial feeders and abnormal
draining veins. However, as traditional CTA lacks
temporal resolution, the key diagnostic feature of
bAVMs, namely, arteriovenous shunting, cannot
be identified. In the absence of temporal resolu-
tion, other intracranial vascular abnormalities
must be necessarily considered in the differential
diagnosis, specifically developmental venous
anomalies (DVAs) (Fig. 12) and capillary telangi-
ectasia. DVAs and capillary telangiectasias have
features that may be mistaken for a nidus, and
DVAs demonstrate a prominent vein, which may
lead to further confusion. Modern CT angiogra-
phy (CTA) provides excellent spatial resolution
allowing characterization of nidal characteristics
including size and to a lesser extent intranidal
aneurysms, though the spatial resolution remains
a fraction of that of conventional angiography
(Figs. 1 and 13) [89].
With proper adherence to the technique, using
individually timed acquisitions to ensure that the
study is acquired in the early to mid arterial phase,
arteriovenous shunting may be inferred by the
presence of contrast within venous structures.
More recently, dynamic 3D-CTA techniques
have been developed which allow even greater
diagnostic accuracy, albeit at higher radiation
doses [90]. As CTA techniques improve, greater
angioarchitectural detail will be gleaned from
these noninvasive studies, allowing for the iden-
tification of associated aneurysms as well as the
interrogation of the venous drainage including the
number and location of draining veins as well as
the possibility of associated angiopathy such as
stenoses. In fact, CTA is currently able to identify
intranidal aneurysms with reliability approaching
that of digital subtraction angiography (DSA), the
gold standard (Fig. 13) [91].
In the setting of rapid clinical decline, CTA
may provide time-efficient information to assist
surgical planning in the emergent setting. The
624
Fig. 11 A 6-year-old girl had a noncontrast head CT for first
seizure (a). Note the calcified, slightly hyperdense pyramidal
shape structured within the anterior left frontal lobe cortex
and subcortical white matter (white arrowheads). Transaxial
CTA confirms the diagnosis (b), with a pyramidal complex
D. Sahlein and N. Manning
mesh-like network of the nidus as well as associated features
such as abnormal, enlarged draining veins (c). Note the
scalloped inner table abutting the venous varix owing to
chronicity (black arrowheads with white center). Transaxial
T2-WI MRI (d, e) demonstrates the classic flow voids
28 Brain Arteriovenous Malformations
routine use of CTA in the investigation of all
intracerebral hemorrhage is being adopted not
only to identify and characterize potential vascu-
lar causes but also to assist in prognostication and
possibly to identify patients that may benefit from
new treatments [92].
Magnetic Resonance Imaging
Magnetic resonance imaging (MRI) offers several
theoretical advantages over CT imaging of
bAVMs. In particular, MRI’s unparalleled soft
tissue contrast allows for the assessment of more
subtle clinically significant epiphenomena of
bAVMs such as clinically silent chronic hemor-
rhage, gliosis, and venous congestion. Further-
more, MRI’s anatomic detail is unmatched
allowing precise localization of the nidus and
characterization of its relationship to adjacent ana-
tomical structures. Using modern tractography
techniques as well as functional imaging, MRI
provides detailed information with regard to func-
tional anatomy that may assist in both open surgi-
cal and radiosurgery planning [93, 94].
The bAVM nidus appears, similarly to CT, as a
mesh-like network. However, due to the flow
within the vascular structures on typical spin
echo sequences, the nidal structures are depicted
as black flow voids (Figs. 7, 11, 14). Typically this
is best appreciated on T2-WI which has been
reported to have a diagnostic sensitivity as high
as 97 % in the case of unruptured bAVM
[91]. Intralesional or perilesional hemorrhage
will be demonstrated dependent upon the age of
the blood. Typically, chronic hemorrhage with
deposition of hemosiderin will appear
hypointense on all sequences with blooming on
susceptibility-weighted imaging or gradient echo
Fig. 11 (continued) associated with the nidus. Note, how-
ever, the high signal within the large venous varix (e)
attributable to turbulent and slower flow (white arrow-
heads). MRI signal characteristics must be interpreted in
the context of the flow physiology. Left ICA LAO angio-
grams in the early arterial and capillary blush phases (f and
g) clarify the detailed angioarchitecture of the bAVM.
There is direct supply from a trunk of the prefrontal divi-
sion of left MCA with additional en passant supply from a
625
imaging with low flip angles. Recent evidence
suggests that those patients with evidence of pre-
vious microhemorrhage are at greater risk for
subsequent clinically significant hemorrhage
[95]. The gliotic parenchymal tissue is generally
best visualized on FLAIR imaging where it
appears hyperintense. Venous hypertension may
lead to vasogenic edema and mass effect in the
involved territory, again best demonstrated by the
FLAIR sequence.
While convenient because of the lack of intra-
venous contrast administration, traditional time-
of-flight (TOF) magnetic resonance angiography
(MRA) has limited utility in bAVM imaging,
being inferior to both CTA and standard MRI in
direct comparisons [91]. Technical limitations in
anatomic coverage and artifacts related to slow
flow result in poor delineation of the nidus itself
and failure to depict associated aneurysms and
some venous drainage pathways. Modified MRA
techniques are under development that allow
superior vascular imaging and possibly even pro-
vide time resolution, critical to competing with
DSA. Postcontrast, dynamic MR subtraction
angiography has shown promise in early work;
however, it is time demanding [96].
More recently, intracranial 4D flow MRI has
been used not only to depict time-resolved MRA
but also to assess flow dynamics within vascular
structures [47]. This may prove of particular inter-
est to endovascular surgeons. Changes in lesion
hemodynamics following partial embolization
may in part be responsible for post-embolization
hemorrhage [6]. Initial results demonstrate
marked changes in flow velocities of
vascular structures both in the immediate vicinity
of the lesion and at a distance from the nidus as
second trunk. Drainage is entirely superficial, with the
deep aspect of the AVM and draining superiorly into the
superior sagittal sinus via cortical veins and the superficial
aspect draining via cortical veins into the superficial
Sylvian territory and out of via the vein of Labbe. There
is a 1.5 cm venous varix (small white arrow) associated
with the deep aspect of the AVM and a 2.5 cm venous varix
(small black arrow) associated with the superficial aspect
626
Fig. 12 A 6-year-old boy was worked up for unexplained
loss of consciousness. Noncontrast CT had revealed a
calcified lesion in the left temporal lobe, and CTA had
demonstrated unusual vascularity in the region. The patient
was referred for angiography for evaluation of bAVM
versus tumor. Left internal carotid artery (ICA) injection
in the LAO view (a) and magnified lateral oblique (b) in
the mid arterial phase demonstrating a normal study. LAO
D. Sahlein and N. Manning
view (c) and magnified lateral oblique (d) in the mid
venous and late venous phases (e) and (f), respectively.
Note the appearance of the classic “caput medusae” (white
arrows on c and d). Note the delayed washout as a result of
a stenosis (black arrow on f). The reported calcification is
likely the result of chronic venous congestion with possible
component of venous infarct. There is no current way to
treat this abnormality
28 Brain Arteriovenous Malformations
Fig. 13 An 11-year-old presented with severe refractory
headaches which began while he was playing soccer. (a)
Noncontrast head CT demonstrates a left temporo-occipital
acute hematoma. (b) CTA was obtained and the axial MIP
demonstrates two potential foci of hemorrhage. A larger,
medial focus (large white arrow) and a smaller, posterior
and slightly lateral focus (small white arrow). The patient
was sent for catheter cerebral angiography to further elu-
cidate. Left vertebral artery LAO oblique angiogram in the
early arterial (c) and late arterial/capillary phases (d). (c)
Note the same two features on the angiogram. The large
white arrow points to the larger outpouching, while the
smaller points to the smaller. In the late capillary/early
venous phase (d), the larger outpouching is clearly the
proximal draining vein (small white arrowheads) with a
small associated varix, whereas the smaller focus looks
well [47]. A greater understanding of the signifi-
cance of these changes may lead to safer and more
effective endovascular treatment.
Digital Subtraction Angiography
Digital subtraction angiography (DSA) remains the
gold standard for both the detection and character-
ization of bAVMs. At the heart of bAVM pathology
is the shunting of the blood between the arterial and
627
pseudoaneurysmal. Repeat angiography 2 weeks later
showed interval thrombosis of the small pseudoaneurysm
which is not uncommon for pseudoaneurysms in the sub-
acute phase following rupture. Also notice the mass effect
with splaying of arterial branches (small black arrow-
heads) and central oligemia, the angiographic appearance
of hematoma. Microcatheterization of pedicles 1 (e) and
2 (f) demonstrates a fairly diffuse border and a single
physiologic compartment, meaning that both pedicles
share the same draining vein. (Catheter tip is depicted by
the black arrowhead.) The point at which each pedicle
injected enters a shared vein is depicted as a white arrow-
head. The bAVM was embolized to cure with nBCA
embolic pushed into the venous outflow on the second
injection
venous vascular compartments. Currently, DSA is
the most reliable modality for evaluation of
bAVMs as a result of its fine temporal resolution.
Furthermore, the extremely high spatial resolution
has the potential to depict the exquisite vascular
detail required for the analysis of angioarchitectural
features as well as related hemodynamics. For
example, voxel size on the latest Siemens equip-
ment is 0.1  0.1 mm (mm) for DSA versus an
628
Fig. 14 A 47-year-old female with seizures. (a)
Transaxial fast spin echo T2-WI images near the vertex.
Note the classic “mesh-like” appearance of black flow
voids within the nidus (white arrow). MRI yields excellent
soft tissue contrast and depicts precise anatomic localiza-
tion. The bAVM nidus is centered within the cortex and
subcortical white matter of the right postcentral gyrus. AP
(b) and lateral (c) catheter angiography in the early arterial
phase from a right ICA injection demonstrate the arterio-
venous shunt associated with the paramedian right parietal
nidus (white arrows depict the nidus) supplied from the
precentral and superior parietal divisions of the ACA
(skinny black arrows), with pial-to-pial collaterals second-
arily supplying from the right MCA (black arrowheads).
Venous drainage is into a single superficial vein (white
average of 0.5  0.5  .5 mm for CTA and
1.0  1.0  0.9 mm for an average clinical MRA.
DSA allows a precise analysis of the origin and
course of arterial feeders as well as supply to
normal brain structures, if present. Aneurysms,
either found proximally about the circle of Willis,
arising from arterial feeders, or arising within the
nidus, are most easily identified on DSA, particu-
larly with superselective microcatheter angiogra-
phy [6, 20]. The lesion itself can be classified as
fistulous or nidiform type, or possessing both
D. Sahlein and N. Manning
arrowhead) that drains into the superior sagittal sinus.
The patient was sent for Gamma Knife treatment. (d)
Depicts the angiography for Gamma Knife targeting pur-
poses. The fiducial markers on the Leksell headframe are
centered within the images (“X” and “+”). At a 6-month
follow-up, there was a signal abnormality surrounding the
lesion seen well on the transaxial FLAIR sequence (e) near
the vertex (white arrow). The FLAIR signal extends into
the precentral gyrus, likely the result of perilesional radia-
tion exposure. The patient remained asymptomatic.
Follow-up catheter angiography at 3 years (f) demonstrates
complete bAVM obliteration. Note that the ACA has
diminished in caliber (skinny black arrow) owing to the
resolution of arteriovenous shunting
characteristics. The bAVM location and by asso-
ciation (as described earlier) arterial supply and
venous drainage are well demonstrated, although
the anatomic position on DSA is inferred via the
proximity to known arterial inflow and venous
outflow, as well as by bony landmarks rather
than directly visualized as it is with CT or MRI.
Finally, the all-important venous drainage can be
assessed. The temporal resolution of DSA allows
easy assessment of not only the bAVM draining
veins but also their relationship to venous
28 Brain Arteriovenous Malformations 629

structures subservient to brain parenchyma. This Table 1 A list of suggested characteristics of bAVMs
distinction is of significant practical importance to Angioarchitectural
the surgeon at resection. The use of superselective characteristics Notes
microcatheter injections allows a further level of AVM location Hemispheric lobe, basal ganglia,
angioarchitectural detail to be revealed. This is the thalamus, caudate, callosal,
intraventricular, cerebellar
most sensitive method for identifying nidal aneu- hemispheric, cerebellar vermis,
rysms [6, 20] as well as nidal flow characteristics subarachnoid, brain stem,
and compartmentalization, which are of critical perimesencephalic
importance in embolization procedures. AVM depth Categorized as hemispheric
supratentorial (with subcategories
cortical, cortical/subcortical,
Angioarchitecture periventricular, cortical/subcortical/
A separate discussion is warranted on angioarch- periventricular, subependymal),
itecture because of its paramount significance in callosal, deep nuclei, ventricular,
both risk estimation and treatment decisions. Fur- cerebellar, brain stem
AVM laterality Right-sided, left-sided, midline
thermore, while it is classically assessed via DSA,
Nidus size Measured in three planes or by
it forms the framework by which all radiologists cross section imaging or
should approach bAVM description irrespective angiography with fiducial
of what imaging modality has been performed. A markers. Additionally categorized
more thorough discussion of angioarchitectural as <3 cm, 3–6 cm, and >6 cm
features and their descriptions is available bAVM flow Nidus-predominant, large fistula-
physiology predominant, or mixed
[1]. We limit this discussion to an overview rele-
Arterial supply Organized by major feeding
vant to diagnostic imaging. Table 1 gives a sum- territory and categorized as
mary of relevant angioarchitectural features to cortical or perforator where
keep in mind during radiologic evaluation. appropriate
Associated The presence, number, and
aneurysms location (nidal, flow-related,
Arteries proximal (circle of Willis), venous)
Arterial feeders may be broadly classified by their Pial collaterals The presence of pial collaterals
relationship to the nidus as either direct feeders or and a description of whether the
en passage feeders (Fig. 3). The former supply the collaterals originated from the
nidus as an end artery, terminating at the nidus, same (i.e., MCA branch to
adjacent MCA branch) or different
while the later give branches to the nidus in pass- (i.e., ACA branch to MCA branch)
ing (en passage) and course distal to the bAVM to major arterial distributions
supply normal structures. This distinction has bAVM border Compact versus diffuse
obvious treatment implications. When en passage morphology
arterial feeders are identified, efforts should be Moyamoya-type The presence of moyamoya-type
changes arterial changes
made to identify downstream structures that may
Venous
be at risk if the artery is lost at surgery or during
characteristics
embolization. Either due to intrinsic changes to Number of Number of principal veins (1, 2,
the bAVM and its arterial feeders or due to partial draining veins 3, or more) directly draining the
or staged treatments, en passage feeders may nidus
evolve, with increasing contribution in bAVM Superficial versus Utilization of superficial, deep, or
supply. In such cases, changes in hemodynamics deep drainage both types of venous drainage
Venous angiopathy Outflow stenosis (defined
secondary to the bAVM’s “sump effect” will alter
as >70 % stenosis) or ectasia.
the location of preexisting watershed regions of Venous or dural sinus thrombosis
the brain. This may lead to vascular steal until a or occlusion
new perfusion equilibrium is achieved with adja- bAVM eloquence The presence and criteria of
cent vascular territories. These dynamic vascular eloquence as defined by Spetzler
and Martin2
changes, which effect normal brain regions, may
630
Fig. 15 AP (Towne projection) (a) and lateral (b) angi-
ography from a left vertebral injection demonstrating a
right occipital bAVM in an 18-year-old found incidentally
after a concussion. There is dominant supply from the
explain the development of seizure activity in
previously asymptomatic lesions.
Dynamic changes in arterial supply may result
when supply from an arterial feeder is reduced,
either from arterial stenosis due to “high-flow
arteriopathy” or as the result of partial emboliza-
tion and atypical and somewhat bizarre arterial
channels of supply may develop. These channels
may cause some diagnostic confusion if the
dynamic and evolving nature of these lesions is
not understood. Arterial supply may develop from
adjacent vascular territories or from vessels not
predicted to be involved (i.e., cortical bAVM with
deep perforator supply). The nidus may even
recruit dural supply in some cases (Fig. 15). This
process reflects non-sprouting angiogenesis prob-
ably in the setting of relative nidal ischemia and
the liberation of pro-angiogenic signaling mole-
cules such as VEGF. If this is suspected on any
imaging modality, it should be conveyed, as this
may increase the risk of surgery or mandate an
atypical endovascular approach.
Arterial Aneurysms
Arterial aneurysms are seen with increased fre-
quency in bAVM patients. The precise incidence
is unclear, being variously reported from 7 % to
D. Sahlein and N. Manning
calcarine and parieto-occipital divisions of the PCA as
well as en passant supply from the posterior temporal
division. There is dural supply from the posterior menin-
geal artery (white arrowheads)
58 %, and is likely dependent upon high-quality
superselective angiography [6, 97]. Redekop and
colleagues proposed a classification system for
bAVM-related aneurysms that remains useful
[98]. In the classification by Redekop et al.,
aneurysms are divided into three broad catego-
ries: intranidal, flow-related, and unrelated (dys-
plastic). Intranidal aneurysms are found within
the boundaries of the nidus and demonstrate rel-
ative early filling during angiography, prior to
opacification of the venous outflow. This should
distinguish intranidal aneurysms from venous
ectasia and aneurysms. Flow-related aneurysms
are those found along the course of arteries that
contribute supply to the bAVM. The authors fur-
ther classified flow-related aneurysms into prox-
imal if the aneurysms were found on or about the
circle of Willis or vertebrobasilar trunk or distal
if the aneurysm occurred beyond these locations.
Unrelated aneurysms were those that occurred on
arteries with no supply to the bAVM
[98]. Unrelated aneurysms are seen in fewer
patients with bAVM (0.8 %) than the general
public, with a generally reported rate of 2–3 %
[98]. This is likely to reflect the younger age of
bAVM patients with respect to the typical aneu-
rysm patient. Both flow-related and intranidal
28 Brain Arteriovenous Malformations
aneurysms are seen at significantly higher rates
suggesting a causative effect of the bAVM or
possibly a shared etiology [24, 99]. It is likely
that a combination of hemodynamic factors
including flow velocity and the hyperdynamic
circulatory environment associated with
bAVMs are at least partly involved. Furthermore,
regression and even resolution of associated
aneurysms have been reported with definitive
treatment of the bAVMs [100]. The risk of hem-
orrhage due to the presence of an associated
aneurysm is also variably reported. There does
appear to be an association with the presence of
aneurysms and initial hemorrhagic presentation;
however, the effect an aneurysm has on the
rate of subsequent hemorrhage remains unclear
[16, 24, 33, 42, 98, 101].
The Nidus
Perhaps the most easily discernible nidal charac-
teristic on noninvasive imaging is the distinction
between compact and diffuse nidi. Diffuse lesions
have intervening brain parenchyma, often demon-
strate multiple, more normal-appearing arterial
feeders, and numerous morphologically
unremarkable draining veins [1]. Diffuse lesions
pose a significant management conundrum as they
are often extensive, with intervening functional
parenchyma.
Nidal size is of importance when considering
treatment options and associated risk. Particularly,
the maximum nidal dimension should be mea-
sured and categorized as 0–3 cm, >3–6 cm,
or >6 cm for the sake of Spetzler-Martin classifi-
cation [5] (see further discussion below). Nidal
size should not be confused with the compact and
diffuse classification as undoubtedly compact
bAVM may be larger than 3 cm.
Using superselective angiography, the nidus
may be further classified as single compartment
(Fig. 13) or multi-compartment bAVM (Fig. 3).
Those lesions that demonstrate a single draining
vein or veins which opacify after every arterial
feeder are injected are considered single compart-
ment. That is to say all arterial feeders have a
common pathway to a shared venous outflow,
whether single or multiple. A multi-compartment
bAVM demonstrates distinct venous drainage,
631
which is unique to one or a few arterial feeders
with distinct draining veins filling from separate
pedicles. This classification is primarily of signif-
icance to embolization.
Of further significance to embolization is the
classification of the abnormal arteriovenous com-
munication as fistulous or nidiform type. In the
former, there is a more direct communication
between the artery and vein. As stated previously,
in practice both fistulous and nidiform types occur
in the same lesion. However, points of shunting
must be identified and monitored to prevent pre-
mature embolization of venous outflow. Both
compartmental characteristics and shunt charac-
teristics can only accurately be assessed on cath-
eter angiography.
Veins
Analysis of venous drainage in the setting of
bAVM should already be part of routine imaging
interpretation given its importance in surgical
morbidity and in endovascular embolization.
The presence of deep venous drainage should be
evaluated on all bAVMs for calculation of the
Spetzler-Martin grade [5]. In addition, the pres-
ence of high-flow venopathy including venous
stenosis and ectasia (Fig. 2) should be
documented, as well as the presence of venous
aneurysm. The presence of venous thrombosis
should be included in any analysis of bAVMs as
well (Fig. 16).
Management
A discussion of the management of bAVMs
should be divided into those patients presenting
with or having a history of clinically significant
hemorrhage and those patients presenting inci-
dentally or with nonhemorrhagic symptoms.
Hemorrhage rates are significantly higher in
patients with a history of hemorrhage estimated
at 4.5 % per year in meta-analyses [42]. The risk is
greatest in the first year following hemorrhage
with rates estimated between 6 % and 32.9 %
[13, 42]. Therefore, patients presenting with
bAVM hemorrhage are generally offered some
form of eradicative treatment in the hope that
632
Fig. 16 A 70-year-old female was found down, and
noncontrast head CT revealed a right brachium pontis and
vermis/medial right cerebellar hemispheric intracerebral
hemorrhage (a). CTA was performed and read as normal.
The patient was sent for catheter angiography because of the
unusual location of the hemorrhage. Lateral mid arterial (b)
and late arterial (c) angiograms from a left vertebral artery
injection show a diffuse superior vermian bAVM (white
arrowheads), likely being compressed by the hematoma.
Supply is from vermian branches of the superior cerebellar
artery (black arrowhead). The lesion drains into a very short
segment of the precentral cerebellar vein and into an
excluded pouch within the vein of Galen (red arrowhead)
which cannot drain into the straight sinus and instead
refluxes into the right internal cerebral vein (yellow arrow-
head) and into the right basal vein of Rosenthal (light blue
arrowheads) which drains out via an infratemporal vein
(white arrow). In a slightly later phase (capillary phase, c),
note the opening of falcine veins for internal cerebral venous
D. Sahlein and N. Manning
outflow (black arrow). AP (d) and lateral projection (e) right
internal carotid injection in the venous phase demonstrate no
opacification of the deep venous system on the right, and
extensive medullary venous congestion, with enlarged
transcerebral veins diverting the deep venous territory out-
flow superficially which can be appreciated as increased
vascularity within the brain parenchyma. Note the lack of
the internal cerebral veins, vein of Galen, or straight sinus on
the lateral projection (white arrows depict location where
the deep venous system should opacify, but does not). AP (f)
and lateral projection left internal carotid injection in the
venous phase demonstrate normal opacification of the left
deep venous system (white arrow) depicted en face. (g) Left
internal carotid injection in the venous phase demonstrates
normal opacification of the left internal cerebral vein (white
arrows), vein of Galen (black arrow), and straight sinus (red
arrow) which proves that the vein of Galen and straight
sinus remain patent and that the primary bAVM drainage
into Galen occurs into an excluded pouch
28 Brain Arteriovenous Malformations
this risk can be eliminated or palliative treatment
in the hope that the risk can be reduced. The
question of what management has to offer
bAVM patients without a history of hemorrhage
is more complex.
The recently published ARUBA explored the
issue of management of unruptured bAVM, com-
paring active treatment with conservative medical
management [3]. The trial was stopped early when
a data and safety monitoring board recommended
halting randomization because of the superiority
of the medical management group. The trial itself
has generated considerable controversy as a result
of poor recruitment, a skewed patient population,
heterogeneous treatment strategies, and relative
short-term follow-up [102–105]. However, it
does generate a need to apply more stringent risk
determination tools, like those discussed previ-
ously, to offer patients the optimum care.
Treatment consists of one or a combination of
three approaches: microsurgical resection,
endovascular embolization, and/or radiotherapy.
Microsurgical Resection
The goal of open surgery is complete excision of
the bAVM and, with that in mind, achieves the
most robust results. The Spetzler-Martin grading
scale was developed to predict surgical difficulty
and complications [5]. The scale ranges from 0 to
5 based on three factors:
1. Assessment of the nidus size (graded as
0 for <3 cm, 1 for 3–6 cm, and 2 for >6 cm),
oftentimes most easily measured on cross-
sectional images.
2. Determination of the pattern of venous drain-
age, either superficial or deep (0 or 1, respec-
tively). (Of note the term “deep venous
drainage” is a surgical one in this context rather
than an anatomical one. As stated earlier, while
the basal veins of Rosenthal are not considered
deep venous structures anatomically, they are
considered deep venous structures at opera-
tion. So too are other venous structures which
are challenging to access at craniotomy such as
the interhemispheric cortical veins.)
3. Involvement of eloquent areas or regions if
injured would cause immediate neurologic
633
deficit (scored as 0 for no eloquence and 1 for
eloquence). Eloquent regions typically
include the sensorimotor cortex, occipital (pri-
mary visual) cortex, speech areas, deep nuclei
of the cerebellum and brain stem, thalamus and
hypothalamus, internal capsule, and cerebral
and cerebellar peduncles. This scale offers a
fairly simple and effective risk stratification
system for patients undergoing surgical resec-
tion, and its effectiveness was prospectively
validated [4]. However, it does presume nor-
mal topographical orientation of brain func-
tions which may not be the case in bAVM
patients. Where possible, fMRI will provide
additional insight. The Spetzler-Martin scale
has been accurate in stratifying risk, with
Grade Is and IIs experiencing less than 5 %
major morbidity or mortality, Grade IIIs
experiencing approximately 5 %, and Grade
IVs and Vs experiencing 7–12 % [5]. The
scale is generalizable between surgeons in
large part because the procedural end point
for surgery is more or less the same – complete
resection of the entire bAVM. The same cannot
be said about the procedural end point for
embolization, a critical factor that will be
discussed below.
Embolization
Unlike microsurgical resection, the role of emboli-
zation is quite variable from institution to institution
or even practitioner to practitioner [6]. Embolization
might be carried out for the following reasons:
(1) endovascular occlusion, (2) devascularization
of as much of the bAVM as possible prior to or
following resection or radiosurgery, and
(3) targeting of a focal angioarchitectural weakness,
either alone or in combination with surgery or
Gamma Knife. Each of these strategies carries its
own unique risk profile and should be assessed
separately with respect to whether the procedural
end point has been met [6]. In the absence of
documented procedural intent, procedural efficacy
cannot be calculated, nor can results be generalized
across medical centers or between practitioners.
This lack of procedural standardization and docu-
mentation makes it impossible to develop a reliable
pre-embolization risk stratification scale [6].
634
The literature on bAVM embolization is limited
by a lack of class 1 evidence and is characterized by
small series, often without a control group or quan-
titative outcome measure. Nevertheless, proposed
benefits of embolization include better long-term
clinical outcomes with embolization and surgery
than surgery alone [106]; the ability to surgically
resect otherwise unresectable large, complex
bAVMs [107]; a reduction in morbidity following
surgery in embolized bAVMs in eloquent brain
areas [108]; equal operating time and blood loss
in larger, more complex embolized bAVMs as
compared to non-embolized smaller AVMs [109];
and diminished surgical difficulty [110].
Traditionally, the principal role of embolization
in most practices is as adjunct treatment prior to
surgical resection or radiotherapy (Fig. 3). A bal-
ance must be struck between the reduction of arte-
rial inflow and the preservation of balanced
outflow. If the arterial inflow is reduced signifi-
cantly but not completely obliterated, venous out-
flow may be reduced to the point of spontaneous
thrombosis. Delayed hemorrhage can result [111].
It should be appreciated from the previous
discussion of the hemodynamically balanced
nature of these lesions that if any arterial inflow
remains in the setting of occluded venous outflow,
hemorrhage becomes increasingly likely. How-
ever, in the setting of attempting cure via emboli-
zation alone, ultimately the venous outflow must
be embolized to prevent recurrence. This tech-
nique is being increasingly used for curative
embolization alone using either transarterial
[112] or transvenous approaches [113].
The liquid embolic agents, N-butyl cyanoacry-
late (NBCA) and onyx, a copolymer of ethylene
vinyl, are the primary agents used for the
endovascular treatment of cerebral AVMs. The
two agents have very different radiographic
appearances as a result of the contrast agents
used in each (Ethiodol versus tantalum).
The risk associated with embolization varies
widely in the literature. The morbidity associated
with embolization in the large nBCA series ranges
from 0.8 % to 9.0 % per patient [6, 114] and
mortality from 0.8 % to 3.7 % [6, 115] and onyx
series morbidity from 4.6 % to 15.5 % [116, 117]
and mortality from 2.1 % to 4.3 % [118, 119].
D. Sahlein and N. Manning
Radiotherapy
The first Gamma Knife surgery (GKS) for bAVMs
was performed in 1970. Multiple research studies
subsequently demonstrated that GKS is safe and
effective in small to medium-sized bAVMs, even
in deep and eloquent territories [120], with oblit-
eration rates of approximately 60–90 %
[121–124], at least in part dependent on bAVM
volume [125]. The major limitations of GKS are
the average latency to bAVM cure (approximately
2–3 years) and potential harm to perilesional brain
parenchyma. During the intervening time between
treatment and cure, the patient remains at risk for
hemorrhage or may even be at slightly greater risk
[126, 127].
While bAVM obliteration rates following GKS
are directly proportional to radiation dose, unfor-
tunately morbidity associated with GKS is also
correlated with radiation dose (as well as the total
volume treated- larger volume bAVMs are associ-
ated with greater morbidity) [128]. In order, there-
fore, to minimize the risk of morbidity, doses at
the periphery of the lesion, abutting normal brain
parenchyma, should be the lowest within the
targeted region, and minimum doses of approxi-
mately 17 Gy at the margin have been demon-
strated to lead to obliteration rates of
approximately 80 % and are an independent pre-
dictor of procedural efficacy (Fig. 14) [121].
The direct correlation between dose delivered
and both obliteration rates and morbidity and the
direct relationship between total volume treated
and morbidity have particular impact on the treat-
ment of large bAVMs using GKS, where total
lesion dose required for obliteration can cause
unacceptable levels of morbidity. There are differ-
ent strategies to safely increase radiation delivery
to the bAVM while limiting perilesional injury,
mostly by fractionating radiation dose by time or
by volume [129]. Multiple fractioned radiation
delivery protocols have been utilized, with
mixed results [130–135]. A more recent trial dem-
onstrated that bAVMs larger than 15 cm3 in vol-
ume can be treated successfully using fractionated
dosing with a staged-volume protocol [136]. Typ-
ically GKS is used alone; however, it has been
used with prior embolization to reduce the volume
of large bAVMs to that suitable for GKS with
28 Brain Arteriovenous Malformations
mixed results [135, 137]. Diminished efficacy of
GKS following embolization is likely related to
obscuration or incomplete opacification of the
nidus following endovascular treatment and there-
fore incomplete nidal targeting [137].
Modern equipment allows excellent dose shap-
ing, achieving high nidal doses with minimal dose
to surrounding tissues. Radiotherapy causes dam-
age to the endothelium and smooth muscle cells of
the bAVM. This induces an inflammatory reaction
and scarring with vascular narrowing and even-
tual obliteration of the nidus. GKS leads to signif-
icant morbidity in approximately 4.8 % [126].
Conclusion
bAVMs are complex and relatively rare vascular
lesions involving a shunt between the artery and
vein within the subpial space. Despite decades of
study, there is scant true natural history data even
among highly cited “natural history” papers, the
source of tremendous controversy about the true
natural history of the disease and secondarily the
most appropriate management. Furthermore, var-
iability from lesion to lesion calls into question the
accuracy of applying “average” data to individual
patients. Risk estimates using individual lesion
characteristics based on parameters with a clear
potential physiologic impact should be utilized.
Current models use anatomic features as physio-
logic surrogates, though advances in MR technol-
ogy might soon avail the neurovascular
community of direct noninvasive physiologic
interrogation of bAVMs. With current technology,
bAVMs are best imaged using digital subtraction
catheter angiography, which has the best temporal
and spatial resolution, though modern dynamic
CT angiography has good temporal resolution,
albeit with a higher radiation dose to the patient.
Management of bAVMs typically comes down
to microsurgical resection, endovascular emboli-
zation, Gamma Knife surgery, or some combina-
tion thereof. The efficacy and risks of each of
these procedures have been well studied, though
new treatment strategies are constantly being
developed, particularly using the latter two
modalities, necessitating ongoing research and
635
publication on the part of treating clinicians.
Given the uncertain natural history of individual
lesions, the significant risks and limitations of the
available management, and the complexity of the
imaging features and their interplay with risk-
benefit assessment, bAVM diagnosis and manage-
ment is best suited to collaborative multidis-
ciplinary teams. This allows management
combinations to be individually tailored to suit
both the characteristics of the lesion and the patient.
References
1. Lasjaunias PL, Berenstein A, Brugge KGT (2001)
Surgical neuroangiography, 2nd edn. Springer,
Berlin/New York
2. Ruiz-Sandoval JL, Cantu C, Barinagarrementeria F
(1999) Intracerebral hemorrhage in young people:
analysis of risk factors, location, causes, and progno-
sis. Stroke 30(3):537–541
3. Mohr JP et al (2014) Medical management with or
without interventional therapy for unruptured brain
arteriovenous malformations (ARUBA): a
multicentre, non-blinded, randomised trial. Lancet
383(9917):614–621
4. Hamilton MG, Spetzler RF (1994) The prospective
application of a grading system for arteriovenous
malformations. Neurosurgery 34(1):2–6,
discussion 6–7
5. Spetzler RF, Martin NA (1986) A proposed grading
system for arteriovenous malformations. J Neurosurg
65(4):476–483
6. Sahlein DH et al (2012) Nidal embolization of brain
arteriovenous malformations: rates of cure, partial
embolization, and clinical outcome. J Neurosurg
117(1):65–77
7. Al-Shahi R, Warlow C (2001) A systematic review of
the frequency and prognosis of arteriovenous
malformations of the brain in adults. Brain
124(Pt 10):1900–1926
8. Garfield E (1972) Citation analysis as a tool in journal
evaluation. Science 178(4060):471–479
9. Fults D, Kelly DL Jr (1984) Natural history of arte-
riovenous malformations of the brain: a clinical study.
Neurosurgery 15(5):658–662
10. Crawford PM et al (1986) Arteriovenous malforma-
tions of the brain: natural history in unoperated patients.
J Neurol Neurosurg Psychiatry 49(1):1–10
11. Brown RD Jr et al (1988) The natural history of
unruptured intracranial arteriovenous malformations.
J Neurosurg 68(3):352–357
12. Ondra SL et al (1990) The natural history of symp-
tomatic arteriovenous malformations of the brain: a
24-year follow-up assessment. J Neurosurg
73(3):387–391
636
13. Mast H et al (1997) Risk of spontaneous haemorrhage
after diagnosis of cerebral arteriovenous malforma-
tion. Lancet 350(9084):1065–1068
14. Cockroft KM (2007) Unruptured brain arteriovenous
malformations should be treated conservatively:
no. Stroke 38(12):3310–3311
15. Stapf C, Mohr JP (2007) Unruptured brain arteriove-
nous malformations should be treated conservatively:
yes. Stroke 38(12):3308–3309
16. Kim H et al (2014) Untreated brain arteriovenous
malformation: patient-level meta-analysis of hemor-
rhage predictors. Neurology 83(7):590–597
17. Choi JH et al (2006) Clinical outcome after first and
recurrent hemorrhage in patients with untreated brain
arteriovenous malformation. Stroke 37(5):1243–1247
18. Duong DH et al (1998) Feeding artery pressure and
venous drainage pattern are primary determinants of
hemorrhage from cerebral arteriovenous
malformations. Stroke 29(6):1167–1176
19. Nataf F et al (1997) Angioarchitecture associated with
haemorrhage in cerebral arteriovenous
malformations: a prognostic statistical model. Neuro-
radiology 39(1):52–58
20. Turjman F et al (1995) Correlation of the angioarch-
itectural features of cerebral arteriovenous
malformations with clinical presentation of hemor-
rhage. Neurosurgery 37(5):856–860, discussion
860–2
21. Kader A et al (1994) The influence of hemodynamic
and anatomic factors on hemorrhage from cerebral
arteriovenous malformations. Neurosurgery
34(5):801–807, discussion 807–8
22. Miyasaka Y et al (1992) An analysis of the
venous drainage system as a factor in hemorrhage
from arteriovenous malformations. J Neurosurg
76(2):239–243
23. Mansmann U et al (2000) Factors associated with
intracranial hemorrhage in cases of cerebral arterio-
venous malformation. Neurosurgery 46(2):272–279,
discussion 279–81
24. Meisel HJ et al (2000) Cerebral arteriovenous
malformations and associated aneurysms: analysis
of 305 cases from a series of 662 patients. Neurosur-
gery 46(4):793–800, discussion 800–2
25. Halim AX et al (2004) Longitudinal risk of intracra-
nial hemorrhage in patients with arteriovenous mal-
formation of the brain within a defined population.
Stroke 35(7):1697–1702
26. Stapf C et al (2000) Incident hemorrhage risk of brain
arteriovenous malformations located in the arterial
borderzones. Stroke 31(10):2365–2368
27. Brown RD Jr, Wiebers DO, Forbes GS (1990)
Unruptured intracranial aneurysms and arteriovenous
malformations: frequency of intracranial hemorrhage
and relationship of lesions. J Neurosurg
73(6):859–863
28. Khaw AV et al (2004) Association of infratentorial
brain arteriovenous malformations with hemorrhage
at initial presentation. Stroke 35(3):660–663
D. Sahlein and N. Manning
29. Marks MP et al (1990) Hemorrhage in intracerebral
arteriovenous malformations: angiographic determi-
nants. Radiology 176(3):807–813
30. Pollock BE et al (1996) Factors that predict the bleed-
ing risk of cerebral arteriovenous malformations.
Stroke 27(1):1–6
31. Todaka T et al (2003) Analysis of mean transit time of
contrast medium in ruptured and unruptured arterio-
venous malformations: a digital subtraction angio-
graphic study. Stroke 34(10):2410–2414
32. Yamada S et al (2007) Risk factors for subsequent
hemorrhage in patients with cerebral arteriovenous
malformations. J Neurosurg 107(5):965–972
33. da Costa L et al (2009) The natural history and pre-
dictive features of hemorrhage from brain arteriove-
nous malformations. Stroke 40(1):100–105
34. Alexander MD et al (2015) Association between
venous angioarchitectural features of sporadic brain
arteriovenous malformations and intracranial hemor-
rhage. AJNR Am J Neuroradiol 36:949–952
35. Stefani MA et al (2002) Angioarchitectural factors
present in brain arteriovenous malformations associ-
ated with hemorrhagic presentation. Stroke
33(4):920–924
36. Hernesniemi JA et al (2008) Natural history of brain
arteriovenous malformations: a long-term follow-up
study of risk of hemorrhage in 238 patients. Neuro-
surgery 63(5):823–829, discussion 829–31
37. Langer DJ et al (1998) Hypertension, small size, and
deep venous drainage are associated with risk of
hemorrhagic presentation of cerebral arteriovenous
malformations. Neurosurgery 42(3):481–486, discus-
sion 487–9
38. Shi X, Wang Z, Yu C (1993) An analysis of the
correlation between angiographic and clinical find-
ings in cerebral arteriovenous malformations. Chin
Med Sci J 8(1):35–37
39. Albert P et al (1990) A study on the venous drainage
of 150 cerebral arteriovenous malformations as
related to haemorrhagic risks and size of the lesion.
Acta Neurochir (Wien) 103(1–2):30–34
40. Nataf F et al (1998) Therapeutic strategy for cerebral
arteriovenous malformations. Proposal for classifica-
tion of individual hemorrhagic risk. Neurochirurgie
44(2):83–93
41. Abecassis IJ et al (2014) Natural history of brain
arteriovenous malformations: a systematic review.
Neurosurg Focus 37(3), E7
42. Gross BA, Du R (2013) Natural history of cerebral
arteriovenous malformations: a meta-analysis.
J Neurosurg 118(2):437–443
43. Sahlein DH et al (2014) Features predictive of
brain arteriovenous malformation hemorrhage:
extrapolation to a physiologic model. Stroke
45(7):1964–1970
44. Stapf C et al (2002) Incidence of adult brain arterio-
venous malformation hemorrhage in a prospective
population-based stroke survey. Cerebrovasc Dis
13(1):43–46
28 Brain Arteriovenous Malformations
45. Zammar SG et al (2014) Precision medicine in brain
arteriovenous malformation management: arteries
steal the show but veins may hold the crystal ball.
Neurosurgery 75(6):N13–N14
46. Wolf RL et al (2008) Arteriovenous shunt visualiza-
tion in arteriovenous malformations with arterial spin-
labeling MR imaging. AJNR Am J Neuroradiol
29(4):681–687
47. Ansari SA et al (2013) Intracranial 4D flow MRI:
toward individualized assessment of arteriovenous
malformation hemodynamics and treatment-induced
changes. AJNR Am J Neuroradiol 34(10):1922–1928
48. Brown RD Jr et al (1996) Incidence and prevalence of
intracranial vascular malformations in Olmsted
County, Minnesota, 1965 to 1992. Neurology
46(4):949–952
49. ApSimon HT et al (2002) A population-based study
of brain arteriovenous malformation: long-term treat-
ment outcomes. Stroke 33(12):2794–2800
50. Al-Shahi R et al (2003) Prospective, population-based
detection of intracranial vascular malformations in
adults: the Scottish Intracranial Vascular Malformation
Study (SIVMS). Stroke 34(5):1163–1169
51. Stapf C et al (2003) The New York Islands AVM
Study: design, study progress, and initial results.
Stroke 34(5):e29–e33
52. Hillman J (2001) Population-based analysis of arte-
riovenous malformation treatment. J Neurosurg
95(4):633–637
53. Weber F, Knopf H (2006) Incidental findings in mag-
netic resonance imaging of the brains of healthy
young men. J Neurol Sci 240(1–2):81–84
54. Berman MF et al (2000) The epidemiology of
brain arteriovenous malformations. Neurosurgery
47(2):389–396, discussion 397
55. Katzman GL, Dagher AP, Patronas NJ (1999) Inci-
dental findings on brain magnetic resonance imaging
from 1000 asymptomatic volunteers. JAMA
282(1):36–39
56. Tsushima Y, Taketomi-Takahashi A, Endo K (2005)
Prevalence of abnormal findings on brain magnetic
resonance (MR) examinations in adult participants of
brain docking. BMC Neurol 5:18
57. Yue NC et al (1997) Clinically serious abnormalities
found incidentally at MR imaging of the brain: data
from the Cardiovascular Health Study. Radiology
202(1):41–46
58. Jessurun GA et al (1993) Cerebral arteriovenous
malformations in The Netherlands Antilles. High
prevalence of hereditary hemorrhagic telangiectasia-
related single and multiple cerebral arteriovenous
malformations. Clin Neurol Neurosurg
95(3):193–198
59. Gabriel RA et al (2010) Ten-year detection rate of
brain arteriovenous malformations in a large, multi-
ethnic, defined population. Stroke 41(1):21–26
60. Moftakhar P et al (2009) Cerebral arteriovenous
malformations. Part 1: cellular and molecular biology.
Neurosurg Focus 26(5):E10
637
61. Gault J et al (2004) Pathobiology of human cerebro-
vascular malformations: basic mechanisms and clini-
cal relevance. Neurosurgery 55(1):1–16, discussion
16–7
62. Meng JS, Okeda R (2001) Histopathological structure
of the pial arteriovenous malformation in adults:
observation by reconstruction of serial sections of
four surgical specimens. Acta Neuropathol
102(1):63–68
63. Raybaud C (2010) Normal and abnormal embryology
and development of the intracranial vascular system.
Neurosurg Clin N Am 21(3):399–426
64. Wittko-Schneider IM, Schnieder FT, Plate KH (2014)
Cerebral Angiogenesis During Development: Who Is
Conducting the Orchestra. In: Milner R (ed) Cerebral
angiogenesis: methods and protocols. Methods in
molecular biology. Springer, New York, PP 3–15
65. Byrne JV (2012) Tutorials in endovascular neurosur-
gery and interventional neuroradiology. In: Byrne JV
(ed) Vascular malformations of the brain. Springer
Heidelberg Dordrecht , London, New York
66. Carmeliet P (2000) Mechanisms of angiogenesis and
arteriogenesis. Nat Med 6(4):389–395
67. Mullan S et al (1996) Embryological basis of some
aspects of cerebral vascular fistulas and
malformations. J Neurosurg 85(1):1–8
68. Risau W (1997) Mechanisms of angiogenesis. Nature
386(6626):671–674
69. Matsubara S et al (2000) Analysis of endoglin expres-
sion in normal brain tissue and in cerebral arteriove-
nous malformations. Stroke 31(11):2653–2660
70. Atri D et al (2013) Endothelial signaling and the
molecular basis of arteriovenous malformation. Cell
Mol Life Sci 71(5):867–883
71. Carmeliet P, Jain RK (2011) Molecular mechanisms
and clinical applications of angiogenesis. Nature
473(7347):298–307
72. Risau W et al (1988) Vasculogenesis and angiogenesis
in embryonic-stem-cell-derived embryoid bodies.
Development 102(3):471–478
73. Houdart E et al (1993) A proposed angiographic
classification of intracranial arteriovenous fistulae
and malformations. Neuroradiology 35(5):381–385
74. Niimi Y et al (1991) Pial arteriovenous fistula of the
brain. Neuroradiology 33 (Suppl) (Proceedings of the
XIV Symposium Neuroradiologicum): 545–547
75. Forsting M, Wanke I (2008) Intracranial vascular
malformations and aneurysms: from diagnostic
work-up to endovascular therapy, 2nd rev. edn. Med-
ical radiology – diagnostic imaging and radiation
oncology. Springer, Berlin, 291 p
76. Stapf C, Mohr JP (2000) New concepts in adult brain
arteriovenous malformations. Curr Opin Neurol
13(1):63–67
77. Choi JH, Mohr JP (2005) Brain arteriovenous
malformations in adults. Lancet Neurol 4(5):299–308
78. Valavanis A, Yasargil MG (1998) The endovascular
treatment of brain arteriovenous malformations. In:
Cohadon F et al (eds) Advances and technical
638
standards in neurosurgery, vol 34. Springer/Wien,
New York
79. Navarrete MG et al (2014) Brain vascular lesions: a
clinicopathologic, immunohistochemistry, and
ultrastructural approach. Ann Diagn Pathol
18(4):193–198
80. McCormick WF (1966) The pathology of vascular
(“arteriovenous”) malformations. J Neurosurg 24
(4):807–816
81. George U et al (2010) Susceptibility-weighted imag-
ing in the evaluation of brain arteriovenous
malformations. Neurol India 58(4):608–614
82. Letteboer TG et al (2006) Genotype-phenotype rela-
tionship in hereditary haemorrhagic telangiectasia.
J Med Genet 43(4):371–377
83. Mahmoud M et al (2010) Pathogenesis of arteriove-
nous malformations in the absence of endoglin. Circ
Res 106(8):1425–1433
84. Pawlikowska L et al (2005) Polymorphisms in
transforming growth factor-beta-related genes ALK1
and ENG are associated with sporadic brain arterio-
venous malformations. Stroke 36(10):2278–2280
85. Shenkar R et al (2003) Differential gene expression in
human cerebrovascular malformations. Neurosurgery
52(2):465–477, discussion 477–8
86. Sure U et al (2004) Hypoxia-inducible factor and
vascular endothelial growth factor are expressed
more frequently in embolized than in nonembolized
cerebral arteriovenous malformations. Neurosurgery
55(3):663–669, discussion 669–70
87. Sure U et al (2001) Endothelial proliferation,
neoangiogenesis, and potential de novo generation
of cerebrovascular malformations. J Neurosurg
94(6):972–977
88. Delgado Almandoz JE et al (2010) Practical scoring
system for the identification of patients with intrace-
rebral hemorrhage at highest risk of harboring an
underlying vascular etiology: the Secondary Intrace-
rebral Hemorrhage Score. AJNR Am J Neuroradiol
31(9):1653–1660
89. Sanelli PC, Mifsud MJ, Stieg PE (2004) Role of CT
angiography in guiding management decisions of newly
diagnosed and residual arteriovenous malformations.
AJR Am J Roentgenol 183(4):1123–1126
90. Matsumoto M et al (2007) Dynamic 3D-CT angiog-
raphy. AJNR Am J Neuroradiol 28(2):299–304
91. Gross BA, Frerichs KU, Du R (2012) Sensitivity of
CT angiography, T2-weighted MRI, and magnetic
resonance angiography in detecting cerebral arterio-
venous malformations and associated aneurysms.
J Clin Neurosci 19(8):1093–1095
92. Meretoja A et al (2014) The spot sign and tranexamic
acid on preventing ICH growth–AUStralasia Trial
(STOP-AUST): protocol of a phase II randomized,
placebo-controlled, double-blind, multicenter trial.
Int J Stroke 9(4):519–524
93. Kikuta K et al (2006) Early experience with 3-T
magnetic resonance tractography in the surgery of
cerebral arteriovenous malformations in and around
D. Sahlein and N. Manning
the visual pathway. Neurosurgery 58(2):331–337,
discussion 331–7
94. Latchaw RE et al (1995) Functional magnetic reso-
nance imaging as a management tool for cerebral
arteriovenous malformations. Neurosurgery
37(4):619–625, discussion 625–6
95. Guo Y et al (2012) Silent intralesional
microhemorrhage as a risk factor for brain arteriove-
nous malformation rupture. Stroke 43(5):1240–1246
96. Griffiths PD et al (2000) Brain arteriovenous
malformations: assessment with dynamic MR digital
subtraction angiography. AJNR Am J Neuroradiol
21(10):1892–1899
97. D’Aliberti G et al (2015) Arterial and venous aneu-
rysms associated with arteriovenous malformations.
World Neurosurg 83(2):188–196
98. Redekop G et al (1998) Arterial aneurysms associated
with cerebral arteriovenous malformations: classifica-
tion, incidence, and risk of hemorrhage. J Neurosurg
89(4):539–546
99. Okamoto S, Handa H, Hashimoto N (1984) Location
of intracranial aneurysms associated with cerebral
arteriovenous malformation: statistical analysis.
Surg Neurol 22(4):335–340
100. Lasjaunias P et al (1988) Cerebral arteriovenous
malformations (C. AVM) and associated arterial aneu-
rysms (AA). Analysis of 101 C. AVM cases, with
37 AA in 23 patients. Acta Neurochir (Wien)
91(1–2):29–36
101. Rutledge WC et al (2014) Hemorrhage rates and risk
factors in the natural history course of brain
arteriovenous malformations. Transl Stroke Res
5(5):538–542
102. Fiehler J, Stapf C (2008) ARUBA–beating natural
history in unruptured brain AVMs by intervention.
Neuroradiology 50(6):465–467
103. Cockroft KM et al (2012) A perfect storm: how a
randomized trial of unruptured brain arteriovenous
malformations’ (ARUBA’s) trial design challenges
notions of external validity. Stroke 43(7):1979–1981
104. Mocco J et al (2013) Randomized clinical trials: the
double edged sword. J Neurointerv Surg
5(5):387–390
105. Amin-Hanjani S (2014) ARUBA results are not appli-
cable to all patients with arteriovenous malformation.
Stroke 45(5):1539–1540
106. DeMeritt JS et al (1995) Outcome analysis of preop-
erative embolization with N-butyl cyanoacrylate in
cerebral arteriovenous malformations. AJNR Am J
Neuroradiol 16(9):1801–1807
107. Spetzler RF et al (1987) Surgical management of large
AVM’s by staged embolization and operative exci-
sion. J Neurosurg 67(1):17–28
108. Pasqualin A et al (1991) Treatment of cerebral arte-
riovenous malformations with a combination of pre-
operative embolization and surgery. Neurosurgery
29(3):358–368
109. Jafar JJ et al (1993) The effect of embolization with N-
butyl cyanoacrylate prior to surgical resection of
28 Brain Arteriovenous Malformations
cerebral arteriovenous malformations. J Neurosurg
78(1):60–69
110. Vinuela F et al (1991) Combined endovascular embo-
lization and surgery in the management of cerebral
arteriovenous malformations: experience with
101 cases. J Neurosurg 75(6):856–864
111. Ovalle F, Shay SD, Mericle RA (2012) Delayed intra-
cerebral hemorrhage after uneventful embolization of
brain arteriovenous malformations is related to vol-
ume of embolic agent administered: multivariate anal-
ysis of 13 predictive factors. Neurosurgery 70(2 Suppl
Operative):313–320
112. van Rooij WJ et al (2012) Curative embolization of
brain arteriovenous malformations with onyx: patient
selection, embolization technique, and results. AJNR
Am J Neuroradiol 33(7):1299–1304
113. Iosif C et al (2015) Endovascular transvenous cure for
ruptured brain arteriovenous malformations in com-
plex cases with high Spetzler-Martin grades.
J Neurosurg 122(5):1229–38
114. Taylor CL et al (2004) Complications of preoperative
embolization of cerebral arteriovenous
malformations. J Neurosurg 100(5):810–812
115. Debrun GM et al (1997) Embolization of the nidus of
brain arteriovenous malformations with n-butyl cya-
noacrylate. Neurosurgery 40(1):112–120, discussion
120–1
116. van Rooij WJ, Sluzewski M, Beute GN (2007) Brain
AVM embolization with Onyx. AJNR Am J
Neuroradiol 28(1):172–177, discussion 178
117. Perez-Higueras A, Lopez RR, Tapia DQ (2005)
Endovascular treatment of cerebral AVM: our experi-
ence with onyx. Interv Neuroradiol 11(Suppl
1):141–157
118. Pierot L et al (2005) Endovascular treatment of brain
arteriovenous malformations using onyx: preliminary
results of a prospective multicenter study. Interv
Neuroradiol 11(Suppl 1):159–164
119. Pierot L et al (2013) Endovascular treatment of brain
arteriovenous malformations using a liquid embolic
agent: results of a prospective, multicentre study
(BRAVO). Eur Radiol 23(10):2838–2845
120. Nagy G et al (2012) Stereotactic radiosurgery for
arteriovenous malformations located in deep critical
regions. Neurosurgery 70(6):1458–1469, discussion
1469–71
121. Flickinger JC et al (1996) A dose–response analysis
of arteriovenous malformation obliteration after
radiosurgery. Int J Radiat Oncol Biol Phys
36(4):873–879
122. Karlsson B, Lindquist C, Steiner L (1997) Prediction
of obliteration after gamma knife surgery for cerebral
arteriovenous malformations. Neurosurgery 40
(3):425–430, discussion 430–1
123. Chang JH et al (2000) Factors related to complete
occlusion of arteriovenous malformations after
639
gamma knife radiosurgery. J Neurosurg 93(Suppl
3):96–101
124. Ding D et al (2013) Radiosurgery for patients with
unruptured intracranial arteriovenous malformations.
J Neurosurg 118(5):958–966
125. Wegner RE et al (2011) A modified radiosurgery-
based arteriovenous malformation grading scale and
its correlation with outcomes. Int J Radiat Oncol Biol
Phys 79(4):1147–1150
126. Kano H et al (2013) Stereotactic radiosurgery after
embolization for arteriovenous malformations. Prog
Neurol Surg 27:89–96
127. Parkhutik Vet al (2012) Postradiosurgery hemorrhage
rates of arteriovenous malformations of the brain:
influencing factors and evolution with time. Stroke
43(5):1247–1252
128. Barker FG 2nd et al (2003) Dose-volume prediction
of radiation-related complications after proton beam
radiosurgery for cerebral arteriovenous
malformations. J Neurosurg 99(2):254–263
129. Pollock BE et al (2000) The rationale and technique
of staged-volume arteriovenous malformation
radiosurgery. Int J Radiat Oncol Biol Phys
48(3):817–824
130. Firlik AD et al (1998) Staged volume radiosurgery
followed by microsurgical resection: a novel treat-
ment for giant cerebral arteriovenous malformations:
technical case report. Neurosurgery 43(5):1223–1228
131. Karlsson B et al (2007) Is repeated radiosurgery an
alternative to staged radiosurgery for very large brain
arteriovenous malformations? J Neurosurg
107(4):740–744
132. Karlsson B et al (2005) Long-term results after frac-
tionated radiation therapy for large brain arteriove-
nous malformations. Neurosurgery 57(1):42–49,
discussion 42–9
133. Foote KD et al (2003) Salvage retreatment after fail-
ure of radiosurgery in patients with arteriovenous
malformations. J Neurosurg 98(2):337–341
134. Lindvall P et al (2003) Hypofractionated conformal
stereotactic radiotherapy for arteriovenous
malformations. Neurosurgery 53(5):1036–1042,
discussion 1042–3
135. Veznedaroglu E et al (2004) Fractionated stereotactic
radiotherapy for the treatment of large arteriovenous
malformations with or without previous partial embo-
lization. Neurosurgery 55(3):519–530, discussion
530–1
136. Huang PP et al (2012) Long-term outcomes after
staged-volume stereotactic radiosurgery for
large arteriovenous malformations. Neurosurgery
71(3):632–643, discussion 643–4
137. Schwyzer L et al (2012) Long-term results of gamma
knife surgery for partially embolized arteriovenous
malformations. Neurosurgery 71(6):1139–1147,
discussion 1147–8
 Dural AVF and Treatment
29
Pierre Guedin

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642 Intracranial arteriovenous dural shunts
(ICADS) account for 10–15 % of all intracra-
Physiopathology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642
nial vascular malformations.
Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 643 The natural history of ICADS is based on
Clinical Symptoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 646 the venous drainage and the presence or
absence of leptomeningeal venous reflux.
Natural History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 646
Benign ICADS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 646 CT and MR can provide useful informations
Malignant ICADS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 647 for diagnosis, but DSA is mandatory to analyze
Imaging Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 647 the lesion and to understand the venous drain-
Benign ICADS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 647 age of the shunt and of the brain.
Malignant ICADS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 648 ICADS are dynamic lesions, i.e., a benign
Digital Subtraction Angiography (DSA) . . . . . . . . . . . . 648 fistula can turn into a malignant ICADS with
Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 648 time, but may also regress spontaneously.
Transvenous Embolization . . . . . . . . . . . . . . . . . . . . . . . . . . . 648 The spontaneous morbi-mortality of the dis-
Transarterial Embolization . . . . . . . . . . . . . . . . . . . . . . . . . . . 650
Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650
ease dramatically increases in case of
Stereotactic Radiosurgery (SRS) . . . . . . . . . . . . . . . . . . . . 655 leptomeningeal venous reflux and require
thus a curative treatment.
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655
Endovascular approach is nowadays the
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655 first intention treatment, either transarterially
or transvenously. This should be proposed in
case of malignant ICADS and benign ICADS
with intolerable symptoms.

Keywords
Intracranial dural fistula • Cortical venous
drainage • Embolization • Glue • Onyx

P. Guedin (*)
Department of Diagnostic and Therapeutic
Neuroradiology, Hôpital Foch, Suresnes, France
e-mail: p.guedin@hopital-foch.org

# Springer Science+Business Media New York 2016 641

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_43
642
Introduction
Intracranial arteriovenous dural shunts (ICADS)
represent abnormal communications between
arteries and veins that are located inside the dura
mater. They account for 10–15 % of all intracra-
nial vascular malformations [1].
Arterial feeders are most frequently meningeal
branches arising from the external carotid artery
(ECA), the internal carotid artery (ICA), or the
vertebral artery (VA), while the venous drainage
can be taken in charge by either dural sinuses or
veins or by cerebral veins.
The term “shunt” encompasses two architec-
tural types: AV fistulas and niduses. Nidus-type
shunts are characterized by a network of abnormal
vessels placed between feeding arteries and
draining veins (either sinuses or leptomeningeal
veins), whereas fistulas are AV connections with
arterial feeders ending directly into draining
veins. In our daily practice, we regularly use the
term “dural fistula” only in this latter
angioarchitecture.
The venous drainage of ICADS influences the
natural history of these lesions [2–4]. If the
venous drainage is exclusively in dural sinuses
or dural veins, the venous drainage of the brain
is not compromised; those lesions are thus con-
sidered as “benign” because they do not carry a
risk of hemorrhage or neurological symptoms. If
however the shunt induces reflux into cortical
leptomeningeal or deep veins, this correlates
with a high spontaneous risk of morbi-mortality
related to venous congestion and rupture giving,
respectively, rise to seizures, deficits, or hemor-
rhage. Those lesions are thus considered as
“aggressive.”
Pediatric dural shunts are different from
adult ones: not only do they vary according
to the age groups that can be affected
(dural sinus malformations in neonates or
infants, juvenile dural shunts in infants,
adult-ype shunts in older children), but they
also have specific symptoms and natural
histories [5]. This group of lesions will not be
addressed here.
P. Guedin
Physiopathology
A wide range of pathologies or conditions (e.g.,
traumatisms, surgery, infectious diseases,
coagulopathy, dehydration, hormonal deficits)
have been reported in the medical history of
patients presenting with ICADS. However, not
all of the patients having suffered from these
diseases develop ICADS, even after many years
of evolution. This points out the fact that an indi-
vidual’s susceptibility influences the occurrence
of these acquired lesions.
Venous hypertension has been supposed to be
the main pathogenic factor in the creation of these
lesions. It can occur associated with
non-thrombotic diseases (like direct sinusal com-
pression by a space occupying lesion, or surgical
sinusal sacrifice) but is mainly considered to be
related to thrombotic events.
Three potential steps in the creation of ICADS
have thus been described. The first step is a sinusal
thrombophlebitis and/or stenosis reducing the
venous outflow and inducing a venous hyperten-
sion. The second step consists in the opening of
microshunts inside the dura mater related to this
venous hypertension and congestion, associated
with the release of angiogenic factors due to the
thrombus itself (as vascular endothelial growth
factor (VEGF), basic fibroblast growth factor
(FGF)) and to the ischemic brain parenchyma.
The third step is the recanalization of the sinus.
Incomplete recanalization associated to the shunts
and arterialization of the venous outflow can pro-
voke a reflux in leptomeningeal veins and give
rise to an aggressive dural shunt with neurological
consequences [6].
Even if conceptually seducing, this initial
hypothesis has been counterbalanced by clinical
evidences. In fact very few patients suffering from
thrombophlebitis will develop an ICADS, and the
vast majority of patients with ICADS do not pre-
sent clinical and radiographical signs of present or
previous thrombophlebitis [1]. Other hypotheses
have thus been imagined to understand the crea-
tion of ICADSs: inflammation occurs at the pen-
etrating point of the emissary vein on the dura due
29 Dural AVF and Treatment 643

to idiopathic or secondary causes. Local inflam- Table 1 Classification of ICADS

matory reactions induce vessel dilatation and ICADS Classification
neovascularization and subsequently create arte- Venous
riovenous connections on the arteriole level. Localization drainage Borden Lariboisière
Although emissary vein (EV) communicating Dural sinus Sinusal, 1 I
with dural arteries might play a role as draining anterograde
Dural sinus Sinusal, 1 IIa
routes at first, they start to degrade due to com-
anterograde/
pression of enlarged emissary arteries or to a retrograde
hemodynamic shift to the drainage pathway of Dural sinus Sinusal, 1 IIa
least resistance following, the occlusion of drain- (occluded) retrograde
age pathway through EV into the sinus or cortical Dural sinus Sinusal, 2 IIb
veins may form, resulting in clinically detectable anterograde,
CVR
ICADS. The shunt then expands to the surround-
Dural sinus Sinusal, 2 II a + b
ing dura associated with recruitment of feeders anterograde/
from distant sites induced by expression of angio- retrograde,
genic factors and a shift in the hemodynamic CVR
balance. In sinus-type ICADS, the sinus is pro- Cortical vein CVR 3 III
gressively compartmentalized and finally Dural sinus CVR 3 III
(tandem
occludes due to thrombogenesis with activated occlusion with
coagulopathy or to hemodynamic hypertrophy of CVR)
the sinus wall. This progression results in the Ectatic CVR 3 IV
mature, aggressive fistula with drainage impair- cortical vein
ments. Previous mechanistic hypotheses focusing Perimedullary CVR 3 V
vein
on sinus hypertension and sinus thromboses can-
not explain the pathogenesis of non-sinus type of
ICADS [7].

Classification

ICADSs are most frequently classified according

to either Lariboisière or Borden classification
(Table 1). Both are based on the analysis of the
angiographic venous drainage and correlate to the
natural history [8, 9]. The leptomeningeal or deep
venous reflux appears to be the best predictive
factor of the severity of the lesion. The “benign”
types are represented by Lariboisière I (Fig. 1) and
IIa and Borden 1, whereas “malignant” types
encompass Lariboisière IIb (Fig. 2), III (Fig. 3),
IV, and V and Borden 2 and 3 [3, 10, 11].
Recently Lasjaunias et al. proposed a revised
classification of cranial and spinal dural shunts
Fig. 1 Cerebral angiogram, lateral view. Selective
based on embryologic and anatomical specific- catheterism of the occipital artery: intracranial dural arte-
ities [12]. This classification considers epidural riovenous shunt of the lateral sinus with exclusive sinusal
compartments in three groups: the ventral drainage
644
Fig. 2 Cerebral angiogram, lateral view, selective
catheterism of the maxillary (a) and occipital artery (b).
Intracranial dural arteriovenous shunt of a thrombosed
epidural, the dorsal epidural, and lateral epidural
group. Three observations of the venous develop-
ment of the brain and spinal cord and adjacent
bony structures contributed to the generation of
this classification.
The venous system of the notochord and
corresponding sclerotome extends from the
basisphenoid (cavernous plexus) to the sacrum
and gives rise to the ventral epidural drainage
group. It collects the blood from spongious bony
structures and has no primary role in the drainage
of the central nervous system. It includes the
basioccipital bone, the petrous pyramid, the
basisphenoid with its adjacent sphenoid wings,
and their related dural structures. The venous
afferents of these regions are closely related to
P. Guedin
lateral sinus with reflux into the vein of Labbé.
Posttreatment lateral view of the cast of glue into the lateral
sinus (“sinus suspendu”)
the bony structures, and they drain outside the
bony limits, thus resulting in no subarachnoid
(spinal or cortical) venous reflux. However,
some factors may precipitate development of
CVR in these regions such as thrombosis within
the epidural space surrounding, remote, or distal
to the shunt. This will produce reflux in the lateral
epidural space as seen in rare cases of ventral
epidural spinal shunts with perimedullary venous
reflux. CVR can also be encountered in high-flow
shunts forcing the emissary-bridging vein open-
ing after reflux in the lateral epidural space. Ven-
tral epidural shunts are characterized by a strong
female predominance and absence of CVR unless
extensive thrombosis of epidural drainage or
high-flow shunts.
29 Dural AVF and Treatment
Fig. 3 Cerebral angiography, lateral view. Selective
catheterism of the middle meningeal artery. Intracranial
dural arteriovenous shunt of the dural wall of the lateral
sinus with direct fistula into a temporal vein with secondary
opacification of the superior longitudinal sinus and lateral
sinus
The dorsal epidural venous space includes the
superior sagittal sinus (SSS), torcular, transverse
sinus, medial occipital sinus, and posterior mar-
ginal sinus at the foramen magnum. It is poorly
developed at the spinal level, and the presence of
dorsally located dural sinuses intracranially is
therefore the major difference between the venous
systems of the brain and spine. Their formation is
linked to the appearance, during evolution, of the
dural falces and tentorium and associated with the
development of the paleo- and neopallial struc-
tures. These sinuses result from the confluence in
the epidural space of two different venous sys-
tems: the osseous system draining the cranial
vault and the leptomeningeal system draining the
brain. The venous pressure within the dural
sinuses is usually low; thus, antegrade and
craniofugal flow of the shunts is normally
observed. CVR may occur after associated venous
outflow restriction or with high-flow shunts; in
case of variations of the venous opening into this
system such as presence of an accessory epidural
sinus, distal restriction of the sinus opening will
rapidly produce CVR. Dorsal epidural shunts
have no sex predominance, present at lower
645
mean age, and can be multiple. Usually no
leptomeningeal venous reflux is seen, and
craniofugal flow is normally observed.
Leptomeningeal reflux may occur following asso-
ciated venous outflow restriction or with high-
flow shunts. Variations of the venous opening
into this system or distal restriction of the sinus
opening will also be responsible for cortical
venous reflux.
The veins draining the central nervous system
(both spinal and cranial) are not related to
the peripheral nerves but use “emissary” or
“bridging veins.” These latter join the lateral
epidural venous spaces as connecting drainage
system. The leptomeningeal venous drainage
has no direct confluent communication with
the ventral and dorsal epidural venous
plexuses, which drain the skull and spine.
While the typical lateral epidural shunts corre-
spond to spinal dural fistulas that drain into emis-
sary or radicular veins joining the perimedullary
venous plexuses, intracranial pathological loca-
tions involve emissary-bridging veins of the
brainstem and their homologs draining deep
cerebral structures such as the condyloid vein
at the foramen magnum, the superior petrosal
vein, the basal vein, the vein of Galen, the veins
of the anterior cranial fossa, and the orbit. Thus,
the following DAVS locations are encountered:
marginal sinus, falco-tentorial angle, petrosal
and basi-tentorial areas, Breschet sinus,
paracavernous and intraorbital regions, and lam-
ina cribriformis. The drainage is always directed
towards leptomeningeal veins, and this type of
DAVS is therefore always considered to be
aggressive. Lateral epidural shunts are character-
ized by a male predominance and present at
later ages.
The significant difference of the patient char-
acteristics (sex and age) between the groups of the
new classification suggests the existence of fun-
damental differences, either biological or hor-
monal, within the epidural venous systems in
each region and may point towards different eti-
ologies of these three different groups of arterio-
venous shunts. Furthermore, this classification
allows a better understanding of various physio-
pathological phenomena.
646 P. Guedin

Clinical Symptoms Table 2 Differential diagnosis of pulsatile tinnitus

Pulsatile tinnitus etiologies
Two modes of presentation can be distinguished: Arterial Venous Nonvascular
those due to hemorrhage and those due to venous Developmental Benign Myoclonia
congestion [2]. Hemorrhagic complications, i.e., Aberrant intracranial
internal carotid hypertension
parenchymal, subarachnoid, or subdural hemor-
artery
rhages, present acutely in a stroke-like manner, Persistent
whereas nonhemorrhagic neurological presenta- stapedial artery
tions have often a subacute or slowly progressive Neurovascular
conflict
onset. Cranial nerve palsy, ocular symptoms,
Arteriopathy Venous Semicircular
dementia, seizures, and encephalopathy are the
Atheroma dehiscent canal
usual symptoms associated to venous Fibromuscular jugular bulb dehiscence
hypertension. dysplasia thrombophlebitis
Those symptoms will vary accordingly to the Dissection
Aneurysm
localization and the type of the shunt.
Vascular tumors
Pulsatile tinnitus (PT) is due to arterialization Paraganglioma
and flow turbulences in the shunting zone and the Endolymphatic
surrounding venous structures, which can be trans- sac tumor
mitted to the petrous bone. They are presented in Meningioma
Hemangioma
60 % of patients, mainly those presenting with Hypervascular
direct sinusal drainages. The differential diagnoses metastasis
of PT are listed in Table 2 [1, 10]. Osseous
Cavernous sinus ICADS typically present with dysplasia
ophthalmoplegia, proptosis, chemosis, retro- Otospongiosis
Paget disease
orbital pain, or decreased visual acuity.
Vascular shunts
During the follow-up of non-treated patients Arteriovenous
with ICADS, any modification of the noise inten- malformation
sity (increase, decrease, or disappearance) should Dural fistulas
be explored by a MRI/MRA and/or DSA. Indeed,
changes in the architecture of the ICADS as
thrombosis of the sinus may reroute the venous 73 patients (62 %). Intolerable bruit or ophthal-
drainage through cortical veins and would put the mological sequelae were deemed indications for
patient in a risky situation because of the transfor- palliative embolization in 43 patients and surgical
mation of the lesion from a benign into an aggres- treatment in one patient. A median follow-up
sive one. period of 27.9 months (range 1 month – 17.5
years) was available in 112 patients (95.7 %),
among whom repeated angiography was
Natural History performed in 50. Overall, observational and palli-
ative management resulted in a benign and toler-
Benign ICADS able level of disease in 110 (98.2 %) of 112 cases.
In two cases managed conservatively CVR devel-
Satomi et al. reported a series of 117 patients oped with spontaneous progressive thrombosis of
harboring benign cranial DAVFs. At the initial venous outlets. The disease course of a cranial
assessment, four patients were asymptomatic. DAVF without CVD is indeed benign, obviating
Most patients presented with benign symptoms: the need for a cure of these lesions. However, a
chronic headache, bruit, or orbital phenomena. benign DAVF carries a 2 % risk of developing
Observational management was instituted in CVD, mandating close clinical follow-up and
29 Dural AVF and Treatment
renewed radiological evaluation in response to
any deterioration in the patient’s condition [4].
Malignant ICADS
The natural history of aggressive (Borden 2 and 3)
cranial dural arteriovenous fistulas is not well
described. Reported annual mortality and hemor-
rhage rates vary widely and range up to 20 % per
year. Van Dijk et al. reported a consecutive single-
center cohort of 236 cases that presented with a
cranial DAVF between June 1984 and May 2001.
A group of 118 cranial DAVFs was selected for the
presence of CVR. All patients were offered treat-
ment aimed at the disconnection of the CVR.
Patients who declined or had partial treatment with
persistence of the CVR had long-term clinical and
angiographic follow-up to study the disease course
of this select group. Fourteen non-treated patients
(11.9 %) and six partially treated patients were
assessed clinically and angiographically over time.
The mean follow-up in this select group was 4.3
years (86.9 patient-years). During follow-up, seven
patients suffered an intracranial hemorrhage (35 %).
The incidence of nonhemorrhagic neurological def-
icit was 30 %. Nine patients (45 %) died: six patients
expired after a hemorrhage, and three patients died
of progressive neurological deterioration. Two
patients demonstrated a spontaneous closure of the
DAVF (10 %). Authors concluded that persistence
of the CVR in dural shunts yields an annual mor-
tality rate of 10.4 %. Excluding events at presenta-
tion, in this series the annual risk for hemorrhage or
nonhemorrhagic neurological deficit during follow-
up was 8.1 % and 6.9 %, respectively, resulting in an
annual event rate of 15.0 % [13].
In a retrospective study of 85 patients suffering
from an ICADS with CVR, Soderman
et al. evaluated the records of 85 patients with
dural arteriovenous shunts with cortical venous
drainage or reflux hospitalized in their institution
from 1978 to 2007.
Fifty-three patients did not have an intracra-
nial hemorrhage as the presenting event. One of
these patients bled after diagnosis. Thirty-two
patients had an intracranial hemorrhage as the
647
presenting event. Three patients bled after diag-
nosis. One of these patients died. Apart from
deficits caused by hemorrhage, no patient
reported adverse neurological symptoms. In
patients presenting with an intracranial hemor-
rhage, the annual risk for hemorrhage is approx-
imately 7.4 %, and in those not presenting with a
hemorrhage, it is approximately 1.5 %. They
conclude that the risk of intracranial hemorrhage
from a dural arteriovenous shunt with cortical
venous drainage is most likely smaller than pre-
viously proposed, but presentation with hemor-
rhage is a risk factor for hemorrhage [14].
Duffau et al. reported a series of 20 patients
suffering from dural fistulas with cortical venous
drainage revealed by intracranial hemorrhage.
The mean duration between the first hemorrhage
and treatment was 20 days. Seven patients (35 %)
presented with acute worsening during this delay
due to radiologically proven early rebleeding with
graver consequences than the first hemorrhage.
They therefore advocate complete and early treat-
ment in all cases of AVF with cortical venous
drainage revealed by an ICH [15].
Imaging Strategy
Benign ICADS
For high-flow shunts, cervical Doppler may find
an increase of the output in the primitive carotid
and a diastolic flow in the external carotid artery.
Cerebral CT is usually normal.
Contrast-enhanced CT may reveal a mottling
of the skull base due to transosseous arterial
feeder arising from the occipital artery in lateral
sinus shunt (Fig. 4).
CTA may be useful in static depiction of
angioarchitecture, showing enlarged arterial
feeders and early opacification of the sinus and
the jugular vein.
MR parenchymal sequence is useful to rule out
the absence of parenchymal edema.
3D TOF may be positive in large shunt with
high flow but may also be negative for smaller or
slow-flow shunts.
648
Fig. 4 T1 post-contrast MR: transosseous arterial feeders
arising from the right occipital artery
3DMRA is useful for gross depiction of
angioarchitecture and dynamics.
For cavernous shunts, the signs are specific,
encompassing a dilated superior ophthalmic vein,
exophthalmos, and a dilated cavernous sinus.
Malignant ICADS
Cerebral non-contrast CT can reveal subarachnoid
hemorrhage, hematoma, and parenchymal edema.
Brain parenchyma calcifications reflecting the
chronicity of the venous hypertension could rarely
be seen. In cerebral atrophy, hydrocephaly can be
pointed out as consequences of the shunt on the
brain parenchyma.
Contrast-enhanced CT may show tortuous
feeding arteries and dilated and tortuous cortical
veins.
MR Flair sequences are useful to depict the
edema due to venous hypertension and ischemia
(Fig. 5). In case of type V fistulas, hyperintensity
in the cervical medulla could be found. As for
benign ICADS 3D TOF and MRA could be useful
P. Guedin
for gross depiction of angioarchitecture and
dynamics.
MRI/MRA is useful in post therapeutic follow-
up [16].
Digital Subtraction Angiography (DSA)
DSA is the gold standard modality to detect and
classify ICADS. It should be performed every
time an ICADS is suspected, in order to make
the diagnosis, to distinguish benign and aggres-
sive type of shunt and then to plan the treatment.
Selective serial should be performed in both
carotid and vertebral arteries, also in both external
carotid arteries. Long serial should be done in
order to understand the venous drainage of the
brain, the patency of the sinus, the flow direction
in each sinus, and the venous drainage of the
fistula itself. Selective serial should be done in
each arterial feeders of the region to depict spe-
cific angioarchitecture of the shunt.
Treatment
The therapeutic strategy will vary accordingly to
the venous drainage pattern. Indications of treat-
ment of benign ICADS are rare. They are limited
to intolerable clinical symptoms, mostly double
vision and pulsatile tinnitus. The risk of the treat-
ment should be extremely low and less than the
natural history of the disease itself. Aggressive
ICADS require curative treatment in order to dis-
connect the leptomeningeal reflux and to prevent
future intracranial hemorrhage or neurological
deficit. Several therapeutic approaches can be pro-
posed: surgery, radiosurgery, and embolization.
Endovascular treatment, transarterially or
transvenously, is often considered as the first
intention treatment.
Transvenous Embolization
Several modalities of transvenous embolizations
have been described: catheterization of the
affected sinus by venous route followed by
29 Dural AVF and Treatment 649

Fig. 5 MRI, T2 coronal sequence (a) and T1 post contrast anteroposterior view (c) and lateral view (d, e): dural
(b): right cerebellar edema in posterior fossa due to venous shunt with cortical reflux in cerebellar vein and venous
congestion of cerebellar vein. Cerebral angiography in congestion in the late venous phase
650
deposition of coils and/or liquid embolic agent
adjacent to the shunt, transcranial approach and
direct puncture of the pathological sinus if selec-
tive endovascular navigation is considered impos-
sible for anatomical or architectural reasons, or
selective obliteration of the CVR after intravenous
navigation without occlusion of the dural sinus
[17–19]. This latter technique remains excep-
tional because of the tortuosity and fragility of
the leptomeningeal veins and the risk of rupture
during selective catheterization. Transvenous
management is reserved to cases in which the
sinus is already partially thrombosed (Fig. 6). A
patent sinus draining ICADS should not be
occluded transvenously in first intention: the nor-
mal venous drainage of the brain could be
compromised and lead to delayed ischemic com-
plications. Furthermore, secondary development
of de novo ICADS can occur, leading then to
potential impairments in the intracranial venous
drainage if one sinus is already suppressed.
Transarterial Embolization
The concept of this approach is to superselectively
catheterize the arterial feeders and occlude via this
route the shunt itself. Nowadays, transarterial
embolization can be achieved with liquid embolic
agent (glue and Onyx) and particles. The use of
particles as sole embolic agent is getting more and
more restricted because of their instability and
short-term effects. They have thus no interest in
ICADS with CVR, and when needed, they are
mostly used in particular lesions as an adjuvant
embolus to reduce the flow in collateral vessels
and help to induce thrombosis.
All vessels potentially giving rise to arterial
feeders of the shunt are catheterized in order to
build up the lesional anatomy and architecture of
the lesion. Particular attention is paid to the
venous drainage of the brain as compared to the
drainage of the lesion.
With glue (n-BCA), when embolization is
decided, and according to the size and aspects of
the vessels, a microcatheter is used to
superselectively catheterize the main feeder of
P. Guedin
the shunt. Progression of the microcatheter is
helped by the use of a microguidewire in order
to allow distal catheterization as close as possible
to the shunt. The ideal injections of glue are
performed under a wedge position [20]. When a
safe embolization position is achieved, selective
injection of contrast is performed in order to
depict the focal angioarchitecture of the
malformative compartment. The dilution of the
cyanoacrylate glue mixed with Lipiodol will
vary accordingly to the position of the catheter,
the size of the vessel, and the flow of the shunt. n-
BCA is injected in liquid form and solidifies on
contact with ionic solutions as blood. The injec-
tion duration is relatively short. Glue presents a
thrombogenic effect that can promote secondary
occlusion of residual shunt [21] (Figs. 7 and 8).
Onyx is a copolymer of ethylene-vinyl alcohol
mixed with a solvent, DMSO and tantalum pow-
der, that reacts as a gelling agent. Mixed with
blood, the solvent is volatile and disappears in
the arterial flow. The deposit of the polymer is
slow. The polymer initially precipitates within the
peripheral area of blood vessel with secondary
occlusion of the central vessel. Onyx injection
can alternatively be stopped and started again,
with the polymer going in another vessel. This
technique allows to maintain the dural sinus pat-
ent with focusing on arterio-arterial reflux sur-
rounding the shunting zone [22].
Surgery
When surgical management is decided, the oper-
ative technique has to be adapted to the type of
venous drainage of the lesion: in aggressive dural
shunts with leptomeningeal venous drainage, its
aim is to disconnect the origin of the draining
vein. First described by Grisoli in 1984, surgery
is considered to be the efficacious way of curing
the ICADS. If the shunt drains first into the sinus
and secondarily in cortical veins, by
disconnecting the CVR, the surgical technique
leaves the sinusal drainage intact, converts the
aggressive DAVF into a benign lesion, and
changes thus the natural history of the shunt [23].
29 Dural AVF and Treatment 651

Fig. 6 (continued)
652
Fig. 6 Cerebral angiogram, selective catheterism of left
maxillary (a, b) and occipital artery (c) revealing a left
dural cavernous shunts with reflux into superior and infe-
rior ophthalmic vein and cortical venous reflux into the left
sylvian vein. Selective catheterism of petrosquamous
branch of the middle meningeal artery (d) showing more
precisely the venous congestion of ophthalmic and left
hemispheric veins. The left inferior petrous sinus is
P. Guedin
occluded. Treatment: transvenous approach through the
inferior petrosal sinus and selective coiling of the cavern-
ous sinus (e), in association with a superselective injection
of glue inside the sinus. (f) Posttreatment selective injec-
tion of internal maxillary artery (g, h) revealed the occlu-
sion of the shunt. Appreciate the specific exophthalmos
with vascular congestion of the left eye (i) pre- and
posttreatment (j)
29 Dural AVF and Treatment 653

Fig. 7 Angiograms obtained in a 49-year-old man who arteriovenous shunt of fistulous type draining initially into a
presented with two recurrent hemorrhages in the posterior petrous vein and resulting in congestion of the posterior fossa
fossa. (a, b) Lateral views of the left internal maxillary (a) veins. Other feeding vessels arose from the left internal and
and occipital (b) arteries demonstrating a petrotentorial dural ascending pharyngeal arteries (not shown). (c): Left VA
654
Fig. 8 Angiograms obtained in a 75-year-old man who
presented with acute headaches due to a right frontal lobe
hemorrhage. (a) Lateral view of a right ICA injection
showing an ethmoidal shunt located at the level of the
lamina cribriformis. The fistula is vascularized by a poste-
rior ethmoidal artery arising from the ipsilateral ophthal-
mic artery (OA). It drains into a frontal vein harboring a
false aneurysm (indicating the rupture point of the lesion)
and reaches secondarily the superior sagittal sinus.
Fig. 7 (continued) angiogram in anteroposterior view
confirming poor filling of left hemispheric cerebellar
veins due to congestion of the veins because of the
ICADS. All of the feeding vessels converged towards a
single shunting zone. (d) Image obtained at the end of the
angiographic session showing the cast of glue injected into
the shunt. Distal catheterization of the basal tentorial
branch of the middle meningeal artery, which extended
into the fistula, was performed with a Magic 1.2
microcatheter and a 50 % mixture of NBCA, and Lipiodol
was injected in a wedge position to cast the shunt and
penetrate into the proximal portion of the petrous vein,
P. Guedin
(b) Image obtained after superselective catheterization of
the ethmoidal feeding vessel was performed via the OA
and confirmed this architecture. (c) Image showing embo-
lization of the whole shunt performed in wedge position
with NBCA. (d) Right ICA angiogram obtained immedi-
ately after treatment confirming cure of the shunt with
respect to the OA and its branches. The patient remained
asymptomatic. Follow-up MR imaging and MR angiogra-
phy obtained 3 months later confirmed cure of the lesion
thus occluding the ICDAVS. (e–g) Immediate postembo-
lization images confirming cure of the fistula. Injection of
all the previously described arteries, internal maxillary
artery (e), occipital artery (f), as well as the internal carotid
artery (ICA) and ascending pharyngeal artery (not shown),
does not opacify the shunt anymore but fills all the regional
arteries without opacification of the venous drainage. Late
phase of the postembolization left VA angiogram (g) shows
normal filling of the hemispheric veins, confirming the
suppression of the venous congestion. A 5-month control
angiogram (not shown) confirmed total obliteration of the
fistula. The patient returned to normal life
29 Dural AVF and Treatment
Stereotactic Radiosurgery (SRS)
SRS provides effective long-term relief of symp-
toms. It cannot be applied however to every type of
ICADS and should be reserved to benign fistulas
because of the delay needed before obtaining the
classical radiation-induced obliterative arteritis that
leads to cure of the shunt. The poor natural history
of ICADS with CVR requires indeed more emer-
gent treatments and rapid results. Furthermore, the
results obtained by radiosurgery seem
unpredictable, as 30 % of lesions are not cured by
this technique. CVR, a risk factor for intracranial
hemorrhage, is considered thus as a relative contra-
indication to radiosurgery; if needed, it should be
proposed as a complementary treatment for failures
of surgical or endovascular treatments [24, 25].
Summary
The natural history of ICADS is based on the
venous drainage. ICADS are dynamic lesions, i.
e., a benign fistula can turn into a malignant
ICADS with time, but may also regress spontane-
ously. The spontaneous morbi-mortality of the
disease dramatically increases in case of
leptomeningeal venous reflux and requires thus a
curative treatment.
CT and MR can provide useful informations
for diagnosis, but DSA is mandatory to perform
the analysis of the lesion and to understand the
venous drainage of the shunt and of the brain.
Endovascular approach is nowadays the first
intention treatment, either transarterially or
transvenously. This should be proposed in case
of malignant ICADS and benign ICADS with
intolerable symptoms.
References
1. Berenstein A, Lasjaunias P, TerBrugge K (2004) Dural
arteriovenous shunts. Therapeutic aspects. In: Surgical
neuroangiography. Springer, Berlin, p 2.2:605
2. Van Rooij WJ, Sluzewski M, Beute GN (2007) Dural
arteriovenous fistulas with cortical venous drainage:
incidence, clinical presentation, and treatment. AJNR
Am J Neuroradiol 28(4):651–655
655
3. Davies MA, TerBrugge K, Willinsky R, Coyne T,
Saleh J, Wallace MC (1996) The validity of classifica-
tion for the clinical presentation of intracranial dural
arteriovenous fistulas. J Neurosurg 85(5):830–837
4. Satomi J, van Dijk JMC, Terbrugge KG, Willinsky RA,
Wallace MC (2002) Benign cranial dural arteriovenous
fistulas: outcome of conservative management based
on the natural history of the lesion. J Neurosurg
97(4):767–770
5. Berenstein A, Lasjaunias P, TerBrugge K (1987) Dural
arteriovenous shunts. In: Surgical neuroangiography.
Springer, Berlin, pp 389–451
6. Kraus JA, St€ uper BK, Nahser HC, Klockgether T,
Berlit P (2000) Significantly increased prevalence of
factor V Leiden in patients with dural arteriovenous
fistulas. J Neurol 247(7):521–523
7. Miyachi S, Izumi T, Matsubara N, Naito T,
Haraguchi K, Wakabayashi T (2011) Mechanism of
the formation of dural arteriovenous fistula: the role
of the emissary vein. Interv Neuroradiol J
Peritherapeutic Neuroradiol Surg Proced Relat
Neurosci 17(2):195–202
8. Borden JA, Wu JK, Shucart WA (1995) A proposed
classification for spinal and cranial dural arteriovenous
fistulous malformations and implications for treatment.
J Neurosurg 82(2):166–179
9. Cognard C, Gobin YP, Pierot L, Bailly AL, Houdart E,
Casasco A et al (1995) Cerebral dural arteriovenous
fistulas: clinical and angiographic correlation with a
revised classification of venous drainage. Radiology
194(3):671–680
10. Davies MA, Saleh J, Ter Brugge K, Willinsky R,
Wallace MC (1997) The natural history and manage-
ment of intracranial dural arteriovenous fistulae.
Part 1: benign lesions. Interv Neuroradiol J
Peritherapeutic Neuroradiol Surg Proced Relat
Neurosci 3(4):295–302
11. Davies MA, Ter Brugge K, Willinsky R, Wallace MC
(1997) The natural history and management of intra-
cranial dural arteriovenous fistulae. Part 2: aggressive
lesions. Interv Neuroradiol J Peritherapeutic
Neuroradiol Surg Proced Relat Neurosci 3(4):303–311
12. Geibprasert S, Pereira V, Krings T, Jiarakongmun P,
Toulgoat F, Pongpech S et al (2008) Dural arteriove-
nous shunts: a new classification of craniospinal epi-
dural venous anatomical bases and clinical
correlations. Stroke J Cereb Circ 39(10):2783–2794
13. Van Dijk JMC, terBrugge KG, Willinsky RA, Wallace
MC (2002) Clinical course of cranial dural arteriove-
nous fistulas with long-term persistent cortical venous
reflux. Stroke J Cereb Circ 33(5):1233–1236
14. Söderman M, Pavic L, Edner G, Holmin S, Andersson
T (2008) Natural history of dural arteriovenous shunts.
Stroke J Cereb Circ 39(6):1735–1739
15. Duffau H, Lopes M, Janosevic V, Sichez JP,
Faillot T, Capelle L et al (1999) Early rebleeding
from intracranial dural arteriovenous fistulas: report
of 20 cases and review of the literature. J Neurosurg
90(1):78–84
656
16. Mossa-Basha M, Chen J, Gandhi D (2012) Imaging of
cerebral arteriovenous malformations and dural arte-
riovenous fistulas. Neurosurg Clin N Am 23(1):27–42
17. Roy D, Raymond J (1997) The role of transvenous
embolization in the treatment of intracranial dural arte-
riovenous fistulas. Neurosurgery 40(6):1133–1141
18. Luo C-B, Chang F-C, Wu H-M, Chung W-Y (2007)
Transcranial embolization of a transverse-sigmoid
sinus dural arteriovenous fistula carried out through
a decompressive craniectomy. Acta Neurochir
(Wien) 149(2):197–200
19. Mironov A (1998) Selective transvenous embolization
of dural fistulas without occlusion of the dural sinus.
AJNR Am J Neuroradiol 19(2):389–391
20. Nelson PK, Russell SM, Woo HH, Alastra AJG,
Vidovich DV (2003) Use of a wedged microcatheter
for curative transarterial embolization of complex
intracranial dural arteriovenous fistulas: indications,
endovascular technique, and outcome in 21 patients. J
Neurosurg 98(3):498–506
21. Guedin P, Gaillard S, Boulin A, Condette-Auliac S,
Bourdain F, Guieu S et al (2010) Therapeutic
P. Guedin
management of intracranial dural arteriovenous shunts
with leptomeningeal venous drainage: report of 53 con-
secutive patients with emphasis on transarterial embo-
lization with acrylic glue. J Neurosurg 112(3):603–610
22. Cognard C, Januel AC, Silva NA, Tall P (2008)
Endovascular treatment of intracranial dural arteriove-
nous fistulas with cortical venous drainage: new man-
agement using Onyx. AJNR Am J Neuroradiol
29(2):235–241
23. Da Costa LB, Terbrugge K, Farb R, Wallace MC
(2007) Surgical disconnection of cortical venous
reflux as a treatment for Borden type II dural arterio-
venous fistulae. Acta Neurochir (Wien)
149(11):1103–1108
24. Koebbe CJ, Singhal D, Sheehan J, Flickinger JC,
Horowitz M, Kondziolka D et al (2005) Radiosurgery
for dural arteriovenous fistulas. Surg Neurol
64(5):392–398 (discussion 398–399)
25. Söderman M, Edner G, Ericson K, Karlsson B, Rähn T,
Ulfarsson E et al (2006) Gamma knife surgery for dural
arteriovenous shunts: 25 years of experience.
J Neurosurg 104(6):867–875
 Carotid Cavernous Fistula
30
Timothy R. Miller, Ravishankar Shivashankar, Gaurav Jindal,
and Dheeraj Gandhi

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658 Carotid cavernous fistulas (CCF) are complex
lesions involving the abnormal shunting of
Anatomy of the Cavernous Sinus
and Cavernous Segment of the Internal Carotid
arterial blood into the cavernous sinus. There
Artery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658 are two distinct types: one is a direct commu-
nication between the internal carotid artery and
Pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
the cavernous sinus, while the other is a dural
Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 660 shunt involving meningeal branches of the
Etiology and Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . 662 external and/or internal carotid arteries.
Clinical Presentations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664 Resulting patient symptoms as well as poten-
tial complications is determined primarily by
Complications and High-Risk Features . . . . . . . . . . . 665
the degree of arteriovenous shunting as well as
Imaging Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 667 the direction of venous outflow from the cav-
Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 670 ernous sinus. Catheter angiography remains
Conservative Management . . . . . . . . . . . . . . . . . . . . . . . . . . 670 the gold standard imaging evaluation of
General Considerations of Endovascular CCFs, although both MR and CT angiography
Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671
Endovascular Treatment of Direct CCF . . . . . . . . . . . . . 673
can be used to screen patients in whom the
Endovascular Treatment of Indirect or diagnosis is suspected. Current treatment for
Dural CCF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 676 these lesions consists primarily of conservative
Surgical Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 677 management and endovascular embolization.
Stereotactic Radiotherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . 678
The latter can be performed via transarterial
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679 and/or transvenous approaches, using various
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 680 materials such as detachable balloons, coils,
and liquid embolic agents. A multidisciplinary
approach, including neurointerventionalists,
neurosurgeons, and ophthalmologists, is often
required to achieve the best outcomes for
T.R. Miller (*) • R. Shivashankar • G. Jindal • D. Gandhi patients.
Department of Diagnostic Radiology, Neuroradiology,
University of Maryland Medical Center, Baltimore, MD, Keywords
USA
Carotid cavernous fistula • Endovascular
e-mail: Tmiller5@unm.edu; trmiller12@gmail.com;
Rshiva@umm.edu; Gjindal@umm.edu; embolization • Cortical venous reflux • Intra-
Dgandhi@umm.edu cranial hemorrhage
# Springer Science+Business Media New York 2016 657
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_7
658
Introduction
A carotid cavernous fistula (CCF) is an abnormal
arteriovenous shunt involving the cavernous sinus
[1, 2]. The term actually applies to two distinct
lesions, which vary in their etiology, potential
clinical presentations, treatment, and outcomes.
The first is a direct CCF, which consists of a tear
in the cavernous segment of the internal carotid
artery (ICA) with direct shunting of arterial blood
into the sinus [2]. The second is an indirect CCF,
which represents a dural arteriovenous fistula
involving the cavernous sinus wall, which is fed
by meningeal arterial branches that normally sup-
ply this region [1, 2]. Despite their differences,
both lesions involve shunting of arterial blood into
the normally low-pressure cavernous sinus, with
resulting symptoms and complications deter-
mined primarily by the degree and route of venous
drainage from the diseased sinus [3, 4]. The fol-
lowing chapter explores the characteristics of both
types of CCF, highlighting their differences as
well as similarities.
Types of Carotid Cavernous Fistula
• Direct fistulas: A direct fistulous com-
munication between the cavernous seg-
ment of the internal carotid artery and the
surrounding cavernous sinus
• Indirect fistulas: A dural arteriovenous
fistula involving meningeal branches of
the internal and external carotid arteries
that normally supply the region of the
cavernous sinus
Anatomy of the Cavernous Sinus
and Cavernous Segment
of the Internal Carotid Artery
Cavernous Sinus: The cavernous sinus is a con-
tiguous network of trabeculated venous sinusoids
surrounded by dura matter that is located centrally
in the skull base adjacent to the pituitary sella and
sphenoid sinus [5–7]. It contains both the
T.R. Miller et al.
cavernous segment of the ICA and cranial nerves
III, IV, V1, V2, and VI [5–7]. Cranial nerves III,
IV, V1, and V2 run in the lateral wall of the sinus,
while VI is located within the sinus itself. The
cavernous sinus normally receives venous blood
from the ipsilateral orbit and brain, including the
superior ophthalmic, inferior ophthalmic, superfi-
cial middle cerebral, and deep middle cerebral
veins, as well as the sphenoparietal sinus
[4, 8]. Venous outflow from the cavernous sinus
is via superior and inferior petrosal sinuses to the
transverse sinus and internal jugular vein respec-
tively, as well as inferiorly via the pterygoid
venous plexus as well as emissary veins associ-
ated with the foramina ovale and rotundum
[4, 8]. Finally, the cavernous sinuses are
interconnected via a venous network centered
around the hypophysis, the so-called circular
sinus [2, 9]. Although the circular sinus generally
consists of a larger anterior and smaller posterior
limbs, either may be absent or hypoplastic in a
particular patient.
Contents of the Cavernous Sinus
• Cavernous segment of the internal
carotid artery
• Cranial nerve VI
• Cranial nerves III, IV, V1, and V2 in the
lateral wall of the sinus
Normal Hemodynamics of the Cavernous
Sinus
• Venous inflow
– Orbital veins: superior, inferior oph-
thalmic, and central retinal
– Superficial and deep middle cerebral
veins
– Sphenoparietal sinus
• Venous outflow
– Superior petrosal sinus
Drains to transverse sinus
– Inferior petrosal sinus
Drains to internal jugular vein
(continued)
30 Carotid Cavernous Fistula
– Pterygoid venous plexus
– Emissary veins around foramina
ovale and rotundum
• Intercavernous flow
– Circular sinus
Internal Carotid Artery: The ICA ascends ver-
tically in the skull base from the foramen lacerum
into the cavernous sinus, before turning anteriorly
into a horizontal segment approximately 2 cm in
length [6]. The artery then ascends superiorly in a
second vertical segment, extending along the
medial aspect of the anterior clinoid process
before exiting the sinus. The cavernous segment
of the ICA can be divided into five segments, as
detailed by Debrun et al. [10]. Going from caudal
to cranial, these include the posterior ascending
segment, junction of posterior ascending segment
and horizontal segment, horizontal segment, junc-
tion of the horizontal segment and anterior
ascending segment, and ascending segment
[10]. Although any segment of the cavernous
ICA may be involved by a direct CCF, the hori-
zontal portion is the most common site of
fistulization [10].
The cavernous segment of the ICA has two
main branches that, along with distal branches of
the internal maxillary artery from the external
carotid, supply the dura surrounding the cavernous
sinus as well as the nearby cranial nerves
[2, 5, 11]. Small anastomoses between these
branches of the external and internal carotid arteries
are normally present in the surrounding dura and
are responsible for the dural arteriovenous shunts
found in indirect CCF [1, 5, 11]. The first is the
meningohypophyseal trunk, present in 100 % of
patients, which arises from the apex of the proximal
curve of the vessel in the cavernous sinus [6]. This
artery in turn has three main branches: the artery to
the tentorium (Bernasconi-Cassonari), the dorsal
meningeal artery to the clivus, as well as the infe-
rior hypophyseal artery [5]. The second branch of
the cavernous ICA is the inferolateral trunk (ILT),
found in 84 % of patients, which originates from
the distal horizontal segment of the vessel and is the
659
most common branch to be involved by an indirect
CCF [5, 6]. The ILT also has three main branches:
superior or tentorial, anterior, and posterior
branches, which supply dura as well as cranial
nerves III, IV, V1, V2, and VI in and around the
cavernous sinus and orbit [5]. Finally, the cavern-
ous ICA may also give rise to McConnell’s capsu-
lar arteries, although these vessels are inconstantly
present [6].
Branches of the Cavernous Internal Carotid
Artery
• Meningohypophyseal trunk
– Artery to the tentorium (Bernasconi-
Cassonari)
– Dorsal meningeal artery
– Inferior hypophyseal artery
• Inferolateral trunk
– Superior or tentorial branch
– Anterior branch
– Posterior branch
• Capsular arteries
Pathophysiology
All CCFs demonstrate the same basic pathophys-
iologic mechanism of abnormal shunting of arte-
rial blood into the normally low-pressure venous
cavernous sinus. The resulting increase in pres-
sure and volume results in reversal of blood flow
in venous structures that normally drain the cav-
ernous sinus, including the superior and inferior
ophthalmic veins as well as the superior and infe-
rior petrosal sinuses [4]. The resulting signs and
symptoms associated with a CCF are determined
both by the degree of arteriovenous shunting and
the route of venous drainage from the cavernous
sinus [3, 4]. The latter is determined by the loca-
tion of the arteriovenous shunt, the presence of
thrombus in the involved sinus, as well as stenosis
or thrombosis of potential venous drainage path-
ways [10]. If the hemodynamic changes associ-
ated with a CCF persist over a long period of time,
the involved veins may become arterialized with
660
Fig. 1 Venous drainage pathways. (a) Frontal projection
and (b) lateral projection of a right internal carotid artery
angiogram demonstrating a direct right-sided CCF with
bilateral venous drainage via the circular sinus. Multiple
venous drainage pathways are illustrated, including the left
superior petrosal sinus (thin black arrows), left inferior
thickened, hyalinized walls [1]. Other chronic
changes include recruitment of capillary connec-
tions in the conjunctiva of the eye, which are
called specific limbal loops [1].
There are various patterns of venous drainage
from the cavernous sinus that can be encountered
with a CCF, which have important implications
for patient symptomatology as well as the risk of
intracranial complications (Fig. 1) [4, 10,
12]. Anterior drainage is commonly present,
with reflux of blood into the ipsilateral orbit via
the superior and inferior ophthalmic veins, with
subsequent drainage into angular and facial veins
[1, 12]. An additional anterior drainage route con-
sists of the sphenoparietal sinus running along the
anterior middle cranial fossa [2]. Posterior drain-
age from the cavernous sinus into the inferior and
superior petrosal sinuses is also often encoun-
tered, either alone or in combination with anterior
as well as contralateral drainage via the circular
sinus [2, 4, 10, 12]. It has been speculated that the
majority, if not all, of low-flow indirect CCF orig-
inally drain posteriorly, with anterior drainage
only commencing once the posterior pathway
becomes obstructed, often due to thrombosis
[12]. Finally, superior and inferior venous drain-
age from the cavernous sinus can proceed via the
Sylvian vein and pterygoid venous plexus,
respectively [4].
T.R. Miller et al.
petrosal sinus (thick black arrows), right superior ophthal-
mic vein (thin white arrows), right inferior ophthalmic vein
(thick white arrows), right superficial middle cerebral vein
(thin blue arrows), right deep middle cerebral vein (thin
red arrow), and pterygoid venous plexus (thick red arrow)
Potential Routes of CCF Venous Drainage
• Anterior drainage via the orbital veins
– Eventually draining into angular and
facial veins
• Posterior drainage
– Superior and inferior petrosal sinuses
• Contralateral drainage
– Circular sinus
• Superior drainage
– Superficial middle cerebral vein
• Inferior drainage
– Pterygoid venous plexus & emissary
veins
Classification
CCFs can be categorized using several criteria
[5, 13]. These include classifications systems
based on etiology (i.e., traumatic vs. spontaneous),
hemodynamics (high-flow vs. low-flow shunts), as
well as angiographic criteria [1, 5, 8, 13]. The latter
was characterized by Barrow et al. [5] and is based
on both the hemodynamics of the shunt as well as
the arterial supply to the fistula as demonstrated on
angiography [5].
30 Carotid Cavernous Fistula 661

Fig. 2 Type A direct CCF.

(a) Frontal projection and
(b) lateral projection of a
left internal carotid artery
angiogram demonstrating a
high-flow direct shunt
between the left internal
carotid artery and the
ipsilateral cavernous sinus.
Venous drainage is
primarily via the circular
and contralateral cavernous
sinuses

Fig. 3 Indirect type B

CCF. (a) Frontal projection
and (b) lateral projection of
a left internal carotid artery
angiogram demonstrating
an indirect CCF supplied by
meningeal branches of the
ICA arising from the
meningohypophyseal trunk
(thin black arrow) and
inferolateral trunk (thick
black arrow)

Fig. 4 Indirect type C CCF.

(a) Frontal projection of a
right internal carotid artery
angiogram demonstrating no
evidence of arteriovenous
early cavernous shunting.
(b) Frontal projection right
external carotid artery
angiogram demonstrating
dural shunts between
external carotid meningeal
branches and the ipsilateral
cavernous sinus (white
arrows)

Type A CCFs are high-flow direct shunts
between the ICA and the cavernous sinus
(Fig. 2). Type B CCFs are dural shunts between
meningeal branches of the cavernous ICA (arising
from the meningohypophyseal or ILT trunks) and
the cavernous sinus (Fig. 3). Type C CCFs are
dural shunts between external carotid meningeal
branches and the cavernous sinus (Fig. 4). Finally,
type D CCFs are dural shunts with arterial supply
from meningeal branches of both the ICA and
ECA (Fig. 5) [5]. Rarely, a direct fistula between
the ophthalmic artery and an ophthalmic vein can
present in a similar fashion to a CCF, a so-called
orbital shunt [1] (Fig. 6; Table 1).
662 T.R. Miller et al.

Fig. 5 Indirect type D

CCF. (a) Frontal projections
right internal carotid and (b)
right external carotid artery
angiograms demonstrating
an indirect CCF with
arterial supply from
meningeal branches arising
from both internal and
external carotid arteries
(white arrows)

Fig. 6 Cortical venous reflux. (a) Frontal projection right
external carotid artery angiogram and (b) flat panel com-
puted tomography demonstrates a type C indirect CCF
with cortical venous reflux (black arrows a, white
arrows b). (c) Axial FLAIR and (d) susceptibility MR
imaging demonstrates a resulting hemorrhagic brainstem
venous infarct (white arrows c, d). (e) Lateral right external
carotid artery angiogram and (f) axial FLAIR imaging
following coil embolization demonstrates no residual arte-
riovenous shunting (black arrows e) as well as resolution
of ischemic changes in the brain stem (white arrows f)

most commonly found in young males, presum-

Etiology and Epidemiology
Direct Fistulas: Direct CCFs often result from
blunt or penetrating trauma that results in a tear
in the cavernous segment of the ICA, with
resulting rapid arteriovenous shunting into the
cavernous sinus [8, 10, 14]. These lesions are
ably due to the increased prevalence of head
trauma in this population [13–15]. Debrun
et al. [10] found that the arterial tear in traumatic
fistulas can vary in size anywhere from 1 to 5 mm,
and rarely multiple rents may be present. Despite
the high-flow arteriovenous shunting associated
with direct fistulas, patients often present in a
30 Carotid Cavernous Fistula
Table 1 Barrow angiographic classification of carotid
cavernous fistulas
Fistula
type Arterial supply
Type A Direct Cavernous ICA
Type B Indirect Meningeal ICA branches
Type C Indirect Meningeal ECA branches
Type D Indirect Meningeal ICA and ECA
branches
somewhat delayed fashion, anywhere from days
to a few weeks following the inciting trauma
[16]. The mechanism of vessel injury in many of
these cases is likely a sudden increase in
intraluminal arterial pressure associated with the
traumatic event [17]. Penetrating trauma is
another possible cause, as well as rarely arterial
injury during skull base surgery [14–16,
18]. Finally, traumatic direct CCFs have been
associated with fractures involving the skull
base, with Liang et al. [16] in a retrospective
review reporting an 8.3 % incidence of direct
fistulas in patients with fractures involving the
middle cranial fossa, particularly those with a
transverse or oblique orientation.
Direct CCF fistulas can also be encountered in
patients without a history of preceding trauma
[8, 19]. These lesions are often encountered in
middle-age women, but are not exclusive to this
age group [20]. These spontaneous direct fistulas
have several potential etiologies, the most com-
mon of which is rupture of a cavernous segment
ICA aneurysm [18, 19]. Cavernous ICA aneu-
rysms represent somewhere between 1.9 % and
9 % of all intracranial aneurysms, are more com-
monly associated with CCF development as
opposed to subarachnoid hemorrhage, and give
rise to roughly 20 % of direct CCFs [14, 19, 21].
In a retrospective review of all cavernous
ICA aneurysms presenting to a single medical
center, Kupersmith et al. [21] found that 13 of
193 lesions were associated with a CCF (6.7 %).
However, it is important to remember that a
predisposing cavernous aneurysm may not be
identified following direct fistula formation due
to the presence of high-flow arteriovenous
shunting [22].
663
Etiologies of Direct Carotid Cavernous
Fistulas
• Traumatic
– Blunt
– Penetrating
– Iatrogenic
• Spontaneous
– ICA aneurysm rupture
– Collagen vascular disorders
Ehlers-Danlos type IV syndrome
Pseudoxanthoma elasticum
Fibromuscular dysplasia
– Minor episode of trauma or valsalva
Coughing and sneezing
An additional cause of spontaneous direct
CCFs includes various genetic syndromes that
can weaken the arterial wall and predispose to
rupture after minor trauma or episodes of valsalva,
including coughing or sneezing [1, 18, 19]. Spe-
cific genetic conditions that can affect vascular
wall structural integrity and have been associated
with CCF include fibromuscular dysplasia,
Ehlers-Danlos type IV syndrome, and
pseudoxanthoma elasticum [23–25]. Finally, a
direct CCF can be associated with a persistent
trigeminal artery extending from the cavernous
ICA to the basilar artery [22]. These so-called
trigemino-cavernous fistulas may arise either
from an inherent weakness in the aberrant vessel
wall or due to aneurysm formation at the
ICA-trigeminal artery branch point [22].
Indirect or Dural Fistulas: Indirect CCFs rep-
resent a subset of dural arteriovenous fistulas
involving the cavernous sinus, which receive
blood flow from meningeal branches of the inter-
nal and/or external carotid arteries that normally
supply the cavernous sinus dura [2, 26]. These
lesions most often occur spontaneously in post-
menopausal women [15, 26, 27]. However, as is
the case with direct CCFs, these lesions may pre-
sent in all age groups, including children and
infants, and can develop following minor trauma
or episodes of valsalva [12]. Altogether, dural
arteriovenous fistulas represent approximately
664
10–15 % of intracranial vascular malformations
[28]. Potential arterial feeders to indirect CCFs
include branches of the external carotid artery,
such as the internal maxillary, middle meningeal,
accessory meningeal, and ascending pharyngeal
arteries, as well as internal carotid artery branches
[26]. The latter include meningohypophyseal,
capsular, inferolateral trunk arteries, as well as
ethmoidal branches arising from the ophthalmic
artery [26].
As is the case with dural arteriovenous fistulas
found elsewhere in the intracranial compartment,
the etiology of indirect CCF remains uncertain
[12, 15]. One theory speculates that these fistulas
may result from breakdown in small thin-walled
dural arteries that normally cross the cavernous
sinus [12]. Alternatively, venous thrombosis
and/or elevated venous pressure may result in the
opening of normally closed, small anastomotic
channels in the dura [8, 12]. Risk factors for the
development of these dural fistulas include athero-
sclerosis, hypertension, diabetes, sinusitis, preg-
nancy, and collagen vascular disease [8, 27, 29].
Etiology of Indirect Carotid Cavernous
Fistulas
• Uncertain, two possible mechanisms
– Breakdown of small thin-walled dural
arteries that transverse the cavernous
sinus
– Opening of small anastomotic dural
channels
Venous sinus thrombosis
Elevated venous pressure
• Risk factors
– Atherosclerosis, diabetes, hyperten-
sion, and pregnancy
Clinical Presentations
The majority of symptomatic CCFs have anterior
drainage via the ophthalmic veins, with resulting
reflux of high-pressure arterial blood into the
ipsilateral orbit [3, 12, 15]. This arteriovenous
T.R. Miller et al.
shunting results in orbito-ocular congestion and
is responsible for the classic clinical triad asso-
ciated with CCFs, namely, pulsatile exophthal-
mos, orbital bruit, and chemosis of the ipsilateral
globe [1, 3, 4, 14, 19]. The contralateral orbit
may also be affected due to reflux of blood across
the circular sinus, occasionally in the absence of
ipsilateral orbital symptoms depending on the
pattern of venous drainage [1, 9, 12]. Other
signs and symptoms of orbito-ocular congestion
include medically refractory glaucoma,
ophthalmoplegia, retro-orbital pain, vision loss,
as well as dilatation and arterialization of con-
junctival and episcleral veins on ophthalmologic
exam [1, 12]. Although this latter finding may be
seen in the setting of other diseases such as
conjunctivitis, a tortuous corkscrew appearance
of these vessels is highly specific for the diagno-
sis of CCF [12]. General differential consider-
ations for physical exam findings suggestive of
orbito-ocular congestion include vascular
malformations involving the orbit or cavernous
sinus, cavernous sinus thrombosis, as well as
inflammatory processes such as scleritis, with
vortex vein blockage [1].
Some CCFs may present without classic symp-
toms, depending on the pattern of venous drainage
as well as the degree of arteriovenous shunting [3,
11, 15, 18, 28]. For example, low-flow indirect
CCFs with exclusively posterior venous drainage
characteristically lack signs and symptoms of
orbito-ocular congestion, including pulsatile
exophthalmos and orbital bruit [3, 4, 12, 29,
30]. Instead, these lesions often are either asymp-
tomatic or associated with nonspecific symptoms
including headache, tinnitus, trigeminal neuropa-
thy, facial nerve palsy, or isolated oculomotor palsy
due to the involvement of the corresponding cranial
nerves [11, 12, 30]. In these instances, patient
symptoms may be due, at least in part, to brainstem
congestion resulting from the posterior venous
drainage [12, 31]. The atypical presentation of
these posteriorly draining fistulas often leads to a
delay in their diagnosis, which is unfortunate
given the association between drainage via the
superior petrosal sinus and cortical venous reflux,
a high-risk feature discussed subsequently in this
chapter [1, 4].
30 Carotid Cavernous Fistula
Clinical Presentation of Carotid Cavernous
Fistulas
• Fistulas with anterior drainage via orbital
veins
– Pulsatile exophthalmos
– Chemosis
– Glaucoma
– Vision loss
– Ophthalmoplegia
• Fistulas with exclusively posterior drain-
age via petrosal sinuses
– Asymptomatic
– Headache
– Tinnitus
– Trigeminal neuropathy
– Isolated oculomotor palsy
Finally, symptom progression in CCFs is often
determined by the degree of associated arteriove-
nous shunting, with high-flow lesions often
presenting acutely with rapid deterioration, while
low-flow fistulas may demonstrate an insidious
onset and subsequent benign course
[1, 15]. Low-flow shunts are most often encountered
in the setting of an indirect, dural CCF, although
direct lesions may also behave in a similar manner
if the tear in the cavernous ICA is relatively small or
if there is partial thrombosis of the involved cavern-
ous sinus [1]. Recurrent fistulas following treatment
may also present insidiously, again depending on the
amount of residual arteriovenous shunting and avail-
able venous drainage pathways [22].
Complications and High-Risk Features
Ophthalmologic complications of CCFs include
vision loss, ophthalmoplegia, and medically
refractory glaucoma [1, 12]. Ophthalmoplegia
can result from either entrapment of the
extraocular muscles due to swelling of these struc-
tures or due to cranial nerve palsy secondary to
mechanical compression of the corresponding
nerve(s) [1, 12]. Glaucoma in CCFs may develop
from high orbital venous pressure, congestion of
665
the choroid or iris, displacement of the iris-lens
diaphragm, or neovascularity secondary to
chronic ischemia [12]. Finally, loss or deteriora-
tion of vision in the involved eye can be secondary
to venous stasis retinopathy with resulting retinal
ischemia, secondary glaucoma with optic nerve
damage, or spontaneous choroidal detachment [1,
11, 12]. Vision loss is more commonly seen with
direct CCF, although this finding may be encoun-
tered in as many as 20–30 % of indirect fistulas,
particularly chronic lesions [12].
Ophthalmologic Complications of Carotid
Cavernous Fistulas
• Glaucoma
– Elevated orbital venous pressure
– Choroid or iris congestion
– Iris-lens diaphragm displacement
– Neovascularity
• Ophthalmoplegia
– Extraocular muscle swelling
– Cranial nerve palsy
• Vision loss
– Venous stasis retinopathy
– Glaucoma with optic nerve damage
– Spontaneous choroidal detachment
Additional complications of CCFs include
intracranial hemorrhage, both intraparenchymal
and subarachnoid, venous infarct, epistaxis
(which may be fatal), as well as increased intracra-
nial pressure with venous hypertension [18, 26,
32]. Intraparenchymal hemorrhage is often a result
of reflux of high-pressure arterialized blood into
cortical veins overlying adjacent brain parenchyma
[1, 15, 18]. Posterior venous drainage of a CCF via
the superior petrosal sinus has been associated with
the development of this high-risk feature and may
be precipitated by the spontaneous thrombosis or
iatrogenic occlusion of alternative venous drainage
pathways [1, 4, 15, 18]. Overall, cortical venous
reflux has been reported in anywhere from 10 % to
55 % of CCFs, is often symptomatic due to
increased intracranial pressure, and carries up to a
30–40 % chance of hemorrhagic stroke if untreated
666 T.R. Miller et al.

[15, 18, 26, 33]. Furthermore, the reported annual

mortality rate of intracranial dural arteriovenous
fistulas with cortical venous reflux, regardless of
location, is 10.4 % [33].
Subarachnoid hemorrhage from a CCF has been
associated with the development of a cavernous
sinus varix or pseudoaneurysm (Fig. 7) [18]. In a
retrospective review of the angiographic features of
155 patients with CCFs, Halbach et al. [18] found
that cavernous sinus venous varix was present in
three of four patients who presented with subarach-
noid hemorrhage, all of which were fatal. Although
cavernous sinus varices were encountered in
patients without intracranial hemorrhage, the
authors argued the risk of fatal subarachnoid hem-
orrhage warranted emergent fistula treatment with
this finding is present [18]. Finally, drainage into
the sigmoid and transverse sinuses has been asso-
ciated with elevation of intracranial pressure from
resulting venous hypertension [18].
High-Risk Features of Carotid Cavernous
Fistulas
• Cortical venous reflux
– Between 10 % and 55 % of CCFs
– Up to 30–40 % risk of hemorrhagic
stroke untreated
– Associated with superior petrosal
sinus drainage
– Can arise from blockage of venous
drainage pathways
• Subarachnoid hemorrhage
– Rare, but often fatal
– Associated with cavernous sinus
varix
• Drainage into sigmoid and transverse
sinuses
– Risk of elevated intracranial pressure
from resulting venous hypertension

Fig. 7 Cavernous sinus

pseudoaneurysm. (a) Axial
and (b) coronal
reconstructed CT
angiography images show
an ectatic left cavernous
sinus (white arrow a) with
multiple lateral projecting
outpouchings (white arrows
b), which may represent
possible pseudoaneurysm
and/or draining venous
tributaries. (c) Frontal
projection and (d) lateral
projection of a left internal
carotid artery angiogram
demonstrate a direct CCF
with associated
pseudoaneurysm (black
arrow c)
30 Carotid Cavernous Fistula
Imaging Evaluation
Noninvasive Cross-Sectional Imaging: Although
many patients with a CCF present with classic
signs and symptoms suggestive of the diagnosis,
others may demonstrate either atypical symptom-
atology and/or an insidious, slowly progressive
course [26]. In these instances, conventional
MR, time-of-flight (TOF) and contrast-enhanced
MR angiography (MRA), Doppler ultrasound, as
well as CT angiography (CTA), may be used to
screen patients for the presence of an arteriove-
nous fistula involving the cavernous sinus
[34–36]. Conventional MR imaging finding sug-
gestive of a CCF includes prominent flow voids in
the involved cavernous or inferior petrosal sinuses
on spin-echo sequences, dilated intercavernous
venous channels, as well as the sequelae of
orbito-ocular congestion including enlargement
of the superior ophthalmic vein and extraocular
muscles and proptosis [37–39]. However, these
findings may only be present in high-flow fistulas,
with conventional MR imaging being relatively
Fig. 8 Conventional MR and MRA finding suggestive of
a right CCF. (a, b) Prominent flow voids are noted on spin-
echo sequences in the right cavernous and inferior petrosal
sinuses (white thick and thin arrows respectively a) as well
as dilated intercavernous venous channels (thin black
arrows b). (c) Sequelae of orbito-ocular congestion are
present in the ipsilateral orbit with enlargement of the
extraocular muscles (white thin arrows) and proptosis
667
insensitive for lesions demonstrating less rapid
arteriovenous shunting [40].
3D TOF MRA can increase the sensitivity for
detection of CCF by depicting flow-related
enhancement in the involved cavernous sinus as
well as arterial feeders in the setting of an indirect
fistula [35, 37, 40]. It is important to note however
that venous flow signal can be normally seen in
the cavernous and inferior petrosal sinuses on 3D
TOF MRA in the absence of a CCF [41]. In addi-
tion, 3D TOF MRA is limited by the lack of
temporal resolution, which precludes full charac-
terization of dynamic shunting lesions such as a
CCF [40]. However, time-resolved contrast-
enhanced MRA can help to overcome some of
these challenges by imaging the suspected dis-
eased cavernous sinus during passage of a gado-
linium contrast bolus [40]. This technique has
been shown to be both sensitive and specific for
the diagnosis of CCF and is a promising tool both
for screening patients as well as surveillance for
fistula recurrence following treatment
(Fig. 8) [40].
(horizontal white line). (d) Post-contrast MRA demon-
strates enhancement of the involved cavernous sinus
(white arrows). (e, f) Time-resolved contrast-enhanced
MRA demonstrates early opacification of right cavernous
sinus (thin white arrow e) as well as the fistulous shunt and
drainage via the superior ophthalmic vein (thick and thin
white arrows, respectively f)
668 T.R. Miller et al.

CT angiography has also been demonstrated to
be a promising technique for screening patients
suspected of having a CCF [42]. With the avail-
ability of 256 and 320 slice CT scanners, excellent
quality dynamic studies can be performed with
high spatial and acceptable temporal resolution.
CT and CT angiography findings of a CCF include
an enlarged cavernous sinus that demonstrates
early enhancement in the arterial phase, as well
as dilatation of draining venous tributaries (e.g.,
superior ophthalmic vein) [42, 43]. Chen
et al. [37] performed a retrospective study com-
paring noninvasive 3D TOF MRA and CTA to the
gold standard of catheter angiography. They
found that CTA performed better at detecting
CCF than TOF MRA, particularly for fistula
involving the more proximal aspect of the cavern-
ous ICA (Table 2).
Catheter Angiography: Despite advances in
cross-sectional imaging of CCF, catheter
angiography remains the gold standard for the
diagnosis and characterization of these lesions
due to its superior spatial and temporal resolution
[26, 40]. The diagnosis of a CCF is readily made
on catheter angiography by the demonstration of
abnormal arteriovenous shunting into the cavern-
ous sinus from either the ipsilateral cavernous
ICA or meningeal branches to the sinus wall in
the case of an indirect fistula. Goals of catheter
angiography when evaluating a CCF include the
identification of the exact site of fistulization,
evaluation of the degree of arteriovenous
shunting, determination of arterial supply to the
lesion, evaluation of the pattern of venous out-
flow, the presence of high-risk features including
cortical venous reflux and venous varix, potential
dangerous external or internal carotid artery anas-
tomoses, as well as the presence of atherosclerotic
disease if carotid compression is contemplated as
a potential treatment [26].

Table 2 Noninvasive imaging modalities for evaluation Catheter Angiography Evaluation of Carotid
of CCFs Cavernous Fistulas
Potential • Determination of type of fistula
Modality Findings disadvantages • Arterial supply with indirect lesions
MRI Flow voids, Insensitive for • Localization of fistulous tear with direct
cavernous and moderate- to
petrosal sinuses low-flow fistulas lesions
Distension of the • Evaluation of degree of arteriovenous
cavernous sinus shunting
Dilated • Analysis of venous drainage pathways
intercavernous • Identification of high-risk features,
channels
including cortical venous reflux
Sequelae of orbito-
ocular congestion
TOF Flow-related Not 100 % specific
MRA enhancement in No temporal Huber’s Maneuver: In the setting of a high-
cavernous sinus resolution flow CCF, the tremendous arteriovenous shunting
CE Flow-related Improved temporal may obscure the underlying ICA tear, and the
MRA enhancement in resolution carotid artery more distally may not fill due to
cavernous sinus May not detect essentially complete diversion of blood flow into
high-risk features
the cavernous sinus [2, 10, 22]. In these instances,
CTA Distension of the Poor temporal
cavernous sinus resolution
injection of a vertebral artery during manual com-
Early sinus May not detect pression of the ipsilateral ICA may allow for
enhancement in high-risk features better characterization of both the location and
arterial phase the size of the fistulous communication (Fig. 9)
Dilatation of [10, 44]. This technique works by allowing a
draining venous
limited amount of contrast to reach the shunt via
tributaries
a posterior communicating artery with subsequent
30 Carotid Cavernous Fistula 669

Fig. 9 Huber maneuver.

(a) Frontal projection left
internal carotid artery
angiogram demonstrates a
direct CCF with the high-
flow arteriovenous shunting
obscuring the underlying
tear in the carotid artery. (b)
Injection of the contralateral
ICA or (c) vertebral artery
during manual or balloon
compression of the
ipsilateral ICA (Huber’s
maneuver) may allow for
better characterization of
both the location and the
size of the fistulous
communication. (d) C-arm
angiographic CT (flat panel
rotational computed
tomography) is also a
exceptional technique for
locating and studying the
anatomy of fistula as well

Fig. 10 Angiographic balloon occlusion test. (a) Frontal
projection left internal carotid artery angiogram demon-
strates high-flow arteriovenous shunting associated with a
left direct CCF. There is relatively poor opacification of the
more distal left internal carotid artery due to vascular steal.
(b) Frontal projection right internal carotid artery angio-
gram opacifies the left anterior cerebral artery, but there is
no significant filling of the left middle cerebral artery. (c)
Frontal projection right internal carotid artery angiogram
during balloon occlusion of the cervical left internal carotid
artery (black arrow) demonstrates brisk collateral flow to
the left middle cerebral artery territory (white arrows)

retrograde flow down the ipsilateral supraclinoid
ICA [10]. Other maneuvers that can help delineate
the fistula size and location include very high
frame rate imaging and 3D subtraction angiogra-
phy. Finally, C-arm angiographic CT (flat panel
rotational computed tomography) is another
exceptional technique for locating and studying
the anatomy of fistula.
Balloon Occlusion Test (BOT): As a significant
minority of direct CCFs may require ipsilateral
ICA sacrifice for successful closure, performing
a balloon occlusion test during the diagnostic
workup can provide invaluable information to
the treating physician (Fig. 10). First, a guide
catheter is placed in the ipsilateral common
carotid artery. Next, following full heparinization,
670
a soft, compliant balloon (typically a HyperForm
or HyperGlide Occlusion Balloon, ev3, Irvine,
California) is navigated into the cervical ICA
proximal to the fistula. The balloon is then care-
fully inflated under fluoroscopic imaging, and a
gentle injection is performed through the guide
catheter to confirm vessel occlusion. Continuous
neurologic monitoring, including testing of
patient speech and contralateral strength, is then
performed for a total of 30 min. The test is imme-
diately stopped if the patient develops symptoms
suggestive of ischemia involving the ipsilateral
cerebral hemisphere. The patient can be further
challenged during a balloon occlusion test by
purposefully dropping the systolic blood pressure
by 30 %, although all operators do not routinely
perform this maneuver. Finally, collateral flow to
the affected vascular territory via the Circle of
Willis (i.e., an angiographic balloon occlusion
test) can be performed during balloon inflation
by injection of the contralateral carotid and verte-
bral arteries using a second diagnostic catheter.
Management
The management of CCFs is often determined by
multiple interdependent factors, including the rate
of arteriovenous shunting, the venous pathways
recruited by the fistula, the degree of associated
orbito-ocular congestion, as well as the presence
of cortical venous reflux or venous varix
[15, 45]. Treatment options include conservative
management with close imaging follow-up as
well as endovascular or surgical repair. Indica-
tions for aggressive fistula treatment include pro-
gressive vision loss, cranial nerve palsies, rapidly
worsening proptosis with corneal exposure, corti-
cal venous reflux, and intractable retro-orbital
pain [14, 19]. It is important to note that these
signs and symptoms may be reversible only when
treatment is initiated early in the course of the
disease, emphasizing the importance of prompt
diagnosis and prompt treatment [1, 27, 46]. Liang
et al. [16] found in their retrospective review of
post-traumatic direct CCFs that the time from first
symptom onset to treatment was significantly
shorter in patients who had no fistula-related
T.R. Miller et al.
disability on follow-up compared to those who
did (41.5 days compared to 140.8 days). Finally,
although complete fistula cure is often attempted,
partial occlusion/embolization may be adequate if
it addresses either intolerable symptomatology or
high-risk features such as cortical venous reflux.
The following sections further explore the var-
ious management options available to patients
with CCFs.
Conservative Management
When discussing possible treatment of a CCF
with a patient, the potential risks and benefits of
therapy should be weighed against the natural
history of the patient’s particular lesion [18, 26].
Spontaneous closure of indirect CCF has been
reported anywhere between 10 % and 73 % of
cases, although spontaneous closure of direct fis-
tulas is thought to be considerably more rare
[9, 14, 18, 19, 26]. Interestingly, spontaneous
fistula closure may be precipitated by catheter
angiography [18, 19, 26]. Accordingly, low-flow
fistulas without significant orbito-ocular conges-
tion, high-risk angiographic features, or other
intolerable symptoms can often be followed con-
servatively [1, 9, 15]. However, close clinical as
well as periodic imaging follow-up is required
though to exclude the development of aggressive
fistula features [1]. Stiebel-Kalish et al. [9] found
in a retrospective review that clinical signs sug-
gestive of the presence of cortical venous reflux in
indirect CCF include bilateral orbital symptoms as
well as the presence of a postauricular bruit.
Patients with these features should undergo cath-
eter angiography for further evaluation.
During this time, patients may help to promote
fistula closure by manually compressing the ipsi-
lateral cervical carotid artery and internal jugular
vein several times a day (Fig. 11)
[47, 48]. Higashida et al. [48] found that this
resulted in cure of indirect CCF in 30 % of
patients anywhere from several minutes to
6 months. This maneuver can even result in lesion
closure in a minority of direct CCF (17 %). Kai
et al. [49] found that factors associated with a
higher likelihood of achieving fistula closure
30 Carotid Cavernous Fistula
Fig. 11 Manual compression for treatment of a low-flow
indirect type D CCF. (a) Frontal and (b) lateral projections
of a right internal carotid and as well as frontal (c, d) lateral
projections of a right external carotid artery angiogram
demonstrating arterial supply to an indirect CCF from
external and internal carotid meningeal branches (black
with manual compression for indirect lesions
included a shorter time between symptom onset
and initiation of treatment, lower ocular pressure,
and venous drainage exclusively via the superior
ophthalmic vein. Patients should be instructed to
be sitting during the maneuver and to use the
contralateral arm [49]. Otherwise, transient lack
of blood flow to the ipsilateral cerebral hemi-
sphere may lead to syncope as well as inadvertent
prolonged compression if the ipsilateral arm is
used (which would not be affected by the resulting
ischemia).
Carotid Manual Compression of Carotid
Cavernous Fistulas
• Appropriate for low-flow lesions without
high-risk features or progressive
symptoms.
• Patient should be sitting in case of
syncope.
• Patients should perform maneuver sev-
eral times a day.
• Patient should be instructed to use con-
tralateral hand in case of ipsilateral cere-
bral ischemia.
671
arrows). Patient was managed conservatively and asked
to perform intermittent manual carotid artery compres-
sions. Follow-up right internal (e, f) and external carotid
(g, h) artery angiogram performed 6 months later demon-
strates resolution of the fistula
General Considerations
of Endovascular Treatment
A large majority of both direct and indirect CCFs
can be successfully treated using modern
endovascular techniques with a low rate of
procedure-related complications [11, 50]. Emboli-
zation materials currently used for the treatment of
CCFs include fiber and platinum microcoils,
detachable latex balloons, as well as liquid
embolic agents including n-butyl cyanoacrylate
(nBCA) or ethylene vinyl alcohol (Onyx, ev3,
Irvine, California) [50–52]. Transarterial,
transvenous, as well as combined approaches
may be utilized for fistula closure (Fig. 12) [11,
50, 53]. The endovascular method used to treat a
given lesion often is determined by multiple fac-
tors, including the type of shunt (direct
vs. indirect), the size of the ICA tear with direct
fistulas, the rate of arteriovenous shunting, as well
as the accessibility of venous pathways to the
involved cavernous sinus [15].
General risks of endovascular treatment of
CCFs include stoke, hemorrhage, vessel injury,
and parent artery occlusion. In addition, patients
672 T.R. Miller et al.

Fig. 12 Endovascular access routes. (a) Transarterial
microcatheter injection of a distal external carotid artery
branch supplying an indirect CCF. Black arrows demon-
strate the distal aspect of the microcatheter. (b)
Transarterial access of a direct left CCF with a
microcatheter passing from the ipsilateral internal carotid
artery (white arrows) and into the cavernous sinus (black
arrows). (c) Transvenous access of a CCF through the
superior ophthalmic vein (black arrows). (d) Transvenous
access of a CCF through the inferior petrosal sinus (black
arrows) (e) Transvenous access of a CCF with a
microcatheter extending through the intercavernous chan-
nels (white arrows). (f) Direct cavernous sinus access
through percutaneous of the inferior orbital fissure (black
arrows)

should be counseled that their symptoms may patients should be advised that some symptoms,
temporarily, or rarely permanently, worsen fol- particularly cranial nerve palsies, may not
lowing therapy as the fistula thromboses, with improve despite successful fistula closure [46].
alterations of venous drainage [15]. Obtaining a
complete diagnostic catheter angiogram prior to
Endovascular Treatment of Carotid
embolization is essential to define important fis-
Cavernous Fistulas
tula characteristics such as arterial supply and
• Materials
venous drainage, as well as to evaluate for high-
– Detachable Balloons
risk features including cortical venous reflux
– Platinum or fiber coils
[45]. In addition, the interventionalist must pay
– Liquid embolic agents
close attention to alterations of venous drainage
n-Butyl cyanoacrylate (nBCA, glue)
from the cavernous sinus during embolization of a
Ethylene vinyl alcohol copolymer
CCF. Inadvertent diversion of blood into the
(Onyx)
sphenoparietal or superior petrosal sinuses after
– Covered Stents
partial fistula embolization may lead to cortical
• Endovascular Access
venous reflux and hemorrhagic stroke [15, 26, 45,
– Ipsilateral carotid artery for direct
54, 55]. Similarly, redirection of venous drainage
fistula
anteriorly into the ipsilateral orbit can lead to
– Venous routes
worsening of orbito-ocular congestion, with pos-
sible resulting loss of vision [28, 55]. Finally, (continued)
30 Carotid Cavernous Fistula
Inferior petrosal sinus
Facial and angular veins
Pterygoid venous plexus
Superior petrosal sinus
Percutaneous puncture of superior
ophthalmic vein
– Transarterial catheterization of men-
ingeal feeders of indirect fistulas
Endovascular Treatment of Direct CCF
Detachable Balloons: The goal of endovascular
treatment of direct CCF is complete fistula closure
with preservation of the ICA [14, 56]. Early
endovascular approaches to direct CCF consisted
of transarterially navigating a catheter with a
detachable balloon mounted at its tip past the tear
in the cavernous ICA and into the cavernous sinus.
Once in position, one or more of these balloons
were sequentially inflated and detached in the cav-
ernous sinus to achieve fistula closure [12, 22,
33, 53]. Detachable balloons used for CCF treat-
ment are made of latex and inflated with contrast
medium in position [19]. Serbinenko [57] was the
first to use such an approach for the treatment of
CCFs, followed by Debrun et al. [58]. This method
is very successful at achieving fistula closure, with
high rates of ICA preservation and low procedure-
related morbidity and mortality [10, 11, 14, 59].
Lewis et al. [14] demonstrated an 88 % rate of
complete occlusion of direct CCF via the
transarterial balloon embolization method, with a
75 % rate of ICA preservation. The permanent
neurologic complication rate was only 4 %.
Challenges to transarterial balloon emboliza-
tion of direct CCFs include difficulties navigating
the relatively stiff balloon catheters, inability to
cannulate small ICA tears, inability to inflate a
balloon in small-sized venous compartments
with the cavernous sinus, balloon compression
and compromise of the cavernous ICA or cranial
nerves, as well as subsequent balloon deflation or
migration [10, 14, 55, 56, 59, 60]. Balloon defla-
tion or migration could result in fistula recurrence,
the development of an ICA pseudoaneurysm, or
673
compression of the ICA and cranial nerves
[14, 56, 60]. In addition, some patients following
detachable balloon treatment have developed
recanalization of a true cavernous ICA aneurysm,
which was presumably responsible for develop-
ment of the fistula [60]. Transarterial balloon
embolization of direct CCFs has been abandoned
in the USA due to the lack of commercially avail-
able detachable balloons.
Coils and Liquid Embolic Agents: An alterna-
tive endovascular treatment of direct CCFs is the
placement of coils into the involved cavernous
sinus via a transarterial approach [11, 19, 55].
Once the microcatheter is in position in the
sinus, multiple fiber or platinum microcoils can
be carefully deployed in the venous
sinusoids closest to the point of arteriovenous
fistulization [55]. Similar, to detachable balloons,
challenges to coil embolization include difficul-
ties navigating the microcatheter into the cavern-
ous sinus if the ICA tear is small, as well as
inadvertent coil prolapse into the ICA with larger
rents [52, 55]. The latter may be particularly dif-
ficult to visualize in the setting of a high-flow
shunt and can result in parent vessel occlusion
and/or stroke. The risk of coil prolapse into the
ICA may be mitigated by placement of an
endovascular stent and/or non-detachable balloon
in the vessel prior to embolization [13, 52]. Finally,
it is possible to occlude a large majority of the
diseased cavernous sinus with coils and still have
residual arteriovenous shunting if the portion of
the sinus closest to the fistula remains patent [50].
Embolization of a direct CCF may be
supplemented or replaced by liquid embolic
agents such as nBCA or ethylene vinyl alcohol
(Onyx) [56, 61]. The non-adhesive nature of
Onyx offers several advantages over nBCA,
including the reduced risk of microcatheter reten-
tion, which allows for longer injection times and
more thorough penetration of the liquid embolic
agent into the lesion [50]. Potential risks of liquid
embolic embolization of direct CCF include non-
target embolization of the ipsilateral ICA and its
branches as well as cranial neuropathy. The risk of
reflux of liquid embolic agent may be mitigated by
inflation of a balloon into the ipsilateral cavernous
ICA during liquid embolic injection [62].
674 T.R. Miller et al.

Fig. 13 Carotid artery

sacrifice. (a) Frontal and (b)
lateral projections of a right
internal carotid artery
angiogram demonstrate a
post-traumatic right internal
carotid artery
pseudoaneurysm (white
arrows) and associated
direct CCF. (C) Frontal
projection of a right internal
carotid artery angiogram
following vessel sacrifice
by coil embolization
demonstrates no residual
filling of the fistula. Coil
mass is outlined by black
arrows. (d) Frontal
projection of a left internal
carotid artery angiogram
following right internal
carotid artery sacrifice
demonstrates cross-filling
of the contralateral right
anterior and middle cerebral
arteries

Alternative Routes to the Cavernous Sinus: If a
transarterial approach is either not technically
possible or fails to adequately close a direct
CCF, transvenous treatment may be attempted
[54]. A transvenous approach can be performed
by placement of a guide catheter in the ipsilateral
internal jugular vein followed by navigation of a
microcatheter into the cavernous sinus via the
inferior petrosal sinus [45, 53, 54]. However, it
is not always possible to access the cavernous
sinus via this route due to either thrombosis or
stenosis of the inferior petrosal sinus [15, 53,
63]. In these instances, alternative routes to the
cavernous sinus include the ipsilateral facial and
angular veins, the contralateral pterygoid venous
plexus, as well as the superior petrosal sinus via
the transverse sinus [15, 29, 53, 63].
On occasion, it may only be possible to reach
the involved cavernous sinus from a transvenous
approach via direct surgical cannulation/percuta-
neous puncture of the ipsilateral superior ophthal-
mic vein [27, 28]. This method has been shown to
be both safe and effective, with cannulation of the
vein possible even if it appears thrombosed on
imaging [27]. Once microcatheter access to the
cavernous sinus has been achieved, embolization
can then proceed with coils and/or liquid embolic
agents [61]. Potential complications of direct
puncture of the superior ophthalmic vein include
rupture of the vessel with retroocular hemorrhage
and rapid visual loss, infection, and damage to
other orbital structures [28]. Finally, there are a
few report of direct percutaneous cannulation of
the inferior ophthalmic vein performed for fistula
treatment [64].
Internal Carotid Artery Sacrifice: If other
endovascular methods of closing a direct CCF
fail, ICA sacrifice should be considered (Fig. 13)
30 Carotid Cavernous Fistula
Fig. 14 Covered stent treatment of a direct CCF. (a)
Lateral projection internal carotid artery angiogram dem-
onstrates a high-flow direct CCF (black arrow) with exten-
sive shunting into the superior ophthalmic (white arrow)
and inferior ophthalmic veins (red arrow) as well as the
[56]. The latter technique has proven to be highly
successful in the treatment of direct CCFs,
although there is a risk of ipsilateral cerebral
hemispheric stroke. This risk may be mitigated,
in part, by performing a balloon occlusion test
prior to artery closure. However, a successful
balloon occlusion test does not guarantee a good
outcome. High-flow direct CCFs that demonstrate
complete diversion of blood into the lesion with
non-opacification of the supraclinoid ICA repre-
sent a special circumstance when carotid artery
sacrifice should be considered [60]. If the patient
has not suffered from stroke or TIA-like symp-
toms, the hemodynamics of the fistula provide
strong evidence that carotid occlusion will be
tolerated. On the other hand, as many of these
patients are relatively young, carotid artery sacri-
fice must be weighed against the risk of the patient
developing stenosis or occlusion of another major
artery in the head and neck later in life.
Covered Stents: Covered stents have also been
used to treat direct CCFs, either alone or in com-
bination with coils and liquid embolic agents
(Fig. 14) [51, 65, 66]. Gomez et al. [51] treated
seven patients with post-traumatic CCFs with
675
pterygoid venous plexus (arrow head). (b) Lateral projec-
tion fluoroscopic native image demonstrates a covered
stent (black arrows) placed in the cavernous internal
carotid artery spanning the defect
polytetrafluoroethylene (PTFE)-covered stents,
all of which had successful fistula closure and no
immediate procedure-related morbidity or mortal-
ity. However, one patient subsequently developed
asymptomatic in-stent occlusion [51]. Tiewei
et al. [65] reported similarly good results in a
group of eight patients with traumatic direct
CCF treated with covered coronary stents. Five
of the eight patients had occlusion of their fistula
following stent placement, while six went on to
have resolution of their symptoms. However,
once again, there was one case of subsequent
asymptomatic in-stent occlusion [65]. Another
group, Wang et al. [66], reported successful clo-
sure of 8 out of 10 direct CCF that failed more
conventional endovascular therapy, with no
in-stent stenosis or occlusion during the follow-
up period of the study. These results suggest that
covered stents are a promising new endovascular
treatment of direct CCFs. However, more data is
needed, particularly in regard to the long-term
patency of these devices. Finally, as is the case
with other endovascular stents, patients must be
maintained on dual antiplatelet therapy for at least
3 months following treatment [51]. This is a
676 T.R. Miller et al.

Table 3 Endovascular treatment options for direct carotid significant drawback to treatment with covered
cavernous fistulas stents, particularly in patients who recently expe-
Treatment Potential Potential rienced significant head trauma (Table 3).
modality advantages disadvantages
Detachable Proven, safe Difficulty navigating
balloons and effective stiff delivery catheter
method to fistula Endovascular Treatment of Indirect or
Extensive Inability to inflate Dural CCF
operator balloon in small
experience venous Transvenous embolization of indirect, dural CCF
compartments is a safe and efficacious treatment for patients who
Potential for balloon
have progressive symptoms and high-risks fea-
deflation or
migration following tures on imaging or who fail conservative man-
deployment agement (Fig. 15) [15, 28, 45, 50, 63, 67]. Similar
Coils More controlled Potential for coil to all intracranial dural arteriovenous fistulas, the
deployment prolapse into goal of therapy is closure of the arteriovenous
compared to cavernous ICA,
liquid embolic especially in setting
shunts connecting meningeal arteries to feeding
agents of high-flow shunts veins [45]. For indirect CCF, this has been most
or large rents commonly accomplished by placement of multi-
Small, flexible Potential for residual ple coils in the diseased cavernous sinus [45,
microcatheters fistula following coil 50]. Kirsch et al. [46] reported either complete
may be used for embolization of
coil placement majority of lesion closure or minor residual shunt without
cavernous sinus cortical or ocular drainage in 94 % of 141 patients
Can be expensive if treated using this approach. Meyers et al. [26]
using platinum coils reported similar results with 90 % of patients
Liquid Excellent Less controlled cured following transvenous embolization of indi-
embolic penetration of embolization with
rect CCF in a retrospective review of 135 patients.
agents small venous potential for
compartments nontarget closure of However, transvenous coil embolization of
of the arterial supply to the indirect CCF may not always be successful due
cavernous sinus brain, eye, cranial to the trabeculated structure of the cavernous
nerves
sinus, which can preclude adequate coil place-
Carotid Effective Potential for
artery method ipsilateral ischemic
ment in the venous compartment(s) involved by
sacrifice Relatively safe stroke, even if patient the shunt [45, 50]. In these instances, emboliza-
if fistula results passes BOT tion may be successfully performed using liquid
in complete embolic agents, either alone or in combination
arterial steal or with coils (Fig. 16) [45, 67, 68]. Furthermore, if
patient passes
BOT a transvenous approach to an indirect CCF is not
Covered Shown to be Concern for long- possible due to either venous stenosis or occlu-
stents effective in term patency of sion, a transarterial approach with a liquid
small series stents embolic agent may also be utilized [50]. Onyx,
Option if more Need for antiplatelet with its ability to penetrate small arterial feeders
traditional therapy
endovascular
during prolonged injections, is uniquely well
methods fail suited to reach the small dural shunts from an
arterial pedicle [50]. Care must be taken however
30 Carotid Cavernous Fistula
Fig. 15 Coil embolization of an indirect CCF. Frontal
projections of a (a) right internal carotid artery, (b) left
internal carotid artery, and (c) left external carotid artery
angiograms demonstrating meningeal supply (black
arrows) to a type D indirect CCF. The patient subsequently
underwent transvenous coil embolization of the fistula. (d)
as retrograde filling of non-catheterized arterial
pedicles is possible with Onyx, which may lead
to nontarget embolization of the arterial supply to
the brain, eye, or other cranial nerves. Finally,
other treatment options for indirect CCF
include transarterial embolization using a particu-
late agent such as polyvinyl alcohol (PVA)
(Fig. 17) [69]. A major limitation of this
latter approach however is a high rate of fistula
recurrence, which is less likely following emboli-
zation using coils or liquid embolic agents [69]
(Table 4).
677
Unsubtracted right internal carotid artery angiogram dem-
onstrating coil mass in the bilateral cavernous and circular
sinuses (black arrows). Follow-up frontal projections of (e)
right internal carotid artery and (f) left external carotid
artery angiograms demonstrate no residual arteriovenous
shunting (white arrows)
Surgical Treatment
Prior to the advent of modern endovascular tech-
niques, surgical closure of CCFs was the standard
of care [11]. Early surgical treatment for these
lesions consisted primarily of carotid artery sacri-
fice by vessel trapping [10, 11, 70]. Parkinson [71]
subsequently developed a direct surgical
approach to the cavernous sinus via a triangular
space in the lateral sinus wall demarcated by the
III and IV cranial nerves superiorly and the V and
VI cranial nerves inferiorly. With the aid of
678
Fig. 16 Percutaneous onyx injection of an indirect CCF.
(a) Frontal and (b) lateral projections of a left internal
carotid artery angiogram demonstrate a type B CCF sup-
plied by left internal carotid artery meningeal branches.
Due to difficult access by transarterial or transvenous
approaches, the lesion was accessed by direct transorbital
puncture of the left cavernous sinus following by instilla-
tion of onyx. Lateral projections of a (c) subtraction road
circulatory arrest and induced hypothermia, he
went on to use this approach to treat both direct
and indirect CCF, often preserving the carotid
artery [72]. Modern surgical treatment of CCFs
consists of a combined extradural-intradural
approach to the cavernous sinus, with disconnec-
tion of arterial feeders in the case of indirect
fistulas and clipping or suture repair of fistulous
rents in the case of direct lesions [73, 74]. In
addition, both direct and indirect fistulas may
be indirectly occluded by packing the cavernous
sinus with various materials, including muscle,
glue, thrombus, and wires [18, 73, 74]. Due to the
T.R. Miller et al.
map and (f) native fluoroscopy demonstrate a needle cours-
ing through the inferior orbital fissure (white arrows c and
d) and onyx cast within the cavernous sinus (black arrows
c, d). A balloon was inflated in the left internal carotid
artery to prevent nontarget onyx embolization (red arrows
c). Follow-up (e) frontal and (f) lateral left internal carotid
artery angiograms demonstrates closure of the fistula
associated morbidity and technical challenges of
surgical repair of CCFs, as well as the success of
endovascular approaches, these procedures are
most often performed only after failure of
endovascular treatment [64].
Stereotactic Radiotherapy
Several small case series have demonstrated the
efficacy of stereotactic radiosurgery for the treat-
ment of indirect, low-flow CCFs without high-
risk features [75–77]. Successful closure of the
30 Carotid Cavernous Fistula
Fig. 17 Particulate embolization of a type C indirect CCF.
(a) Frontal and (b) lateral projections of a right external
carotid artery angiogram demonstrating dural shunts
between external carotid meningeal branches and the cav-
ernous sinus (white arrows). Microcatheter injections of
arteriovenous shunt has been reported between
12 and 36 months following treatment [1].
Onizuka et al. [76] reported their results of a
small case series consisting of four elderly
women with symptomatic indirect CCF who
received stereotactic radiosurgery targeted to the
compartment of the involved cavernous sinus
with a marginal dose of 13–15 Gy and a maximum
dose of 36–30 Gy. All four fistulas were success-
fully closed, and patients experienced symptom
relief in 1–3 months. There were no instances of
lesion recurrence or adverse events from treat-
ment during the follow-up period [76]. Pollock
et al. [77] reported similarly high rates of symp-
tom improvement and fistula closure in 20 patients
with indirect CCF (95 % and 93 %, respectively).
679
two right external carotid artery pedicles (c, d) again dem-
onstrate supply to the fistula, which were subsequently
embolized with polyvinyl alcohol particles (PVA). (e)
Frontal and (f) lateral projections of a right external carotid
artery angiogram demonstrates occlusion of the fistula
However, stereotactic radiotherapy was
supplemented by transarterial particulate emboli-
zation in 13 of these patients [77].
Conclusion
Carotid cavernous fistulas are complex lesions
that can be challenging to diagnose and manage.
A multidisciplinary approach, including
neurointerventionalists, neurosurgeons, and oph-
thalmologists, is often required to achieve the best
outcomes for patients. The benefits as well as the
risks of treatment need to be carefully weighed
against the natural history of these lesions,
680 T.R. Miller et al.

Table 4 Endovascular treatment options for indirect 3. Kurata A, Takano M, Tokiwa K et al (1993) Spontaneous
carotid cavernous fistulas carotid cavernous fistula presenting only with cranial
nerve palsies. AJNR Am J Neuroradiol 14:1097–1101
Treatment Potential Potential
4. Stiebel-Kalish H, Setton A, Nimii Y et al (2002) Cav-
modality advantages disadvantages
ernous sinus dural arteriovenous malformations: pat-
Transvenous Proven, safe and Trabeculated terns of venous drainage are related to clinical signs
coiling effective method cavernous sinus and symptoms. Ophthalmology 109:1685–1691
may preclude 5. Barrow DL, Spector RH, Braun IF et al (1985) Classi-
fistula closure fication and treatment of spontaneous carotid-
More controlled May not always cavernous sinus fistulas. J Neurosurg 62:248–256
deployment be possible due to 6. Harris FS, Rhoton AL (1976) Anatomy of the cavern-
compared to lack of venous ous sinus. A microsurgical study. J Neurosurg
liquid embolic access to 45:169–180
agents cavernous sinus 7. Hashimoto M, Yokota A, Yamada H et al (2000)
Transvenous Excellent Potential for Development of the cavernous sinus in the fetal period:
liquid penetration of nontarget closure a morphological study. Neurol Med Chir 40:140–150
embolic small venous of arterial supply 8. Ellis JA, Goldstein H, Connolly ES Jr et al (2012)
embolization compartments of to the brain, eye, Carotid-cavernous fistulas. Neurosurg Focus 32:E9
the cavernous cranial nerves 9. Stiebel-Kalish H, Setton A, Berenstein A et al (2002)
sinus Bilateral orbital signs predict cortical venous drainage
Can be used in May not always in cavernous sinus dural AVMs. Neurology
combination of be possible due to 58:1521–1524
coils lack of venous 10. Debrun G, Lacour P, Vinuela F et al (1981) Treatment
access to of 54 traumatic carotid-cavernous fistulas. J Neurosurg
cavernous sinus 55:678–692
11. Barry RC, Wilkinson M, Ahmed RM et al (2011) Inter-
Transarterial Excellent Potential for
ventional treatment of carotid cavernous fistula. J Clin
liquid penetration of nontarget closure
Neurosci Off J Neurosurg Soc Australas 18:1072–1079
embolic small venous of arterial supply
12. Miller NR (2007) Diagnosis and management of dural
embolization compartments of to the brain, eye,
carotid-cavernous sinus fistulas. Neurosurg Focus
the cavernous cranial nerves due
23:E13
sinus to either reflux of
13. Gemmete JJ, Ansari SA, Gandhi D (2009) Endovascular
Option when agent or
treatment of carotid cavernous fistulas. Neuroimaging
venous access to retrograde filling
Clin N Am 19:241–255, Table of Contents
cavernous sinus of
14. Lewis AI, Tomsick TA, Tew JM Jr (1995) Management
not possible non-catheterized
of 100 consecutive direct carotid-cavernous fistulas:
arterial pedicles
results of treatment with detachable balloons. Neuro-
Particulate Safe method High rate of surgery 36:239–244; discussion 244–235
embolization fistula recurrence 15. Annesley-Williams DJ, Goddard AJ, Brennan RP
et al (2001) Endovascular approach to treatment of
indirect carotico-cavernous fistulae. Br J Neurosurg
particularly low-flow, indirect fistulas. Finally, 15:228–233
16. Liang W, Xiaofeng Y, Weiguo L et al (2007) Traumatic
identification of high-risk features on catheter carotid cavernous fistula accompanying basilar skull
angiography, including cortical venous reflux fracture: a study on the incidence of traumatic carotid
and cavernous sinus venous varix, is essential to cavernous fistula in the patients with basilar skull fracture
appropriately triage these patients. and the prognostic analysis about traumatic carotid cav-
ernous fistula. J Trauma 63:1014–1020; discussion 1020
17. Helmke K, Kruger O, Laas R (1994) The direct carotid
cavernous fistula: a clinical, pathoanatomical, and
References physical study. Acta Neurochir 127:1–5
18. Halbach VV, Hieshima GB, Higashida RT et al (1987)
Carotid cavernous fistulae: indications for urgent treat-
1. De Keizer R (2003) Carotid-cavernous and orbital ment. AJR Am J Roentgenol 149:587–593
arteriovenous fistulas: ocular features, diagnostic and 19. Kobayashi N, Miyachi S, Negoro M et al (2003)
hemodynamic considerations in relation to visual Endovascular treatment strategy for direct carotid-
impairment and morbidity. Orbit 22:121–142 cavernous fistulas resulting from rupture of
2. Peeters FL, Kroger R (1979) Dural and direct cavern- intracavernous carotid aneurysms. AJNR Am J
ous sinus fistulas. AJR Am J Roentgenol 132:599–606 Neuroradiol 24:1789–1796
30 Carotid Cavernous Fistula
20. Gossman MD, Berlin AJ, Weinstein MA et al (1993)
Spontaneous direct carotid-cavernous fistula in child-
hood. Ophthal Plast Reconstr Surg 9:62–65
21. Kupersmith MJ, Stiebel-Kalish H, Huna-Baron R
et al (2002) Cavernous carotid aneurysms rarely
cause subarachnoid hemorrhage or major neurologic
morbidity. J Stroke Cerebrovasc Dis Off J Natl Stroke
Assoc 11:9–14
22. Miller TR, Jindal G, Mohan S et al (2014) Diagnosis
and management of trigemino-cavernous fistulas: case
reports and review of the literature. J Neurointerv Surg
45:674–687
23. Hieshima GB, Cahan LD, Mehringer CM et al (1986)
Spontaneous arteriovenous fistulas of cerebral vessels
in association with fibromuscular dysplasia. Neurosur-
gery 18:454–458
24. Rios-Montenegro EN, Behrens MM, Hoyt WF (1972)
Pseudoxanthoma elasticum. Association with bilateral
carotid rete mirabile and unilateral carotid-cavernous
sinus fistula. Arch Neurol 26:151–155
25. Schievink WI, Piepgras DG, Earnest FT et al (1991)
Spontaneous carotid-cavernous fistulae in Ehlers-
Danlos syndrome Type IV. Case report. J Neurosurg
74:991–998
26. Meyers PM, Halbach VV, Dowd CF et al (2002) Dural
carotid cavernous fistula: definitive endovascular man-
agement and long-term follow-up. Am J Ophthalmol
134:85–92
27. Goldberg RA, Goldey SH, Duckwiler G et al (1996)
Management of cavernous sinus-dural fistulas. Indica-
tions and techniques for primary embolization via the
superior ophthalmic vein. Arch Ophthalmol
114:707–714
28. Quinones D, Duckwiler G, Gobin PY et al (1997)
Embolization of dural cavernous fistulas via superior
ophthalmic vein approach. AJNR Am J Neuroradiol
18:921–928
29. Klisch J, Huppertz HJ, Spetzger U et al (2003)
Transvenous treatment of carotid cavernous and dural
arteriovenous fistulae: results for 31 patients and
review of the literature. Neurosurgery 53:836–856;
discussion 856–837
30. Acierno MD, Trobe JD, Cornblath WT et al (1995)
Painful oculomotor palsy caused by posterior-draining
dural carotid cavernous fistulas. Arch Ophthalmol
113:1045–1049
31. Bussiere M, Lownie SP, Pelz DM et al (2009) Direct
carotid-cavernous fistula causing brainstem venous
congestion. J Neuro-Ophthalmol Off J N Am Neuro-
Ophthalmol Soc 29:21–25
32. Turner DM, Vangilder JC, Mojtahedi S et al (1983)
Spontaneous intracerebral hematoma in carotid-
cavernous fistula. Report of three cases. J Neurosurg
59:680–686
33. Cognard C, Januel AC, Silva NA Jr et al (2008)
Endovascular treatment of intracranial dural arteriove-
nous fistulas with cortical venous drainage: new man-
agement using Onyx. AJNR Am J Neuroradiol
29:235–241
681
34. Akiba H, Tamakawa M, Hyodoh H et al (2008) Assess-
ment of dural arteriovenous fistulas of the cavernous
sinuses on 3D dynamic MR angiography. AJNR Am J
Neuroradiol 29:1652–1657
35. Chen JC, Tsuruda JS, Halbach VV (1992) Suspected
dural arteriovenous fistula: results with screening MR
angiography in seven patients. Radiology
183:265–271
36. Flaharty PM, Lieb WE, Sergott RC et al (1991) Color
Doppler imaging. A new noninvasive technique to
diagnose and monitor carotid cavernous sinus fistulas.
Arch Ophthalmol 109:522–526
37. Chen CC, Chang PC, Shy CG et al (2005) CT angiog-
raphy and MR angiography in the evaluation of carotid
cavernous sinus fistula prior to embolization: a com-
parison of techniques. AJNR Am J Neuroradiol
26:2349–2356
38. Elster AD, Chen MY, Richardson DN et al (1991)
Dilated intercavernous sinuses: an MR sign of
carotid-cavernous and carotid-dural fistulas. AJNR
Am J Neuroradiol 12:641–645
39. Hirabuki N, Miura T, Mitomo M et al (1988) MR
imaging of dural arteriovenous malformations with
ocular signs. Neuroradiology 30:390–394
40. Meckel S, Maier M, Ruiz DS et al (2007) MR angiog-
raphy of dural arteriovenous fistulas: diagnosis and
follow-up after treatment using a time-resolved 3D
contrast-enhanced technique. AJNR Am J Neuroradiol
28:877–884
41. Ouanounou S, Tomsick TA, Heitsman C et al (1999)
Cavernous sinus and inferior petrosal sinus flow signal
on three-dimensional time-of-flight MR angiography.
AJNR Am J Neuroradiol 20:1476–1481
42. Coskun O, Hamon M, Catroux G et al (2000) Carotid-
cavernous fistulas: diagnosis with spiral CT angiogra-
phy. AJNR Am J Neuroradiol 21:712–716
43. Ohtsuka K, Hashimoto M (1999) The results of serial
dynamic enhanced computed tomography in patients
with carotid-cavernous sinus fistulas. Jpn J Ophthalmol
43:559–564
44. Huber P (1976) A technical contribution of the exact
angiographic localization of carotid cavernous fistulas.
Neuroradiology 10:239–241
45. Wakhloo AK, Perlow A, Linfante I et al (2005)
Transvenous n-butyl-cyanoacrylate infusion for com-
plex dural carotid cavernous fistulas: technical consid-
erations and clinical outcome. AJNR Am J Neuroradiol
26:1888–1897
46. Kirsch M, Henkes H, Liebig T et al (2006)
Endovascular management of dural carotid-cavernous
sinus fistulas in 141 patients. Neuroradiology
48:486–490
47. Halbach VV, Higashida RT, Hieshima GB et al (1987)
Dural fistulas involving the cavernous sinus: results of
treatment in 30 patients. Radiology 163:437–442
48. Higashida RT, Hieshima GB, Halbach VV et al (1986)
Closure of carotid cavernous sinus fistulae by external
compression of the carotid artery and jugular vein. Acta
Radiol Suppl 369:580–583
682
49. Kai Y, Hamada J, Morioka M et al (2007) Treatment of
cavernous sinus dural arteriovenous fistulae by exter-
nal manual carotid compression. Neurosurgery
60:253–257; discussion 257–258
50. Gandhi D, Ansari SA, Cornblath WT (2009) Success-
ful transarterial embolization of a Barrow type D dural
carotid-cavernous fistula with ethylene vinyl alcohol
copolymer (Onyx). J Neuro-Ophthalmol Off J N Am
Neuro-Ophthalmol Soc 29:9–12
51. Gomez F, Escobar W, Gomez AM et al (2007) Treat-
ment of carotid cavernous fistulas using covered stents:
midterm results in seven patients. AJNR Am J
Neuroradiol 28:1762–1768
52. Moron FE, Klucznik RP, Mawad ME et al (2005)
Endovascular treatment of high-flow carotid cavernous
fistulas by stent-assisted coil placement. AJNR Am J
Neuroradiol 26:1399–1404
53. Biondi A, Milea D, Cognard C et al (2003) Cavernous
sinus dural fistulae treated by transvenous approach
through the facial vein: report of seven cases and
review of the literature. AJNR Am J Neuroradiol
24:1240–1246
54. Halbach VV, Higashida RT, Hieshima GB
et al (1988) Transvenous embolization of direct
carotid cavernous fistulas. AJNR Am J Neuroradiol
9:741–747
55. Siniluoto T, Seppanen S, Kuurne T et al (1997)
Transarterial embolization of a direct carotid cavernous
fistula with Guglielmi detachable coils. AJNR Am J
Neuroradiol 18:519–523
56. Luo CB, Teng MM, Chang FC et al (2006)
Transarterial balloon-assisted n-butyl-2-cyanoacrylate
embolization of direct carotid cavernous fistulas.
AJNR Am J Neuroradiol 27:1535–1540
57. Serbinenko FA (1974) Balloon catheterization and
occlusion of major cerebral vessels. J Neurosurg
41:125–145
58. Debrun G, Lacour P, Caron JP et al (1978) Detachable
balloon and calibrated-leak balloon techniques in the
treatment of cerebral vascular lesions. J Neurosurg
49:635–649
59. Norman D, Newton TH, Edwards MS et al (1983)
Carotid-cavernous fistula: closure with detachable sil-
icone balloons. Radiology 149:149–157
60. Marques MC, Caldas JG, Nalli DR et al (2010) Follow-
up of endovascular treatment of direct carotid-
cavernous fistulas. Neuroradiology 52:1127–1133
61. Goto K, Hieshima GB, Higashida RT et al (1986)
Treatment of direct carotid cavernous sinus fistulae.
Various therapeutic approaches and results in
148 cases. Acta Radiol Suppl 369:576–579
62. Zhao P, Liu L, Jiang C et al (2012) Coils and onyx
embolization of traumatic carotid-cavernous fistula
caused by an intradural internal carotid artery
pseudoaneurysm. Neuroradiol J 25:231–236
63. Mounayer C, Piotin M, Spelle L et al (2002) Superior
petrosal sinus catheterization for transvenous
T.R. Miller et al.
embolization of a dural carotid cavernous sinus fistula.
AJNR Am J Neuroradiol 23:1153–1155
64. White JB, Layton KF, Evans AJ et al (2007)
Transorbital puncture for the treatment of cavernous
sinus dural arteriovenous fistulas. AJNR Am J
Neuroradiol 28:1415–1417
65. Tiewei Q, Ali A, Shaolei G et al (2010) Carotid cav-
ernous fistulas treated by endovascular covered stent
grafts with follow-up results. Br J Neurosurg
24:435–440
66. Wang C, Xie X, You C et al (2009) Placement of
covered stents for the treatment of direct carotid cav-
ernous fistulas. AJNR Am J Neuroradiol
30:1342–1346
67. Arat A, Cekirge S, Saatci I et al (2004) Transvenous
injection of Onyx for casting of the cavernous sinus for
the treatment of a carotid-cavernous fistula. Neurora-
diology 46:1012–1015
68. Suzuki S, Lee DW, Jahan R et al (2006) Transvenous
treatment of spontaneous dural carotid-cavernous fis-
tulas using a combination of detachable coils and
Onyx. AJNR Am J Neuroradiol 27:1346–1349
69. Vinuela F, Fox AJ, Debrun GM et al (1984) Spontane-
ous carotid-cavernous fistulas: clinical, radiological,
and therapeutic considerations. Experience with
20 cases. J Neurosurg 60:976–984
70. Vitek JJ, Smith MJ (2009) The myth of the Brooks
method of embolization: a brief history of the
endovascular treatment of carotid-cavernous sinus fis-
tula. J Neurointerv Surg 1:108–111
71. Parkinson D (1967) Transcavernous repair of carotid
cavernous fistula. Case report. J Neurosurg 26:420–424
72. Parkinson D (1973) Carotid cavernous fistula: direct
repair with preservation of the carotid artery. Technical
note. J Neurosurg 38:99–106
73. Day JD, Fukushima T (1997) Direct microsurgery of
dural arteriovenous malformation type carotid-
cavernous sinus fistulas: indications, technique, and
results. Neurosurgery 41:1119–1124; discussion
1124–1116
74. Tu YK, Liu HM, Hu SC (1997) Direct surgery of
carotid cavernous fistulae and dural arteriovenous
malformations of the cavernous sinus. Neurosurgery
41:798–805; discussion 805–796
75. Barcia-Salorio JL, Soler F, Barcia JA et al (1994) Ste-
reotactic radiosurgery for the treatment of low-flow
carotid-cavernous fistulae: results in a series of
25 cases. Stereotact Funct Neurosurg 63:266–270
76. Onizuka M, Mori K, Takahashi N et al (2003) Gamma
knife surgery for the treatment of spontaneous dural
carotid-cavernous fistulas. Neurol Med Chir
43:477–482; discussion 482–473
77. Pollock BE, Nichols DA, Garrity JA et al (1999) Ste-
reotactic radiosurgery and particulate embolization for
cavernous sinus dural arteriovenous fistulae. Neurosur-
gery 45:459–466; discussion 466–457
 Imaging of the Neurovasculitides
31
Adam J. Davis

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684 Neurovasculitis is a term used to describe a
diverse spectrum of diseases characterized by
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 685
Classification of the Vasculitides Affecting
inflammation of the blood vessels that may
the Central and Peripheral Nervous Systems . . . . . . . . 685 progress to ischemic injury of the central or
Diagnostic Approach to the Neurovasculitides . . . . . 687 peripheral nervous system. The overlapping
Laboratory and Histopathology . . . . . . . . . . . . . . . . . . . . . 690 clinical and radiographic features and the com-
Clinical Signs and Symptoms . . . . . . . . . . . . . . . . . . . . . . . 693
Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 693
plex pathophysiology make it difficult to
quickly or clearly identify neurovasculitis as
Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 694 the etiology of a patient’s symptoms. Nonethe-
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 694
Indirect Imaging Signs: Parenchymal . . . . . . . . . . . . . . . 694 less, simple yet comprehensive categorization
Indirect Imaging Signs: Vessel Lumen . . . . . . . . . . . . . . 700 schemes established by consensus meetings
Direct Imaging Signs: Vessel Wall . . . . . . . . . . . . . . . . . . 705 and based upon the anatomy, pathophysiology,
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 709 and imaging findings help to clarify the situa-
Mimics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 710 tion. For both the radiologist and the treating
Atherosclerosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 710 clinician, a systemic approach considering the
Tumor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 710
Reversible Cerebral Vasoconstrictive Syndrome . . . . 710
triad of organ systems involved, laboratory
results, and vessel size and type affected sim-
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 713
plifies the situation. Vessel size is particularly
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715 important for the radiologist to consider as it
broadly categorizes the neurovasculitides
within the differential diagnosis. Imaging
plays an essential role in the diagnosis of
these diseases and may be divided into those
studies that investigate the vessel lumen, the
vessel wall, and the parenchyma. Demonstra-
tion of the vessel wall or perivascular inflam-
matory change is a specific and compelling
finding, which should be specifically pursued
by the radiologist. All imaging modalities con-
A.J. Davis (*)
tribute to the analysis; conventional catheter
Neuroradiology, Department of Radiology, NYU Langone
Medical Center, New York, NY, USA angiography is no longer the preeminent diag-
e-mail: adam.davis@nyumc.org nostic standard. Despite the uncommon
# Springer Science+Business Media New York 2016 683
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_22
684 A.J. Davis

occurrence of neurovasculitis, it is important to vessel stenosis or occlusion. Endothelial inflam-

be familiar with these diseases. They play a mation promotes intraluminal coagulation and
particularly important role in the differential thrombosis [1]. Perivascular inflammatory
diagnosis of stroke in young and middle-aged changes and edema contribute to the pathologic
adults and typically have devastating clinical picture. Arterial and venous components may be
consequences. If diagnosed early, they may be involved separately or together, and dural sinuses
effectively treated, relieving symptoms and may be affected. Generalized inflammatory pro-
avoiding permanent central and peripheral cesses may produce a secondary encephalitis or
neurologic damage. myelitis. Ischemia of the small vessel vasa
nervorum typically causes mononeuritis multi-
Keywords plex or polyneuropathy.
Neurovasculitides • Neurovasculitis • Cerebral Neurovasculitis has a wide clinical, radio-
vasculitis • Central nervous system vasculitis • graphic, and pathologic spectrum.
Stroke • Cerebral ischemia • Cerebral The overlapping clinical and radiographic fea-
infarction tures as well as the complex pathophysiology often
make it difficult to clearly identify neurovasculitis
List of Abbreviations as the etiology of the patient’s symptoms. The
ACA Anterior cerebral artery central and peripheral inflammatory and ischemic
CA Catheter angiography events associated with vasculitis produce symp-
CT Computed tomography toms that are similar to the more common
CTA Computed tomography angiography noninflammatory occlusive vascular diseases. The
18
FDG- F-fluorodeoxyglucose positron differential diagnosis of neurovasculitis may be
PET emission tomography overwhelming and confusing. The vasculitides
GCA Giant cell arteritis are sometimes confused with noninflammatory pri-
ICA Internal carotid artery mary vasculopathies both clinically and in the med-
MCA Middle cerebral artery ical literature. This includes noninflammatory
MRA Magnetic resonance angiography collagen vascular diseases and reversible vasocon-
MRI Magnetic resonance imaging strictive syndromes. Some authors will only con-
PACNS Primary angiitis of the central ner- sider the neurovasculitides within the context of
vous system autoimmune or autoinflammatory disease, exclud-
PCA Posterior cerebral artery ing vascular inflammation secondary to general-
RCVS Reversible cerebral vasoconstrictive ized disorders such as infection. Other authors
syndrome disregard the more generalized autoimmune and
SLE Systemic lupus erythematosus autoinflammatory diseases that are not primarily
TA Takayasu arteritis related to the vasculature but nonetheless directly
US Ultrasound affect the blood vessels, including systemic lupus
erythematosus (SLE). These disparities lead to
confusion and misunderstanding regarding the
Introduction nomenclature and classification of the neurovas-
culitides. This situation makes it difficult for phy-
Neurovasculitis is a term used to describe a sicians to have a common point of reference and
diverse spectrum of diseases characterized by interferes with the logical and efficient process of
inflammation of the blood vessels that may pro- diagnosis and treatment. Reiterating simple yet
gress to ischemic injury of the central or periph- comprehensive categorization, as outlined by con-
eral nervous system, resulting in a myriad of focal sensus meetings and based upon the anatomy, path-
and generalized neurologic symptoms. The injury ophysiology, and imaging findings helps to clear
is usually secondary to mural changes resulting in the conundrum.
31 Imaging of the Neurovasculitides
The primary vasculitides, inflammatory dis-
ease affecting only the vasculature, are uncom-
mon, even more so for clinicians interested in the
nervous system since only a fraction of these
patients have central or peripheral CNS involve-
ment. The estimated incidence of Churg–Strauss
syndrome is 20 per million, granulomatosis with
polyangiitis (Wegener’s) 10 per million, and
polyarteritis nodosa 0.9 per million people
[2, 77]. Takayasu arteritis (TA) has a worldwide
estimated incidence of 2.6 per million people [3]
although this varies widely by geography;
autopsy studies in Japan report an incidence of
1 per 3,000 population (333 per million) [4],
while studies in England report an incidence of
0.3 per million population [5]. It is apparent that
the frequency of presentation for any given phy-
sician will vary by practice locale. Secondary
vasculitides, typically secondary to connective
tissue diseases including SLE, affect organ
parenchyma in addition to vasculature and are
more common.
Despite the uncommon occurrence of
neurovasculitides, it is important to be familiar
with these diseases. They play a particularly
important role in the differential diagnosis of
stroke in young and middle-aged adults, where
the prevalence of atherosclerotic disease is small
[6, 7]. The neurovasculitides typically have dev-
astating clinical consequences for the afflicted
patient. If diagnosed early, they may be effectively
treated with a myriad of immunosuppressive
agents, relieving symptoms and avoiding perma-
nent damage to the brain, spinal cord, and periph-
eral nerves.
Overview
Classification of the Vasculitides
Affecting the Central and Peripheral
Nervous Systems
The simplest classification systems divide the
neurovasculitides into primary vascular inflam-
matory diseases affecting only the central and
peripheral nervous systems from primary vascular
685
inflammatory diseases that are systemic, affecting
the vasculature throughout the entire body. Next
are the generalized systemic autoimmune and
autoinflammatory diseases that produce vascular
inflammation in addition to a generalized inflam-
matory process affecting the body organs. These
include the connective tissue diseases. These cat-
egories are distinct from those diseases that pro-
duce vasculitis secondarily by proximity
including infection, radiation, drugs, toxins, and
neoplasm. Table 1 organizes the more common
examples of these diseases.
The neurovasculitides are typically approached
as a long list of disease entities, which is inevitable
as the etiologies of neurovasculitis are protean.
Unfortunately, this leaves the clinician with no
practical coherent organizational system to refer
to. A systemic approach to these diseases based
primarily upon the triad of:
1. Clinical presentation particularly the organ
systems involved
2. Laboratory results
3. Vessel size and type afflicted
organizes and simplifies the subject for the radiol-
ogist and clinician alike. Vessel size is particularly
important for the radiologist to consider as it helps
to broadly categorize the neurovasculitides within a
differential diagnosis (Fig. 1).
Discussion of vessel size is problematic as there
are many different classifications based on histo-
logic, pathologic, and clinical definitions, generally
in reference to the entire human body and systemic
disease processes. The cerebrovascular system
skews this paradigm. Division into large elastic
arteries, medium highly muscular arteries, and
small thin muscular arteries/arterioles as is utilized
by the prevailing systemic vasculitis classification
systems [8, 9] fails to adequately describe the dis-
tribution and pathology of neurovasculitides. The
entire cerebrovascular system would coarsely be
divided into large elastic arteries from the aorta
through the common carotid and vertebral arteries
and then the muscular arteries, which correspond to
everything else distally to the capillary level. These
classifications tend to disregard the unique
686 A.J. Davis

Table 1 Common etiologies of the neurovasculitides by which may effect the distribution of disease within
disease category the brain.
I. Nonsystemic primary vasculitis affecting only the In order to better organize the neurovas-
central or peripheral nervous system culitides, a more relevant classification is
Primary angiitis of the central nervous system (PACNS) required. Kuker [10, 25] proposed that the internal
Nonsystemic peripheral nervous system vasculitis
carotid arteries through their intracranial skull
(NSPNSV)
II. Systemic primary vasculitis
base segments including their primary divisions
Large vessel into the anterior cerebral artery (ACA) A1 seg-
Giant cell arteritis ments; the middle cerebral artery (MCA) M1 seg-
Takayasu arteritis ments; the vertebral artery, basilar artery, and
Medium vessel posterior cerebral artery (PCA) P1 segments; and
Polyarteritis nodosa their structural collaterals including the anterior
Kawasaki disease and posterior communicating arteries all be con-
Small-vessel vasculitis sidered to be of the large-vessel type. All cerebral
(More arteriolar) cortex, convexity, and branching arteries and their
Granulomatosis with polyangiitis (Wegener’s) penetrating segments are medium-vessel type,
Churg–Strauss syndrome discernable by catheter angiography (CA) and
Microscopic polyangiitis depending upon the technique by magnetic reso-
(More capillary venule) nance angiography (MRA) and computed tomo-
Henoch–Schönlein purpura graphic angiography (CTA) at the more proximal
Essential cryoglobulinemic vasculitis
second- and third-order branching segments.
Cutaneous leukocytoclastic angiitis
More distally, the small vessels are those not dis-
(Diverse involvement with emphasis on capillaries,
venules, and veins) cernible by imaging, corresponding to the smallest
Behçet’s disease muscular arteries and arterioles within the brain
III. Systemic autoimmune/autoinflammatory diseases parenchyma as well as the capillaries and proximal
affecting organ systems with a vasculitic component venules. Although Kuker did not discuss vascular
Systemic lupus erythematosus dimensions, these small vessels beneath the reso-
Rheumatoid lution of imaging correspond to 200 μm or less [11]
Scleroderma although even CA is severely limited below
Mixed connective tissue disease 500 μm [12]. Using this helpful and relevant clas-
Sjogren’s disease sification system, etiologies tend to cluster into the
IV. Secondary vasculitis following three groups:
Infection
Viral, bacterial, fungal, protozoan, spirochetes
1. Extracranial elastic, extracranial large vessel,
Neoplasm
and intracranial skull base large-vessel
Lymphoma
involvement
Lymphomatoid granulomatosis
2. Intracranial medium-vessel involvement
Lymphocytic vasculitis
Drugs
3. Intracranial small-vessel involvement
Amphetamines, hydralazine, levamisole, cocaine,
synthetic marijuana and other marijuana substitutes, Additionally, some etiologies tend to involve
immunosuppressive drugs for transplant patients the venous system with greater frequency and
(Tacrolimus), interferon sometimes even exclusively, though these small
Radiation venular diseases are difficult if not impossible to
detect with direct vascular imaging. One should
characteristics of brain arteries including a constantly bear in mind that these are categorical
decreased wall thickness to lumen ratio, lack of definitions to aid diagnosis; in reality, there is
external elastic lamina, diminished adventitia, and considerable overlap in vessel size involvement
perivascular subarachnoid fluid environment, all of among the neurovasculitides.
31 Imaging of the Neurovasculitides 687

Fig. 1 Giant cell arteritis. 78-year-old female with severe
worsening headache, diplopia, and jaw pain. MRA at pre-
sentation (a) demonstrates multifocal luminal narrowing
and contour irregularity within the bilateral ICAs (white
arrows). CTA volume rendering (b) reveals multifocal
short-segment luminal narrowing of the superficial
temporal arteries (white arrows). Maximum intensity pro-
jection images of the ICAs (c) demonstrate intraluminal
linear defects (white arrows) consistent with dissection, a
known complication of advanced GCA (Case courtesy of
Adam Davis, MD, New York University Langone Medical
Center, New York)

Table 2 Primary neurovasculitic diseases by distribution of vascular size and type

Aorta, large Medium Small arteries, Capillaries Venules Veins
arteries arteries arterioles
Takayasu arteritis ++ +
Giant cell arteritis ++ +
Kawasaki disease ++ +
PAN ++ +
ANCA + arteritides (CS, + ++ ++ +
MPA, GPA)
Henoch–Schonlein + ++ ++
Behçets + + ++ ++ ++
PACNS ++ ++ ++ ++ +
Based on reference [76]

A comprehensive clinical organization enables Diagnostic Approach to the

both the radiologist and clinician to arrive more
quickly and assuredly at the diagnosis, providing
a clear mandate for biopsy and/or treatment and
resulting in improved outcome for the patient. An
organization of the most common neurovas-
culitides by vessel size is a helpful starting point
for practicing clinicians who may encounter these
patients, typically radiologists, internists, rheuma-
tologists, dermatologists, neurologists, and
neurosurgeons.
Neurovasculitides
The neurovasculitides tend to cluster into groups
characterized by the size and type of vessel
afflicted, an attribute that may be capitalized
upon for organizing these diseases (Table 2). Dis-
eases affecting the large extracranial and skull
base arteries may be differentiated from those
involving the intracranial medium and small arter-
ies and arterioles and, in a similar, fashion from
688
those preferentially affecting capillaries, venules,
and veins. There is sometimes considerable ana-
tomic overlap although clinical and laboratory
characterizations help to further define these
entities.
This approach is most helpful for the radiolo-
gist who typically is able to separate the radio-
graphic findings into broad vascular anatomic
categories, depending on the appearance of the
vessels and the anatomic distribution of the imag-
ing abnormalities. For example, an imaging eval-
uation, including an MRI of the brain
demonstrating multifocal small and punctate
regions of T2 or FLAIR hyperintensity within
the cerebral white matter and irregularity of the
most distal discernible intracranial vessels on
CTA or CA, is more likely to correlate with a
medium and/or small-vessel arteritis than an
imaging demonstrating a large territory infarction
and mural enhancement of the aorta or
brachiocephalic arteries, which likely is second-
ary to a large-vessel arteritis.
The radiologic findings of the neurovas-
culitides may be similar to those of the more
common noninflammatory vasculopathies
including atherosclerosis. Noninflammatory
vasculopathies are associated with central
and peripheral nervous system symptoms
(Table 3), and separating inflammatory from
noninflammatory entities may be difficult, partic-
ularly when imaging characteristics are nearly
indistinguishable. In these cases, clinical criteria
typically help. For example, intracranial athero-
sclerosis may be difficult to separate from a small
artery-type vasculitis such as primary angiitis of
the central nervous system (PACNS), SLE, or
Churg–Strauss syndrome. However, young age
or the absence of large-vessel, skull base, and
calcified mural thickening should cast doubt on
the diagnosis of atherosclerosis. Similarly, dis-
tinctive clinical features such as the presence of
a facial malar rash with elevated renal function
tests or serum eosinophilia with asthma would
steer the diagnosis toward SLE (Fig. 2) or
Churg–Strauss disease (Fig. 3). Alternatively, in
some cases, the imaging characteristics of
neurovasculitis are distinctive and compelling,
allowing for a primary radiographic diagnosis.
A.J. Davis
Table 3 Differential diagnosis of the common neurovas-
culitides and noninflammatory neurovasculopathies orga-
nized by vessel size and type
I. Predominantly large artery extracranial disease
Vasculitis
Systemic with CNS involvement
Takayasu/giant cell arteritis
Atherosclerosis
Arterial dissection
Fibromuscular dysplasia
II. Predominantly medium and small artery intracranial
disease
Vasculitis
1. Systemic with CNS Involvement – PAN and the
ANCA-positive diseases: Churg–Strauss/GPA/
microscopic polyangiitis
2. Autoimmune/autoinflammatory diseases –
connective tissue diseases (SLE, rheumatoid, Sjögren’s)
and Behçets
3. Primary angiitis of the CNS
4. Vasculitis 2 , infection, radiation
Atherosclerosis
Diabetes
Moyamoya
Sickle cell
Neurofibromatosis
Vasospasm – subarachnoid hemorrhage
Reversible cerebral vasoconstrictive syndrome
III. Diseases involving the venules and veins
Vasculitis
Behçets, ANCA-positive diseases,
Henoch–Schonlein, hypersensitivity vasculitis,
lymphomatoid granulomatosis, lymphocytic vasculitis
Hypercoagulable state, neurofibromatosis
IV. No vessel size predilection
Thromboembolic occlusion
Cardio-embolic including septic emboli, arterioarterial
2 to atherosclerotic plaque, hypercoagulable states
As an example, vascular mural thickening with
intramural or perivascular enhancement associ-
ated with a segmental luminal narrowing at extra-
cranial sites would suggest TA or giant cell
arteritis (GCA). In this circumstance, the more
common noninflammatory vasculopathies includ-
ing atherosclerosis or spontaneous dissection sec-
ondary to connective tissue disease would be
unlikely.
Peripheral neuropathy is an area of consider-
able clinical overlap between the inflammatory
31 Imaging of the Neurovasculitides
Fig. 2 Systemic lupus erythematosus. 39-year-old female
with known systemic lupus erythematosus and transient
worsening, bilateral lower extremity weakness, and numb-
ness. MRI FLAIR sequence at presentation (a left) reveals
bilateral hemispheric white matter signal abnormality with
689
involvement of the parietal and occipital gray matter. Right
medial parietal cortical hyperintensity on diffusion-
weighted imaging (a, middle, black arrow) with ADC
low signal intensity (not shown) is consistent with acute
ischemia in the right PCA parieto-occipital division.
690
and noninflammatory causes of vasculopathy.
Many noninflammatory vasculopathies, including
hypertension, atherosclerosis, and diabetes, may
cause extremity sensory neuropathies and cerebro-
vascular abnormalities that can be difficult to dis-
tinguish from vasculitis secondary to polyarteritis
nodosa, granulomatosis with polyangiitis
(Wegener’s), microscopic polyangiitis,
Churg–Strauss, SLE, or a myriad of other inflam-
matory vascular diseases that have a similar pre-
sentation. Again, the presence and pattern of
associated systemic organ involvement in conjunc-
tion with the laboratory results will help to steer the
differential diagnosis in the correct direction.
Laboratory and Histopathology
Familiarity with the basics of the laboratory eval-
uation for vasculitis is helpful to the radiologist.
They are a constant and essential component in
the diagnostic evaluation of these diseases. Perti-
nent laboratory tests should be well regarded by
the radiologist as they help to guide the imaging
differential diagnosis toward the correct
conclusion.
A complete blood count is invariably
performed. Anemia and leukocytosis are frequent
components of the systemic vasculitides. Eosino-
philia may be present, but when excessively ele-
vated, may indicate Churg–Strauss arteritis in the
right clinical setting. A coagulation panel is
Fig. 2 (continued) Gradient echo imaging (a, right, black-
outlined arrow) demonstrates focal susceptibility effect,
indicating hemosiderin deposition from microhemorrhage.
The patient was treated for presumed hypercoagulable
thrombotic complications of SLE and discharged. The
patient presented 3 months later with profound left
hemiparesis and sensory deficits. Diffusion MRI of the
brain (b) reveals restricted diffusion characterized by
increased T2 diffusion and decreased ADC signal intensity
within a large confluent region of the medial right parietal
lobe and occipital lobes, extending to the splenium of the
corpus callosum, consistent with the extension of acute
infarction within the right PCA territory. MRA (c) demon-
strates multifocal luminal narrowing, mostly short segment
and occasionally severe, as well as long-segment stenoses
A.J. Davis
typically performed to evaluate for thrombotic
disorders that may be mistaken for vasculitis.
This typically includes prothrombin time, partial
thromboplastin time, protein C, protein S, and
fibrinogen levels. More sophisticated clotting
tests are pursued if suspicion is high and a hema-
tology consultation is involved.
Acute-phase reactants, usually erythrocyte
sedimentation rate and C-reactive protein, may
be elevated. Serologic/antibody testing typically
includes rheumatoid factor, antinuclear antigen,
antibody to double-stranded DNA, anticardiolipin
antibodies, and cryoglobulins. These tests are
helpful for evaluating the more common systemic
autoinflammatory and autoimmune diseases
including SLE and rheumatoid arthritis. Antinu-
clear cytoplasmic antibodies (p-ANCA and
c-ANCA) are routinely obtained to evaluate for
primary systemic vasculitis including
granulomatosis with polyangiitis (Wegener’s),
Churg–Strauss, and microscopic polyangiitis.
These tests are negative or weakly positive in
polyarteritis nodosa, an important differentiating
factor.
Infectious etiologies may be the primary or
secondary causative factor in neurovasculitis.
Bacterial infection leading to leptomeningitis is a
particularly common etiology that may produce a
secondary vascular inflammation resulting in
ischemia and infarction. Infectious agents may
not only lead directly to inflammatory changes
of the blood vessels but have also been postulated
within the bilateral MCA M1 segments (black arrows).
Similar changes are present within the distal ICAs (not
marked). The right PCA demonstrates multifocal luminal
narrowing and abrupt occlusion of the parieto-occipital
branch (black-outlined arrows). The extensive, large- and
medium-vessel, multisegmental, tandem, luminal irregu-
larity and narrowing are more indicative of a vasculitic
process than a thromboembolic process. CA (d) confirms
the long-segment irregular luminal narrowing with inter-
spersed segments of normal luminal caliber and contour
within all vascular territories. Immunosuppressive therapy
was intensified for treatment of presumed lupus vasculitis.
The patient’s clinical condition stabilized over the follow-
ing several months (Case courtesy of Adam Davis, MD,
New York University Langone Medical Center, New York)
31 Imaging of the Neurovasculitides
Fig. 3 Churg–Strauss disease vs. eosinophilic vasculitis.
45-year-old female with a history of rheumatoid arthritis,
asthma, and colon cancer in situ (PET confirmed not met-
astatic) presents with abdominal pain, nausea, vomiting,
lower extremity pain, and cyanosis. WBC 39,000, eosino-
phil count >15,000, C-reactive protein 110, rheumatoid
factor elevated, and pANCA (positive). MRA of the
extremities (a) demonstrates multifocal brachial, radial,
posterior tibial, and dorsalis pedis artery stenoses and
occlusions. Emergent attempt at thrombolysis was unsuc-
cessful and ischemia progressed to a gangrenous right foot
requiring amputation. Initially assumed to be a multi-
embolic event, no cardiac source was uncovered by
transesophageal echocardiography. MRI of the brain
691
performed for complaints of headache (b, FLAIR right,
ADC left) reveals a punctate right basal ganglia chronic
lacunar infarct (black arrows). CTA of the brain (c) dem-
onstrates multifocal short- and long-segment luminal
narrowing within the ACA (black arrows), MCA (white
arrows), and superficial temporal artery (black arrow-
heads). Intracranial vessels demonstrate marked luminal
contour irregularity, while the extracranial arteries demon-
strate a more long-segment smooth and concentric
narrowing, reflecting the variability of the radiographic
findings of these diseases even in the same individual
(Case courtesy of Adam Davis, MD, New York University
Langone Medical Center, New York)
692
Fig. 4 Lymphocytic vasculitis. 15-year-old male with
X-linked lymphoproliferative disease. Worsening confusion.
MRI FLAIR axial imaging (a, left) demonstrates extensive
confluent abnormal hyperintense signal throughout the bilat-
eral hemispheric white matter as well as more patchy
involvement of the deep nuclei and cortical gray matter.
Diffusion imaging (not shown) failed to demonstrate an
evidence for acute infarction. Post-gadolinium T1-weighted
sagittal images (b, right) demonstrate striking radial linear
enhancement corresponding to the periventricular venous
system (white arrows). Additional regions of linear
to be a predisposing factor in the development of
several vasculitic diseases including PACNS
[1, 13]. Consequently, CSF and serum investiga-
tion for infection is often performed. Examination
of the CSF including routine cell count, glucose,
protein, Gram stain, and serologic testing for
syphilis, HIV, hepatitis B, Epstein–Barr virus,
varicella or other suspected infectious agents is
routinely acquired. Granulomatous entities, par-
ticularly tuberculosis and sarcoidosis, may need to
be excluded as their imaging signs, including
multifocal regions of parenchymal signal abnor-
mality as well as leptomeningeal and parenchy-
mal enhancement, may overlap with vasculitic
processes including granulomatosis with
polyangiitis (Wegener’s) and PACNS.
Conversely, a negative laboratory evaluation is
a very important diagnostic component in the
evaluation of neurovasculitis. Lack of systemic
inflammatory markers is strongly supportive for
the diagnosis of PACNS, although even in these
patients up to two-thirds may have nonspecific
inflammatory serum markers including elevated
erythrocyte sedimentation rate [14].
CSF evaluation is usually nonspecific as an
elevated protein and leukocytosis may be present
in many infectious and inflammatory
A.J. Davis
enhancement within the deep cerebellum and brain stem
were also present (not shown). The patient did not respond
to ganciclovir for suspected Epstein–Barr virus encephalitis;
a common complication of X-linked lymphoproliferative
disease and concern for lymphocytic vasculitis or lymphoma
were raised. Brain biopsy was performed. Histopathology (b,
black arrow) confirms a perivascular polyclonal lymphocytic
infiltrate consistent with vasculitis. Treatment with cortico-
steroids resulted in rapid clinical improvement (Case cour-
tesy of Yvonne Lui, MD, New York University Langone
Medical Center, New York)
encephalopathies. Regardless, it is important to
investigate CNS infection as well as exclude
demyelinating disease as a possible etiology.
Biopsy is widely considered the gold standard
in the evaluation of neurovasculitis. When labora-
tory and imaging tests do not provide a clear diag-
nosis or when the systemic evaluation is normal
and PACNS is suspected, brain biopsy is required.
Typically, sampling of the affected brain paren-
chyma and particularly the adjacent leptomeninges
with a wedge biopsy guided by the imaging find-
ings [15, 16] provides the highest rate of success,
although nearly 15–38 % of biopsies are reportedly
negative [14–18] in studies investigating all causes
of neurovasculitis. Children tend to have a higher
rate of positive brain biopsies than adults [19]. It is
uncertain whether random sampling decreases the
yield; it is generally not recommended but may be
necessary if no focal lesions are detected, if the
lesions identified occupy eloquent regions of the
brain, or if the lesions are located deep within the
spinal cord parenchyma. It is important to consider
that in approximately one-half of positive brain
biopsies performed for the evaluation of PACNS,
an alternative diagnosis is made, including neo-
plasm (Fig. 4), demyelinating disease, and infec-
tion [15, 18].
31 Imaging of the Neurovasculitides
The risk of biopsy is not negligible although it
varies widely within the literature. A mortality
rate of up to 1.5 %, a morbidity of up to 2 %,
and a complication rate of up to 20 % have been
reported, although most complications are tran-
sient [20, 21]. Very few patients suffer permanent
neurologic deficit [15]. This relatively small risk
needs to be weighed against the risk of erroneous
diagnosis and possible inappropriate or unneces-
sary treatment with potentially toxic immunosup-
pressive medications. Marsh et al. [2] warn that in
a series of five patients referred for a treatment of
primary CNS vasculitis, in which four of whom
had already began immunosuppressive treatment,
none were ultimately found to have vasculitis.
With this in mind, it is important to consider that
75 % of cases of neurovasculitis are diagnosed
and managed without biopsy [22]. Alternatively,
when systemic organs are involved and a primary
vasculitis or systemic autoimmune disease is con-
sidered, a biopsy of non-neurologic tissues may
provide a less morbid means to confirm the diag-
nosis. This should be a strong inclination to per-
form neck, chest, and abdomen imaging in
patients with suspected CNS vasculitis.
Clinical Signs and Symptoms
The signs and symptoms of neurovasculitis are
reflections of the many diseases that cause it [20,
23, 24]. It is roughly divided into symptoms
related to central nervous system and peripheral
nervous system involvement, the latter typically
presenting as mononeuritis multiplex or sensory
polyneuropathy. Cranial nerve involvement is a
prominent component of some of the neurovas-
culitides, particularly granulomatosis with
polyangiitis (Wegener’s), and is a distinguishing
feature. The involvement of the central nervous
system results in a more diverse symptomology
and may be generalized or focal. Headache,
encephalopathy, seizure, stroke-like events, cog-
nitive impairment, and movement disorders par-
ticularly chorea and myoclonus are described.
Behavioral abnormalities may occur, particularly
in children. Visual change may result from an
impairment of the optic nerve, tract, radiation, or
693
cortex. It may be acute and fulminant in presenta-
tion or chronic and indolent.
Systemic signs and symptoms frequently
accompany the neurologic dysfunction and may
at first be misleading. Fever, night sweats, weight
loss, weakness, rash, and arthropathy may initiate
an evaluation for infection or cancer. In fact, many
of the imaging studies that the radiologist encoun-
ters that ultimately result in a diagnosis of vascu-
litis will initially be requested for other
indications. Even when signs and symptoms are
isolated to the neurologic system, demyelinating
disease, infectious meningitis, acute disseminated
encephalomyelitis, atherosclerosis, and cardio-
genic or thrombophilic embolic stroke are often
the working diagnoses. When vasculitis is sys-
temic and not suspected, imaging of the brain is
often performed in symptomatic patients to eval-
uate for hemorrhage or infarct. In these patients,
other organ systems are frequently involved and
the neuroradiologist should take particular note
when imaging is performed for the indication of
neurovascular disease or encephalopathy and
there is a concurrent evaluation for suspected pul-
monary, renal, liver, musculoskeletal, or dermato-
logic disease. The same is true when there is an
evaluation for elusive, cryptic, oncologic, or
infectious disease in a patient with profound cen-
tral or peripheral CNS symptoms.
Treatment
The treatment of the neurovasculitides is beyond
the scope of this chapter. Nonetheless, it is impor-
tant for the radiologist to consider that prompt and
confident diagnosis of neurovasculitis and its
underlying etiology is critically important for
these patients. Many of these illnesses may be
effectively treated with immunosuppressive
drugs [20], typically corticosteroids and cyclo-
phosphamide. Alternative agents including aza-
thioprine, methotrexate, thalidomide, or
interferon may be used. Plasmapheresis may be
helpful for certain diseases including
cryoglobulinemia. Early treatment may induce
long-term remission or cure, while delayed treat-
ment may result in permanent vascular and
694
neurologic damage. As these drugs may have
serious and debilitating side effects, increased
confidence in the diagnosis of vasculitis via imag-
ing provides a clear mandate to the referrer for
their use.
Imaging
Introduction
Imaging plays an important if not essential role in
the diagnosis of neurovasculitis. Nearly all
patients suspected of harboring an inflammatory
vasculopathy with neurologic signs and symp-
toms undergo imaging of the brain and/or spinal
cord as well as dedicated imaging of the
cervicocerebral vasculature. Unfortunately,
parenchymal imaging provides little, if any,
pathognomic findings of neurovasculitis. The
presence of ischemia, edema, intraparenchymal
and subarachnoid hemorrhage, and abnormal
parenchymal and leptomeningeal enhancement
may be present in many different cerebral vaso-
occlusive, infectious, inflammatory, demyelinat-
ing, and neoplastic conditions. Additionally, the
radiographic findings and the specificity and sen-
sitivity of neuroimaging are variable and depen-
dent upon the specific disease entity causing the
neurovasculitis. PACNS has a vastly different
appearance than large-vessel primary vasculitis
such as GCA or a secondary vasculitis from infec-
tion. Nonetheless, the constellation of imaging
features may lead the radiologist to the correct
diagnosis or at least a focused differential diagno-
sis within the given clinical context.
Kuker [13, 25] divided the imaging findings of
neurovasculitis into indirect and direct signs, the
former related to ischemia, hemorrhage, and vas-
cular stenoses thought to be unlikely secondary to
hypertensive or atherosclerotic disease. Direct
signs include findings related to vessel wall
inflammatory changes including mural thickening
and enhancement. The radiologist should bear in
mind that regardless of the imaging modality
employed (CT, CTA, MRI, MRA, CA, US,
nuclear medicine), the imaging findings will
have varying importance and specificity when
A.J. Davis
evaluating neurovasculitis. Gomes [26] conve-
niently divides the imaging into three groups:
studies that investigate the vessel lumen, the ves-
sel wall, and the parenchyma. Parenchymal find-
ings are the least specific but are usually necessary
to detect the presence of disease as well as follow
progression or remission. Vessel lumen abnormal-
ities are highly suggestive for vasculitis when
present but are not specific and the sensitivity is
low for the small-vessel vasculitides. Demonstra-
tion of vessel wall or perivascular inflammatory
changes are a more specific and compelling find-
ing, although it is only inconsistently found, likely
due to the limitations of imaging technique. MRI
and nuclear medicine imaging provide the most
sensitive evaluation for vessel wall inflammation
although extracranial large elastic vessel inflam-
mation is seen particularly well with CT.
Indirect Imaging Signs: Parenchymal
CT of the brain or spine is often employed as the
first imaging examination in patients presenting
with acute or progressively worsening symptoms.
It provides a rapid evaluation for the etiology of
neurologic emergencies including hemorrhage,
infarction, or mass. Other than hemorrhage,
CT is both relatively nonspecific and insensitive
to the findings that are most indicative of
neurovasculitis.
MRI of the brain or spinal cord is typically
employed as the next examination, although for
patients with nonurgent, subacute, or chronic
signs and symptoms, it is likely to be the first
diagnostic modality performed. MRI has become
the standard of imaging for the initial evaluation
of neurovasculitis [13] since the sensitivity to
parenchymal change is unequaled by other modal-
ities. Furthermore, MRI may still be positive in
patients with reversible symptoms.
MRI findings are diverse [10, 13, 25, 27,
28]. The most common findings include
T2/FLAIR hyperintense lesions secondary to
ischemia distributed throughout the subcortical
and deep white matter, the deep gray nuclei, and
the cortex. The MCA territory is the most
involved vascular distribution [29, 30].
31 Imaging of the Neurovasculitides
Diffusion-weighted imaging helps to distinguish
acute, subacute, and chronic ischemia and is man-
datory. Lesions are frequently bilateral and of
differing ages. Involvement of multiple vascular
territories or lesions within a frankly nonvascular
territorial distribution is a clue to the diagnosis
although it should not be confused with a
thrombophilic or cardiogenic multi-embolic pro-
cess producing ischemia. Ischemic lesions are
reportedly present in nearly half of patients with
PACNS [28] and are a constant feature in all the
neurovasculitides (Fig. 5a).
These nonspecific white matter changes may
be the only finding in symptomatic patients [21,
23]. In the younger patient population, these find-
ings are unlikely to be attributed to atherosclerotic
hypertensive disease as in older patients but may
be difficult to distinguish from demyelinating
disease.
Occasionally, diseases exhibit a more charac-
teristic radiographic appearance. Behçet’s disease
is a notable example [31, 32]. Both ischemic and
nonischemic changes secondary to inflammation
or demyelination occur in either a mesodien-
cephalic junction pattern (46 % of patients) or a
pontobulbar pattern (40 % of patients). It is typi-
cally tegmental in distribution. More posterior
mesodiencephalic lesions tend to extend caudal
to the pontine tegmentum and superior cerebellar
peduncles, while more anterior lesions tend to
extend along the corticospinal tracts. The red
nucleus is spared. Upward extension typically
involves the posterior limbs of the internal cap-
sules and the contiguous globus pallidus and/or
putamen. There is no consensus as to the preva-
lence of basal ganglia involvement; it is reported
as occurring in a majority of patients in one study
[32] and within a minority of patients in another
[31]. T2 hyperintensity at the onset eventually
disappears or becomes less prominent, leaving a
T1 iso- or hypointense lesion that is smaller in
size. Nodular enhancement in the acute phase is
frequently present, while hemorrhage is rare. In
isolation from the clinical scenario, the radio-
graphic appearance might be confusing. Mass-
like radiographic characteristics and/or nodular
enhancement may raise concern for a primary
glial neoplasm, metastasis, or CNS lymphoma.
695
However, when the clinical information is consid-
ered, including patient ethnicity along the geo-
graphic “silk route” from the Middle East to
Japan, HLA-B51 variation, and stereotypical
physical signs such as oral and genital ulcers or
uveitis, the imaging findings are virtually patho-
gnomonic (Fig. 6a–c).
Intracranial hemorrhage, typically intrapar-
enchymal or subarachnoid, may be a presenting
or associated radiographic finding for nearly all of
the neurovasculitides [33] although it occurs less
commonly than ischemia. Ischemia is present in
nearly 40 % of patients with PACNS, while hem-
orrhage is present in only 4–12 % [16, 34,
35]. There is uncertainty regarding the signifi-
cance of micro-hemorrhage in these patients.
Prior literature mainly focused on frank hemor-
rhagic lesions easily visualized with CT or con-
ventional MRI sequences. Cortical petechial
hemorrhage, best visualized with T2* technique
(gradient echo or susceptibility weighted imag-
ing), has been suggested to be more commonly
associated with vasculitis than the previously con-
sidered [34] although in one series [36] of
25 patients with variable etiologies of intracranial
neurovasculitis, none demonstrated T2* changes
to suggest acute or chronic blood products
(Fig. 2a). This may be a helpful distinguishing
radiographic characteristic as diseases frequently
included in the differential diagnosis of
neurovasculitis, particularly hypertension and
amyloid angiopathy, are commonly associated
with cortical and white matter T2* susceptibility
artifacts secondary to micro-hemorrhages.
Leptomeningeal enhancement by MRI is pre-
sent in up to 9 % of patients with PACNS [27, 35]
and may be associated with granulomatous histol-
ogy at biopsy (Fig. 5b). This finding is important
as this subset of patients may have negative MRA
and CA [35]. Leptomeningeal enhancement may
also be secondary to infection, itself a separate
etiology for vasculitis, as well as carcinomatous
meningitis and granulomatous inflammatory dis-
eases including neurosarcoidosis. This subset of
patients with PACNS typically present with an
acute clinical onset and cognitive dysfunction,
creating a clinical scenario that may be further
confused with infection or neoplasm. Laboratory
696
Fig. 5 Primary angiitis of the central nervous system.
59-year-old male with remote history of cocaine and alco-
hol abuse. New onset lethargy and confusion progressing
to delirium and agitation. Clinical and laboratory evalua-
tion for systemic disease, including toxicology, was nega-
tive. Initial imaging evaluation reveals numerous infarcts
of varying ages, situated within multiple vascular terri-
tories, with both small perforator and large artery etiology.
Non-contrast CT (a, top) demonstrates multifocal regions
of low attenuation. Those in the right frontal subcortical
A.J. Davis
white matter and left basal ganglia (black arrows) are
sharply defined, without mass effect and likely reflect old
infarctions. Both the cortex and underlying white matter of
the right occipital lobe are involved, as is the right
splenium of the corpus callosum (white arrows). In these
locations, the margins are more ill defined and there is
subtle mass effect characterized by sulcal and ventricular
effacement, suggesting acute ischemia in the right PCA
territory. MR FLAIR imaging (a, middle) demonstrates
central low and peripheral high signal intensities within
31 Imaging of the Neurovasculitides
Fig. 6 Behçet’s disease. 38-year-year old Jordanian male
presenting with ataxia, dysphagia, and dysarthria. Patient’s
brother carried a diagnosis of Behçet’s disease. Axial and
sagittal FLAIR imaging (a, left and right respectively)
demonstrates a strikingly geographic abnormal
hyperintense signal intensity within the tegmentum of the
midbrain and pons. There is little mass effect. This brain
stem distribution is a common appearance of Behçet’s
disease and often extends cephalad to the basal ganglia,
internal capsules, and diencephalon. Areas of diffusion
restriction characterized byT2 diffusion hyperintense sig-
nal and ADC hypointense signal (b left and right
Fig. 5 (continued) the frontal and periventricular white
matter lesions (black arrows) consistent with chronic
encephalomalacia from old infarctions. The FLAIR
hyperintense signal within the right occipital lobe is more
confluent and extends to the posterior temporal lobe and
splenium, involving both the cortex and white matter
(white arrows), and better delineates the extent of the
acute infarct. Diffusion imaging (c, bottom) demonstrates
restricted diffusion confirming the acuity of the occipital
ischemia and the chronicity of the other lesions.
T1-weighted imaging pre- and post-gadolinium (b) dem-
onstrates extensive leptomeningeal enhancement along the
cortical surface of the posterior temporal and occipital
lobes, a well described feature of granulomatous-type
PACNS. CTA (c) demonstrates multifocal vascular
narrowing within several branches of the MCA (white
697
respectively, white arrows) within the left paramedical
caudal midbrain reflect acute or early subacute infarction.
Behçet’s disease often affects the venous system, account-
ing for the “non-territorial” vascular distribution. Hemor-
rhage is unusual. Axial and coronal post-gadolinium
T1-weighted images (c, left and right respectively) dem-
onstrate peripheral, curvilinear, and punctate nodular
enhancement typically associated with Behçet’s disease.
The appearance may sometimes be confusing – this patient
was initially referred for evaluation of a brain stem tumor
(Case courtesy of Adam Davis, MD, New York University
Langone Medical Center, New York)
arrows) with intervening regions of normal-appearing vas-
culature. At the bottom of the image, vascular narrowing
within the PCA (not marked) is present. More than two
lesions within at least two territories is an important diag-
nostic criterion for MRA to avoid false-positive results. CA
(d and e) reveals completely normal extracranial vascula-
ture, helping to narrow the differential diagnosis toward the
diagnosis of PACNS. The ACA (black arrowheads), MCA
(black arrows), and PCA (black-outlined arrows) vascula-
ture demonstrate short-segment mild-to-severe stenoses.
The larger vessels are not involved nor are the branch
points, a feature more commonly associated with athero-
sclerotic disease further helping to arrive at the diagnosis of
PACNS (Case courtesy of Adam Davis, MD, New York
University Langone Medical Center, New York)
698
evaluation including CSF examination and
leptomeningeal biopsy may be required to differ-
entiate these patients.
Unusual MRI findings associated with cerebral
vasculitis have been reported and may be mislead-
ing [27]. For example, a tumor-like appearance
characterized by single or multiple discrete
lesions with increased T2/FLAIR and decreased
T1 signal intensity, surrounding white matter
edema pattern, enhancement, and mass effect
was seen in vasculitis patients, particularly those
with Behçets, PACNS, and lymphocytic vasculitis
[37–40] (Figs. 4, 5, and 6). Up to 4 % of patients
with PACNS present with this finding and may be
associated with amyloid angiopathy at
histology [35].
The variability of imaging findings of cerebral
vasculitis is nowhere more clearly demonstrated
than in a case report by Qu et al. [40] of a patient
with the ultimate diagnosis of PACNS who over a
7-month period of time was sequentially diag-
nosed with idiopathic headache following a nor-
mal head CT; multifocal basal ganglia infarcts
secondary to tuberculous meningitis following
an abnormal CT with lesions in the basal ganglia;
viral meningoencephalitis following an abnormal
CT and MRI with ischemic lesions in the basal
ganglia; malignant glioma following an MRI
demonstrating a parietal lobe mass lesion with
midline shift; and, finally, cerebral vasculitis diag-
nosed only when the MRA and conventional
angiogram demonstrated diffuse vascular steno-
ses. Brain biopsy confirmed the diagnosis.
Figure 7 demonstrates a case of HIV vasculitis
presenting with ischemia which initially was diag-
nosed and followed as congenital fusiform aneu-
rysms and thromboembolic events.
Perfusion CT or MRI imaging may be helpful
in the evaluation of vasculitis since symptomatic
patients with normal MRI findings may have per-
fusion abnormalities [36]. These abnormalities
include delayed perfusion and reduced relative
cerebral blood volume [27, 39]. Perfusion imag-
ing is suggested for all symptomatic patients with
initial negative CT or MR imaging studies not
only in the evaluation of cryptic vasculitis but to
exclude the presence of ischemia secondary to the
more common vasculopathies. Magnetic
A.J. Davis
resonance spectroscopy may reveal elevated glu-
tamate and glutamine and reduced N-acetyl aspar-
tate [39, 41]. While there is no consensus on the
role of choline [39], absence of a choline elevation
helps to exclude a neoplastic etiology, although
the presence of a choline peak does not exclude
PACNS or inflammatory lupus vasculopathy
[41, 42].
MRI is the most sensitive imaging modality
available to the clinician for the evaluation of the
neurovasculitides. Its utility and reliability is
nonetheless dependent on the clinical scenario.
For example, patients with focal sensorimotor
deficits secondary to infarction will have 100 %
sensitivity on MRI, while patients with more
nonspecific symptoms such as headache second-
ary to vascular or leptomeningeal inflammation
will have a lower MRI sensitivity. The literature
reports the sensitivity for MRI and parenchymal
abnormalities for all types of cerebral vasculitis
ranges from 83 % to 100 %, mostly exceeding
90 % [15, 28, 29, 36, 43–46]. Some studies report
a 100 % sensitivity of MRI for all types of auto-
immune neurovasculitis (though more accurately,
this sensitivity is based on cross-sectional imag-
ing since some of these studies were combined
with CT) and propose that normal MRI excludes
the diagnosis [30, 44–46]. Other studies conclude
that MRI is sensitive but not infallible, citing
series of patients with angiographic, clinical, or
biopsy-proven cerebral vasculitis but negative
MRI [29, 30, 36, 47]. However, many of these
series are problematic in that FLAIR imaging,
commonly considered the most sensitive of the
long TR sequences for detecting gray and white
matter and T2 signal intensity abnormalities, was
not included in the imaging evaluation.
Despite the high sensitivity of MRI, specificity
has been reported to be as low as 19 % [15]. It
might be expected that CA, which provides the
greatest vascular luminal resolution (see “Indirect
Imaging Signs: Vessel Lumen” below), would
significantly improve the diagnostic sensitivity
and specificity. Yet surprisingly, studies demon-
strate that CA provides little additional benefit in
specificity when differentiating vasculitis from
noninflammatory vasculopathies [15, 21, 36]. Fur-
thermore, CA does not improve sensitivity as
31 Imaging of the Neurovasculitides
Fig. 7 HIV vasculitis. 19-year-old female with congenital
HIV. History of cerebral infarction at the age of 10 years old.
Imaging evaluation at age 10 with MRA (not shown)
revealed intracranial bilateral ICA supraclinoid segment
fusiform dilatation with an irregular patulous contour
extending into the proximal middle and anterior cerebral
arteries. Findings were presumed to represent congenital
aneurysmal malformations and the patient was followed
periodically with MRA. MRI FLAIR imaging (a, right)
demonstrates abnormal hyperintense signal and volume
loss within the cortex and underlying white matter of the
right parietal and far lateral occipital lobes consistent with
remote right MCA territory infarction. T1-weighted imaging
(a, left) demonstrates gyriform increased signal intensity
which was stable for years likely secondary to dystrophic
calcification or laminar necrosis. MRA (b)) demonstrates
699
bilateral, long-segment, irregular, fusiform dilatation of the
ICA and the most proximal MCA and ACA segments (white
arrows). The right ICA and MCA reveal areas of luminal
stenosis not appreciated on earlier examinations. CTA
reveals a more detailed and diagnostic view of the intracra-
nial vasculature. The right MCA (c) demonstrates regions of
short- and long-segment concentric vascular stenoses (black
arrows) with intervening luminal dilatation (black-outlined
arrows), a hallmark of vasculitis. Infectious thromboem-
bolic disease may produce mycotic aneurysm that result in
vascular dilatation secondary to pseudoaneurysm formation
but may be distinguished from vasculitis as it is typically
eccentric, not fusiform as in this case, and often causes
small-vessel occlusion. The ACA demonstrates short- and
long-segment concentric stenoses (white arrows) and a
small pericallosal saccular aneurysm (white-outlined
700
compared with MRI. There are numerous case
reports and series [15, 21, 30, 41, 44] of patients
with clinical or biopsy-proven cerebral vasculitis
with a positive MRI and a negative catheter angio-
gram, traditionally considered the imaging gold
standard.
In summary, the literature confirms the impor-
tant, if not essential, role of MRI in the diagnostic
evaluation of the neurovasculitides. A negative
MRI may not definitively exclude cerebral vascu-
litis in a symptomatic patient, but it makes the
diagnosis extremely unlikely. Given a negative
MRI, other etiologies for the patient’s symptoms
should be investigated. If strong clinical suspicion
for vasculitis remains, dedicated vascular imaging
including CTA, MRA, or CA should be
performed.
Indirect Imaging Signs: Vessel Lumen
Indirect imaging techniques in the investigation of
neurovasculitis characterize changes in the vessel
lumen secondary to mural or perivascular inflam-
matory processes and include MRA, CTA, and
CA. While they do not provide direct evidence
of inflammation, these studies convey salient
information about the pathologic process causing
ischemia, infarction, and hemorrhage. The infor-
mation is crucial to the diagnosis, risk stratifica-
tion, and management of these patients. While
debate exists as to the sensitivity and specificity
of each of these techniques, their utilization
remains essential to the diagnosis of cerebral
vasculitis.
Fig. 7 (continued) arrow). This constellation of radio-
graphic findings moved the diagnosis from congenital
aneurysm to HIV vasculitis. The patient presented one
year later with right hemiparesis. MRI now demonstrates
extensive abnormal FLAIR hyperintensity within the left
basal ganglia and left frontal lobe cortex and underlying
white matter, with restricted diffusion (d left and right
respectively). Findings are consistent with left MCA and
ACA acute infarctions. CTA (e, maximum intensity pro-
jection, AP on the left and oblique on the right) now
demonstrates occlusion of the skull base, left petrous
A.J. Davis
Spatial resolution and temporal resolution are
important performance benchmarks to consider
when comparing the common angiographic tech-
niques. Catheter angiography, a minimally inva-
sive procedure, provides the highest spatial
resolution (up to 0.2 mm) and fastest temporal
resolution (0.5–0.25 s for a typical study)
[11]. By comparison, the spatial resolution of
modern multi-detector CTA, which is dependent
on detector row thickness, is approximately
0.4–0.75 mm. CA provides detailed information
regarding hemodynamics that is generally absent
from both CTA and MRA. However, the most
recent 320 row-detector CT technology does pro-
vide limited hemodynamic information with a
temporal resolution of 1 s and a spatial resolution
of 0.5 mm [48]. Indices for MRA spatial resolu-
tion are even less precise than these other
techniques.
Catheter Angiography
CA is the oldest of the indirect imaging modalities
currently utilized for the evaluation of the
neurovasculitides. It had been considered the
gold standard for diagnosis but now plays a less
important role with the evolution of noninvasive
vascular imaging techniques that provide similar
and, in some circumstances, equivalent
information.
The hallmark of vasculitis is multifocal lumi-
nal narrowing, vascular contour irregularity, and
vascular dilatation occurring within multiple ves-
sels and multiple vascular territories. The classic
imaging finding is segmental luminal narrowing
followed by vessel dilatation, an appearance
ICA, and cervical ICA segments (white arrows). The bilat-
eral ACAs are occluded (open white arrows) and the left
MCA is now markedly stenotic (white arrowhead). Find-
ings are consistent with disease progression and luminal
thrombosis. CTA (g, volume-rendered imaging) demon-
strates extensive collateral circulation to the ACA (open
white arrows) from the right PCA via cortical
leptomeningeal branches and posterior choroidal branches
(white arrows) (Case courtesy of Adam Davis, MD,
New York University Langone Medical Center, New York)
31 Imaging of the Neurovasculitides
Fig. 8 Primary angiitis of the central nervous system.
74-year-old male with hypertension and no other signifi-
cant medical problems. Recurrent cerebral infarctions.
Systemic evaluation including CSF analysis was negative.
CA (a and b) reveals multifocal long- and short-segment
stenoses. Note the concentric appearance and relatively
smooth contour of the vessel lumen consistent with
referred to as “string of beads” (Figs. 3c and 7b,
c, d). Unfortunately, vascular dilatation is an
inconstant finding. More typically, angiography
of vasculitis reveals multifocal stenoses with
intervening vessel ectasia or normal luminal
diameter (Figs. 5e, and 8a, b). Luminal contour
may be smooth (Figs. 5e, and 8a, b) or irregular
(Fig. 2c, d). Stenoses are classically short segment
and discrete, but occasionally may be more
elongated.
The size of the vessels affected and the distri-
bution of lesions within each of the vessels vary
with the etiology of the neurovasculitis and help to
differentiate them. Large-vessel extracranial
narrowing is typically long segment and undulat-
ing when associated with TA (Fig. 9) and GCA
(Fig. 1). Intracranial skull base large-vessel and
cortical medium-vessel involvements are more
typical of the systemic primary vasculitides such
as the ANCA positive diseases and the autoim-
mune diseases such as SLE and PACNS. Intracra-
nial luminal manifestations tend to be more
variable than extracranial disease.
PACNS does not have a predilection for vessel
branch points in contradistinction to the autoim-
mune inflammatory diseases [49] and atheroscle-
rosis. Luminal narrowing secondary to vasculitis
is typically not at regions of laminar flow disrup-
tion or regions of high shear stress, which includes
701
PACNS, as opposed to the more irregular appearance asso-
ciated with atherosclerotic disease. Branch points are not
involved. There is no abrupt distal vessel occlusion as
might be expected with thromboembolic disease (Case
courtesy of Adam Davis, MD, New York University
Langone Medical Center, New York)
the ICA bulb, ICA petrocavernous junction, and
ICA cavernous segment. This helps to differenti-
ate vasculitis from atherosclerosis which com-
monly occurs at these locations. Multifocal
stenosis and luminal irregularity within the same
vessel segment with intervening normal vessel
contour favor vasculitis over atherosclerosis. Sim-
ilarly, isolated stenoses within separate vascular
territories and otherwise normal vessel contours
(Figs. 5c, e, and 8a, b) are an unusual appearance
for atherosclerosis and favor vasculitis. It is
important to remember that discrete lesions within
multiple vascular territories may occur secondary
to multi-embolic disease, although emboli are
typically more obstructive than stenotic and tend
not to occur in tandem within the same vessel as
does vasculitis.
There are some attributes which help to distin-
guish medium-vessel disease secondary to sys-
temic vasculitis and autoimmune disease from
PACNS. Abrupt vascular truncation and occlu-
sion are typically seen with medium- and small-
vessel autoimmune vasculitides, but less so for
PACNS. Microaneurysmal formations, likely
pseudoaneurysms, are more commonly associated
with autoimmune inflammatory diseases, particu-
larly polyarteritis nodosa (Fig. 10a–d) and in the
author’s experience with drug-induced vasculitis,
SLE and infectious vasculitis. Microaneurysmal
702 A.J. Davis

Fig. 9 Takayasu arteritis and collateral circulation. Case aortic arch. The brachiocephalic, subclavian, common
1: 25-year-old Indian female with long-standing TA and carotid, and vertebral arteries are among the most fre-
syncopal episodes. CA and CTA (a and b) demonstrate quently diseased in TA. The vertebral arteries are opacified
complete occlusion of the great vessels arising from the by collateral circulation (a, black arrows). Note the
31 Imaging of the Neurovasculitides
formation is uncommon with PACNS; none were
seen in an angiographic series reported by Cloft
et al. [50].
Vessel dissection is a rare and serious event and
may be associated with the transmural necrotizing
vasculopathies including polyarteritis nodosa
(Fig. 10b, c, d). It is an unusual complication of
extracranial GCA (Fig. 1c).
When stenoses are severe or vascular occlusion
is present, typically within the large and medium
vessels, collateral circulation should be investi-
gated (Fig. 9a–f). Abnormal cerebral hemody-
namics, typically characterized by slow
antegrade flow, diminished distal luminal size
(Fig. 9e, f), or prolonged circulation time, is com-
monly described. When present, it may be addi-
tionally quantified with CT or MR perfusion
imaging.
Despite the long experience with CA and the
widespread notion that it represents the “gold
standard” of diagnostic imaging, the utility of
CA has been questioned. Numerous case reports
and studies describe patients with clinical, MRI,
or biopsy-proven cerebral vasculitis and negative
angiography. The sensitivity of CA for PACNS
has been reported to be from 40 % to 100 %
[18, 21, 27, 28, 30, 44, 46, 49, 50] depending
upon whether clinical, radiographic, or histologic
criteria are utilized to ultimately confirm the diag-
nosis. For example, the specificity of CA for mak-
ing the diagnosis of PACNS as compared with
histology is reported to be no higher than 37 %
Fig. 9 (continued) residual stumps of the brachiocephalic,
common carotid, and left subclavian arteries (b, black
arrows). CA (c) demonstrates extensive collaterals to the
cephalic circulation via the intercostal arteries (black
arrows) to the deep cervical and vertebral artery (black
arrowheads). The distal deep cervical branches (d, black
arrows) opacify the occipital artery (black arrowhead)
which via retrograde flow fills the ICA (black-outlined
arrows). CTA (e) documents the same collateral circulation
although flow dynamics are absent (deep cervical artery,
white arrows; occipital artery, white arrowheads; ICA,
white-outlined arrows). Even very small vasculature that
provides clinically significant collateral circulation can be
demonstrated by CTA (f) including the ascending cervical
artery to the vertebral artery to a spinal radicular artery
703
[15]. This is an important benchmark since imag-
ing is essential to the diagnosis of PACNS because
there are no systemic laboratory tests or clinical
signs or symptoms to otherwise confirm the diag-
nosis (Figs. 5, and 8).
Additionally, PACNS and other autoimmune
vasculitides may present with a predominant
small-vessel involvement that is beneath the res-
olution of CA. Many of these patients are symp-
tomatic, with a positive MRI but negative
angiography. Multiple studies have confirmed
the poor correlation between findings on MRI
and CA [15, 30, 50]. Pomper et al. noted that
only 65 % of lesions detected by MRI had a CA
correlate and only 44 % of lesions detected by
angiography were evident on MRI. Similarly, in a
series of patients with PACNS, of 41 territories
involved by MRI, approximately 15 % were nor-
mal on angiography and conversely within 50 vas-
cular territories involved by CA, 34 % were
normal on MRI.
The utility of CA is important to consider as
these studies carry risk, with reported permanent
and transient neurologic morbidities of 1 % and
10 %, respectively, during the evaluation for vas-
culitis [51]. In the author’s experience, intrave-
nous corticosteroids prior to the procedure may
help reduce complications. The benefit of the
study needs to be carefully considered with regard
to the potential harm. Nonetheless, when clinical
suspicion remains high and noninvasive imaging
is inconclusive, the high spatial resolution of CA
(white arrows) to the anterior spinal artery (white arrow-
heads). Case 2: 46-year-old Indian female with worsening
dizziness. CTA demonstrates isolated involvement of the
left common carotid artery, beginning at the origin and
extending the length of the vessel. Note the severe contin-
uous long-segment luminal narrowing characteristic of the
disease (g, left, maximum intensity projection; middle,
volume-rendered image, white-outlined arrows). The
internal and external carotid arteries (g, right, volume-
rendered image, black-outlined arrows) are diminished in
caliber but with an otherwise smooth lumen and normal
morphology, likely resulting more from decreased flow
than from direct disease involvement (Case courtesy of
Adam Davis, MD, New York University Langone Medical
Center, New York)
704 A.J. Davis

Fig. 10 Polyarteritis
nodosa. 43-year-old man
with known polyarteritis
nodosa presents with neck
swelling and bruit. CA of
the celiac axis performed
5 years prior to presentation
(a) demonstrates multifocal
luminal narrowing (black
arrows) and aneurysmal
dilatation (black
arrowhead). Both MRA
and CTA (b, c) demonstrate
a collection of prominent,
tortuous vasculature at the
posterior cervical space
enveloping the distal left
vertebral artery. Prominent
draining veins are noted
(white arrows). CA of the
left vertebral artery (d,
black arrow) reveals the site
of arteriovenous shunting
(black arrowhead) into a
dilated, tortuous draining
vein confirming the
diagnosis of an
arteriovenous fistula. This
was likely due to vascular
necrosis or aneurysmal
rupture. Both CTA and CA
(e, left and right
respectively) demonstrate a
discrete region of luminal
irregularity and narrowing
within the ICA proximal
cervical segment (black
arrows). In isolation, this
might be confused with
fibromuscular dysplasia
(Case courtesy of Adam
Davis, MD, New York
University Langone
Medical Center, New York)
31 Imaging of the Neurovasculitides
continues to make it an important adjunct in the
assessment of the neurovasculitides.
CT Angiography
The utility of CTA has been well established for
the large-vessel vasculitides, particularly TA
(Fig. 9) and GCA. Common conventional angio-
graphic findings including stenosis, occlusion,
dilatation, and luminal irregularity are well dem-
onstrated by CTA as are the distribution and
severity of involvement [52]. The sensitivity and
specificity of CTA for TA are 93–95 % and
98–100 % respectively [53]. Furthermore, CTA
provides direct imaging evidence of inflammatory
vasculopathies by detecting mural changes that
are not evident on CA (Fig. 11). This is particu-
larly important in the early stages of the disease
when mural inflammatory changes may precede
luminal contour changes. The role of CTA in the
medium-vessel intracranial vasculitides is not as
well defined. CTA is routinely utilized to assess
the vessels at the circle of Willis and the 2nd- and
3rd-order branches of the anterior, middle, and
posterior cerebral arteries. These segments are
frequently involved by vasculitis (Fig. 3c). Steno-
sis, dilatation, and occlusion are well seen, while
luminal irregularity is less consistent, given the
limits of the resolution of the modality. CTA is a
non-dynamic study, and collateral flow informa-
tion is not easily discernable as with CA but can
still be inferred by the angioarchitecture (Fig. 7g)
of the opacified vessels. Overall, CTA is noninva-
sive, with high spatial resolution, and may be
obtained quickly at any time of day, prompting
its frequent use as the initial imaging study for
suspected neurovasculitis. The radiation dose
penalty, generally slightly greater than a routine
head CT, should always be considered. This may
dissuade the clinician from utilizing this modality
with children and young adults or as a screening
examination for adults.
MRA
MRA is a frequently utilized modality in the
assessment of extracranial and intracranial vascu-
lar diseases, owing to its distinct advantages as
compared with CA and CTA: it is noninvasive,
lacks ionizing radiation, and for intracranial
705
studies does not require contrast administration.
It possesses the lowest spatial resolution of the
common vascular imaging modalities and is prone
to overestimation of stenosis secondary to dimin-
ished signal intensity from vessel tortuosity or
slow flow. MRA reveals vascular findings that
are similar to the other luminal imaging modalities
including intracranial stenoses and occlusions
(Fig. 2c). More subtle findings including intracra-
nial vessel irregularity are more difficult to assess
due to lower spatial resolution. Compared with
CA in a series of patients with suspected vasculitis
(not all of whom were ultimately diagnosed with
the disease), MRA revealed a sensitivity and spec-
ificity of 79 % and 87 % for diagnosing vasculitis
while CA was 95 % and 97 % [54]. In a separate
study of patients with various etiologies for
neurovasculitis, MRA provided a sensitivity of
92 %. Given the criteria that more than two ste-
noses are identified within at least two vascular
territories, the false-negative rate of MRA
improves markedly, approaching that of CA [54]
eliminating the need for further indirect vascular
luminal imaging. Unfortunately, if MRA is nega-
tive or if these stringent criteria are not met, CA or
CTA is still required.
Direct Imaging Signs: Vessel Wall
The literature clearly supports the utility of direct
vessel investigation in the diagnosis of
neurovasculitis since demonstrating vessel wall
and/or perivascular inflammatory changes pro-
vides direct evidence of the disease. The sensitiv-
ity and specificity depend not only upon the
modality utilized but the etiology as well. Large-
vessel extracranial vascular disease is typically
more obvious than intracranial vasculopathy
where the smaller vessel size and close proximity
of arteries and veins make investigation more
difficult.
CT and CTA
TA and GCA, both large-vessel extracranial gran-
ulomatous transmural arteritides, demonstrate
vessel wall abnormalities with all imaging modal-
ities (Fig. 11). CT readily identifies abnormal
706
Fig. 11 Large-vessel inflammatory disease with features of
both Takayasu arteritis and giant cell arteritis. 51-year-old
male from Mali with fever, chills, night sweats, neck swell-
ing, and jaw pain. CTA (a) demonstrates marked mural
thickening with heterogeneous enhancement within the bilat-
eral cervical vasculature, (left to right) left subclavian artery,
left CCA, and left ICA. The black arrows indicate the outer
wall contour, normally a barely imperceptible margin, while
the white arrows indicate the enhancing, narrowed vessel
lumen. The appearance is consistent with Takayasu arteritis.
CTA volume rendering (b) demonstrates long-segment
smooth narrowing of the left ICA (black arrow) with more
A.J. Davis
localized narrowing within the external carotid artery
branches (arrowheads) typically found with TA. CTA vol-
ume rendering (c) demonstrates narrowing of the superficial
temporal artery (black arrow) as would be expected in giant
cell arteritis. Biopsy demonstrated internal elastic lamina
fragmentation and microaneurysm formation although with-
out granulomatous change as would be expected for either
disease – failing to provide a definitive diagnosis. Following
prompt and aggressive treatment with corticosteroids, the
mural thickening within the left subclavian artery, CCA,
and ICA resolved (d) (Case courtesy of Adam Davis, MD,
New York University Langone Medical Center, New York)
31 Imaging of the Neurovasculitides
mural thickening as defined by a thickness greater
than 1 mm in 93 % of patients with clinical evi-
dence of TA [52]. Seventy-three percent of these
patients demonstrate changes within the
cervicocerebral vessels, most commonly the
aorta, brachiocephalic artery, and common carotid
artery where wall thickening of up to 10 mm is
observed. CA, once considered the gold standard
for detecting abnormal vasculature, may be nega-
tive in the early stages of TA when mural abnor-
malities are present [52, 55]. This is an important
consideration as the disease is most effectively
treated with immunosuppressive therapy during
the earliest phase of the illness; a time when CA
may be nondiagnostic.
CTA is an excellent option in these circum-
stances as it provides not only information regard-
ing luminal abnormalities, including stenosis,
occlusion, aneurysmal dilatation, and contour
irregularities similar to CA, but direct diagnostic
findings including wall thickening, calcification,
and abnormal enhancement [52, 55]. While not
specific to the cerebrovascular circulation,
Yamada et al. [53] demonstrated thoracic CTA to
have a sensitivity and specificity of 95 % and
100 %, respectively, for the diagnosis of TA. The
utility of direct imaging is well documented by
Park et al. [55] who correlate the vessel wall CT
abnormalities with the histopathologic findings.
The authors demonstrated that heterogeneous
mural enhancement and an inner concentric low
attenuation ring which enhances at delayed imag-
ing correlate with the acute-phase histopathologic
findings of tunica media, vascularization, and inti-
mal inflammatory swelling.
MRI
MRI provides similar and in some cases more
sensitive information than CT regarding vessel
wall changes, particularly in those neurovas-
culitides affecting the extracranial vasculature. In
patients with TA, mural thickening and enhance-
ment of the aorta, as compared with the myocar-
dium as a reference, correlate significantly with
clinical disease activity as defined by erythrocyte
sedimentation rate and C-reactive protein [56]. T2
707
signal intensity changes are less sensitive than
enhancement. Similar to CT, MRI provides
radiopathologic correlation. Utilizing both mural
morphology and signal intensity, these same
authors identified six distinct MRI patterns that
likely correlate with histopathologic inflamma-
tory changes, vascularization, and scarring/
fibrosis.
High-resolution MRI evaluation of extracra-
nial vessel walls for inflammatory changes can
play an important role in the diagnosis and
assessment of patients with suspected GCA.
Utilizing high-resolution, gadolinium-enhanced,
fat-saturation, T1 spin-echo technique, Bley
et al. [57] stratified MRI findings based on the
severity of wall thickening and prominence of
mural enhancement within the frontal and parietal
superficial temporal arteries and occipital arteries.
They demonstrated 81 % sensitivity and 97 %
specificity of MRI as compared with the final
rheumatologists’ diagnosis based on clinical
exam, laboratory studies, and/or biopsy results.
Some positive MRI results were biopsy negative,
although the final diagnosis of GCA was con-
firmed clinically. This may be attributed to the
intermittent vessel involvements (“skip lesions”)
that occur with GCA (Fig. 1b), leading to false-
negative biopsy results when a normal segment of
artery is surgically sampled. MRI is able to assess
the entire length of the vessel, leading to a more
confident diagnosis.
Obtaining clinical history regarding immuno-
suppressive therapy is mandatory when imaging
patients with suspected neurovasculitis. When
patients are imaged following treatment with cor-
ticosteroids, false-negative MRI results may
occur. Bley et al. [57] report that if only GCA
patients who received treatment with steroids for
less than 10 days are considered, sensitivity
increased from 81 % to 85 %. For those receiving
steroids for more than 10 days, sensitivity falls to
33 %. Therefore, radiologists should question
patients regarding steroid therapy since patients
typically present for imaging days to weeks after
seeing the referring physician. Vigilance for false-
negative tests is important, although it is
708
reassuring to know that MRI still provides sensi-
tive and specific information if imaging is
performed within days following the initiation of
immunosuppressive treatment.
MRI provides direct evidence of intracranial
vasculitis [10, 36]. Findings include asymmetry
and thickening of the vessel wall, an eccentric or
narrowed lumen, and mural enhancement. Thin-
sliced images and flow saturation technique are
mandatory since images may be marred by an
artifact from laminar flow adjacent to the vessel
wall or by periarterial venous enhancement.
Enhancement may be limited to the vessel wall
or extend into the adjacent leptomeninges. Kuker
et al. [36] reported on 27 patients with
neurovasculitis secondary to varying etiologies
in which 25 patients had mural thickening and
23 demonstrated abnormal enhancement, con-
firmed in more than one plane of imaging within
the same study. Approximately, the same number
of patients demonstrated abnormal stenosis with
MRA or abnormal parenchymal T2 signal inten-
sity with MRI, although these findings were rela-
tively nonspecific. Kuker [36] contends that for
intracranial large-vessel disease, the greater spec-
ificity of direct vascular wall inflammatory
changes provides more convincing evidence than
secondary imaging findings and perhaps exceeds
that of biopsy.
Nuclear Medicine
Nuclear medicine evaluation of neurovasculitis
remains promising but problematic. Conceptually,
the ability to monitor metabolic activity within the
vessel wall should be a good indicator of inflam-
matory activity. 18F-fluorodeoxyglucose positron
emission tomography (FDG-PET) has been the
most well-studied radionuclide for this indication,
particularly large-vessel diseases including TA
and GCA. A meta-analysis of the literature [58]
demonstrated a wide variability of diagnostic sen-
sitivities for TA ranging from 28 % to 100 %,
while the range of specificity was 50–100 %.
The pooled analysis nonetheless indicates modest
results with a diagnostic sensitivity and specificity
of 70 % and 77 % respectively. FDG-PET CT
provides superior anatomic localization and
improved sensitivity and specificity of 91 % and
A.J. Davis
89 % for this disease [78]. Greater diagnostic
sensitivity (80 %) and specificity (89 %) from
FDG-PET in cases of GCA have been obtained
[59]. The specificity of FDG-PET is degraded by
the presence of atherosclerosis as active inflam-
matory plaques may produce false-positive find-
ings [60] and nuclear medicine studies should be
interpreted with caution in the elderly.
The role of FDG-PET for monitoring disease
activity in patients with known vasculitis is still
not well defined. Some patients deemed inactive
by clinical criteria have biopsy-proven active
inflammation [61]. Studies indicate that
FDG-PET may be a sensitive and helpful diag-
nostic study for identifying these patients with
subclinical active disease [66]. Eighty-three per-
cent of patients with biopsy-proven GCA had
positive FDG-PET studies [62]. The sensitivity
and specificity of FDG-PET to identify active
disease as compared with both clinical signs and
laboratory criteria are 100 % and 89 % [63]. Inter-
estingly, one of two patients considered false pos-
itive due to lack of clinical and laboratory
evidence went on to develop active disease within
several weeks, suggesting that subclinical disease
may be uncovered with nuclear medicine studies.
The issue is clouded by the fact that these patients
are treated with immunosuppressive agents that
may halt overt clinical manifestations and prompt
false-negative nuclear medicine studies, despite
the presence of indolent subclinical disease activ-
ity. This is similar to the difficulties with MRI in
detecting mural enhancement following immuno-
suppressive therapy discussed earlier. Nonethe-
less, nearly all practitioners recommend the
utilization of nuclear medicine studies to assess
disease progression, treatment efficacy, and dis-
ease recurrence particularly when there is uncer-
tainty in the clinical assessment as defined by
physical signs and symptoms or laboratory
markers.
Ultrasound
Ultrasound (US), specifically color duplex imag-
ing, may be a helpful modality in the diagnosis of
vasculitis by providing direct imaging evaluation
of the vessel wall. It has been most extensively
studied in the more common extracranial large
31 Imaging of the Neurovasculitides
artery vasculitides, TA and GCA. It is an appeal-
ing technique as it is noninvasive, without ioniz-
ing radiation; can be easily administered; and may
be repeated frequently to follow disease progres-
sion or treatment response. Additionally, US pro-
vides a high-resolution imaging evaluation as
compared with MRI, detecting wall thickness dif-
ferences with a resolution of 0.1 mm [64].
Concentric homogeneous mural thickening,
stenosis, and occlusion of the aorta and
brachiocephalic branches are typical US findings
of TA [60, 65, 66] which may be differentiated
from atherosclerotic disease by the absence of
plaque formation, concentric morphology, long
segment involvement, and location. US reveals
abnormalities more frequently in the common
carotid artery than the internal carotid artery with
vessel wall thickness greater than that of control
subjects. In patients with TA, the common carotid
artery wall thickness is greater than that of the
carotid bulb, the reverse of normal control sub-
jects. US demonstrates subtle mural changes char-
acterized by a homogenous, circumferential,
mid-echoic wall thickening within the subclavian
and carotid arteries in the early stages of the dis-
ease – before abnormalities may be detected by
CA [64]. Common carotid artery intima–media
thickness ratio is increased in patients with TA
as compared with normal control patients and
the degree correlates with disease activity
[67]. The sensitivity and specificity of this finding
as compared with clinical criteria are 82 % and
70 %.
Ultrasound has been well studied in patients
with GCA. The wall diameter of the superficial
temporal artery (common, frontal, and parietal
divisions) is significantly greater in patients with
GCA than in symptomatic patients without the
disease or asymptomatic age-matched controls
[68]. A hypoechoic halo surrounding a patent
vessel lumen was found in 73 % of biopsy-proven
vasculitis patients but not in any symptomatic
patients without GCA or asymptomatic controls,
indicating that it may be a very specific imaging
finding. Interestingly, histopathologic findings
characterized by mural cellular infiltrate did not
correlate with the hypoechoic halo which has been
attributed to edema. In all treated patients, the halo
709
disappeared at a mean of 16 days. Stenoses and
occlusions are commonly seen in all symptomatic
patients, and while significantly more common in
patients with GCA (80 %) than those without, it is
not as specific a finding as the halo. Similar find-
ings are present in the occipital arteries [69]
although the sensitivity is less when compared
with the superficial temporal arteries. Nonethe-
less, it is a useful exam for symptomatic patients
presenting with nuchal pain, occipital headache,
or occipital scalp tenderness when occipital artery
involvement may be the only imaging manifesta-
tion of the disease. Specificity is increased when
more than 3 arterial segments are involved. A
meta-analysis of the literature [70] demonstrated
that if all US findings are considered – halo sign,
stenosis, and occlusion – the sensitivity and spec-
ificity are 88 % and 78 %, respectively, as com-
pared with biopsy. The specificity increases to
82 % when only the halo sign is considered.
Conclusion
The data for parenchymal, luminal, and vessel
wall imaging may at first glance suggest that the
evaluation of vasculitis is a daunting task, partic-
ularly for PACNS and the primary autoimmune
vasculitides, but this is true only if each imaging
modality is considered in isolation. Diagnosis of
neurovasculitis should be considered on a
per-patient basis and not a per-territory or
per-modality basis. The largest single series of
PACNS [28] demonstrated that 77 % of CT,
97 % of MRI, 59 % of MRA, and 90 % of
conventional angiographic studies were ulti-
mately positive. By comparison, only 62 % of
the brain biopsies were positive. Consequently,
when all imaging modalities including those that
provide parenchymal, luminal, and mural evalua-
tion are brought to bear on a given patient with
suspected vasculitis, the entire constellation of
findings typically brings clarity to the situation.
While not specific, few patients are missed by
imaging evaluation. When combined with the
clinical and laboratory results as well as biopsy,
the diagnosis becomes more assured even in the
most difficult of clinical scenarios [15].
710
Mimics
There are numerous diseases that present with
clinical manifestations and imaging findings that
are easily confused with and difficult to distin-
guish from neurovasculitis. Any vasculopathy
resulting in extracranial or intracranial luminal
stenosis, occlusion, or irregularity and associated
with cerebral or spinal cord ischemia, hemor-
rhage, or inflammation may be suspect. Marsh
et al. [2] report on 5 patients referred for treatment
of PACNS who were ultimately found to have
intracranial atherosclerosis, B-cell lymphoma,
cerebral amyloid angiopathy, and RCVS. Neel
et al. [71] report that RCVS, intracranial athero-
sclerosis, infection, malignancy, and embolism
may be confused with PACNS. More frequently,
a patient will present with ischemia secondary to a
common noninflammatory vasculopathy (e.g.,
atherosclerosis) and suffer from an incidental sys-
temic inflammatory disease (e.g., rheumatoid
arthritis). The combination of cerebral ischemia,
vasculopathy, and systemic inflammatory labora-
tory abnormalities may mistakenly lead to a diag-
nosis of vasculitis. These cases may be difficult to
clarify. Careful search for direct vascular imaging
signs of mural inflammation or a characteristic
pattern of vasculitic involvement may help.
Sometimes biopsy is required to differentiate.
Atherosclerosis
Atherosclerosis particularly when it is predomi-
nantly or exclusively intracranial is the most com-
mon and sometimes most difficult disease to
differentiate from vasculitis. Likewise, hyperten-
sive vasculopathy of the medium and small
intracranial vessels may cause confusion. Indirect
luminal imaging, whether MRA, CTA, or CA,
provides subtle differentiating clues. Atheroscle-
rotic plaque tends to occur at branch points and
is typically better defined, shorter in length, and
more eccentric than the smooth, concentric, tapered
morphology and distal location of primary and
secondary vasculitides. The presence of mural cal-
cification is helpful as it is more commonly asso-
ciated with atherosclerosis. Venous involvement is
A.J. Davis
consistent with inflammatory vasculopathy. Mural
enhancement by CT or MRI may definitively dif-
ferentiate vasculitis. A positive PET-CT will be
helpful but it should always be remembered that
atherosclerosis may be falsely positive.
Tumor
Neoplastic involvement of the vasculature is an
important entity to exclude. Tumor may directly
invade the vessel wall or propagate within the
lumen and may be difficult to differentiate from
vasculitis.
Intravascular B-cell lymphoma is a rare type of
extranodal large B-cell lymphoma typically
occurring in older patients which may present
with CNS manifestations and constitutional
symptoms mimicking a systemic vasculitis
[72]. CNS symptoms are typically secondary to
ischemia. Morphologically, the tumor growth is
restricted to the lumen of small vessels and is
often widely disseminated at presentation. Biopsy
is required for diagnosis. Secondary to the
nonspecificity of the symptoms and the elusive
intravascular location of the disease, the diagnosis
is frequently made at autopsy. Radiologic findings
in the brain (Fig. 12) are strikingly similar to
vasculitis [72] and may be partially reversible
with treatment.
Lymphomatoid granulomatosis is a rare
Epstein–Barr virus associated lymphoproli-
ferative disease characterized by the presence of
T cells and EBV-infected B cells that produce an
angiocentric and angiodestructive pattern. It typi-
cally affects the lungs but has been associated
with the central nervous system, skin, and kid-
neys. Similar to intravascular B-cell lymphoma,
the appearance may mimic vasculitis.
Reversible Cerebral Vasoconstrictive
Syndrome
Reversible cerebral vasoconstrictive syndrome
(RCVS) [73–75] is a disorder characterized by
severe acute onset (“thunderclap”) headache,
peaking within 1–2 min, with or without focal
31 Imaging of the Neurovasculitides
Fig. 12 Intravascular B-cell lymphoma. 49-year-old male
with new onset memory difficulties. Patient was referred for
evaluation of demyelinating disease. MRI FLAIR and T2
diffusion imaging (top and bottom respectively) demon-
strate extensive abnormal hyperintense signal within the
bilateral hemispheric white matter, cortex, and splenium of
the corpus callosum (black arrows) without mass effect. A
focal area of cortical/leptomeningeal high attenuation on CT
with corresponding post-gadolinium enhancement on MRI
T1-weighted imaging was noted (b left and right respec-
tively) and retrospectively represented a region of lympho-
matous involvement. Other areas of abnormal CT
attenuation and abnormal MRI signal intensity were
non-enhancing (black-outlined arrow). Findings were not
felt to be consistent with demyelinating disease. Restricted
diffusion raised concern for multifocal, acute, and chronic
ischemia. Thromboembolic disease was considered most
711
likely as there is no predilection for a single vascular terri-
tory and the distribution is sporadic, affecting both cortical
and deep perforating vascular territories. An alternative
diagnosis of infection was considered as well. Systemic
evaluation and CSF analysis were unrevealing and concern
for PACNS was raised. CTA (c) reveals multifocal, discrete,
eccentric luminal narrowing and irregularity affecting the
MCA and ACA distal vasculatures (white arrows). Normal
intervening vascular segments are noted. CA (d) confirms
multifocal luminal narrowing within the distal vasculature.
The proximal vessels and branch points are normal. There is
no luminal dilatation. Suspicion for vasculitis was height-
ened. Brain biopsy revealed intravascular B-cell lymphoma.
Tumor invasion is rare but must always be included in the
differential diagnosis when the neurovasculitides are con-
sidered (Case courtesy of Adam Davis, MD, New York
University Langone Medical Center, New York)
712
Fig. 13 Reversible vasoconstrictive syndrome. 31-year-
old female with history of schizophrenia treated with
multiple psychotropic medications including a selective
serotonin reuptake inhibitor (SSRI). The patient presented
with acute-onset severe headache and right hemiparesis,
slowly improving. MRI FLAIR imaging on presentation
(a) demonstrates multifocal abnormal hyperintense signal
within the bilateral hemispheric white matter, more prom-
inent in the parietal and occipital lobes where it extends to
the cortex. T2 diffusion imaging (b) demonstrates
restricted diffusion consistent with acute ischemia. The
white matter distribution within the left hemisphere strad-
dles the ACA, MCA, and PCA vascular territories, a
“watershed” distribution. The cortical ischemia is
A.J. Davis
consistent with a PCA parieto-occipital artery distribution.
MRA at admission (c) demonstrates short-segment
multifocal narrowing within the distal bilateral vertebral
arteries, the basilar artery, and the bilateral MCA and PCA
vasculatures (white arrows). CTA demonstrates moderate
narrowing within the right PCA P2 segment and a more
severe narrowing distally within the P3 parieto-occipital
segment (d, left, black arrows). Mild narrowing is present
within the ACA A1 segment (d, right, black arrow). CA
confirms multifocal narrowing within the bilateral PCAs
(e, black arrows). The systemic evaluation including lum-
bar puncture was normal. The patient demonstrated rapid
and near-complete clinical improvement with only limited
steroid treatment. Follow-up CTA (e, at presentation on the
31 Imaging of the Neurovasculitides
neurologic symptoms or seizures. Vascular imag-
ing demonstrates abnormal multifocal segmental
vascular constriction and dilatation that resolve
completely within 3 months (Fig. 13). It is asso-
ciated with cortical subarachnoid hemorrhage and
ischemia – including TIA and infarction, intrapar-
enchymal hemorrhage, and posterior reversible
encephalopathy syndrome. The abrupt headache
is an essential component of the diagnosis and
helps to differentiate the disease from inflamma-
tory vasculopathies, which typically have a more
subacute onset of headache. The patient median
age is 45 years old and it occurs in females more
than males. Some cases are spontaneous. Many
occur in the postpartum period or are secondary to
vasoactive substances, the most common being
cannabis, serotonin reuptake inhibitors, and
nasal decongestants. Numerous other etiologies
are cited. The presence of neurologic deficits
varies widely by study, but when permanent,
they are due to infarction or hemorrhage. Hemor-
rhage and posterior reversible encephalopathy
tend to occur early, within days of the illness
onset, while ischemia may be more delayed,
occurring a week or so after the onset of headache.
Pathologically, it is differentiated from
neurovasculitis as the abnormal vasoconstriction
is not inflammatory and is presumed to be second-
ary to a loss of control of vasomotor tone.
In one study of 67 patients with RCVS [74],
MRA revealed multifocal segmental arterial
narrowing in 88 % of patients at a mean of
8 days following the onset of headache. It is
important to note that a small percentage of
patients did not demonstrate vascular narrowing
until two weeks post the onset of symptoms. Mean
flow by transcranial Doppler typically increases
Fig. 13 (continued) left, 5-month follow-up on the right)
reveals complete resolution of the original findings. Note
the normal caliber and contour of the vertebral, basilar,
superior cerebellar, and posterior cerebral arteries on the
follow-up exam (white arrows). Severe (thunderclap)
headache is the most common presentation of RVCS;
focal neurologic signs and symptoms are common. Infarc-
tion is more unusual. A posterior frontal, parieto-occipital,
or temporo-occipital junctional zone distribution on
713
for several weeks following the onset of headache,
indicating a progressive worsening of stenosis
even when headache resolves. This finding corre-
lates with the somewhat delayed appearance of
ischemia. Angiography demonstrates diffuse,
multisegmental, arterial narrowing in all patients,
typically with intervening regions of dilatation
(“string of bead”) or normal luminal diameter. It
may be difficult or impossible to distinguish this
appearance from vasculitis, particularly PACNS,
although the inflammatory conditions may some-
times appear more isolated, asymmetric, and
eccentric with areas of vessel occlusion. The hall-
mark of the disease is the complete remission of
vascular narrowing within 3 months, an arbitrary
milestone but consistent with the literature.
CT and MRI reveal cortical subarachnoid hem-
orrhage most frequently (22 %) followed by
intraparenchymal hemorrhage (6 %) and infarc-
tion, often with a vascular watershed distribution
(4 %). Bilateral symmetric FLAIR signal
hyperintensity (10 %), usually in a posterior cir-
culation distribution or involving the occipital/
temporal–parietal/posterior frontal lobe distribu-
tion, may correlate with the associated posterior
reversible encephalopathy.
Summary
Neurovasculitis is a term used to describe a
diverse spectrum of diseases characterized by
inflammation of the blood vessels that may pro-
gress to ischemic injury of the central or periph-
eral nervous system. The overlapping clinical and
radiographic features and the complex pathophys-
iology make it difficult to quickly or clearly
imaging is a common appearance. RVCS is frequently
secondary to medication or drug use and SSRI have been
implicated. The resolution of vascular imaging abnormal-
ities within 3 months is an essential criterion of the diag-
nosis of RVCS; it would be a very rare outcome with
PACNS, the leading alternative diagnosis (Case courtesy
of Adam Davis, MD, New York University Langone Med-
ical Center, New York)
714
identify neurovasculitis as the etiology of a
patient’s symptoms. Nonetheless, simple yet com-
prehensive categorization schemes established by
consensus meetings and based upon the anatomy,
pathophysiology, and imaging findings help to
clarify the situation. For both the radiologist and
the treating clinician, a systemic approach consid-
ering the triad of (1) organ systems involved,
(2) laboratory results, and (3) vessel size and
type affected simplifies the situation.
Vessel size is particularly important for the
radiologist to consider as it broadly categorizes
the neurovasculitides within the differential diag-
nosis. Current systemic classification systems fail
to adequately describe the distribution and pathol-
ogy of the neurovasculitides as they tend to disre-
gard the unique characteristics of brain
vasculature. A more specific cerebrovascular clas-
sification is more helpful, which roughly divides
the anatomic involvement of the neurovas-
culitides into the following three areas:
1. Extracranial elastic, extracranial large vessel,
and intracranial skull base large-vessel
involvement
2. Intracranial medium-vessel involvement
3. Intracranial small-vessel involvement
This approach is most helpful for the radiolo-
gist who typically is able to separate the radio-
graphic findings into broad vascular anatomic
categories, depending on the appearance of the
vessels and the anatomic distribution of the imag-
ing abnormalities. Even so, there is considerable
anatomic overlap and the radiologic appearance
of the neurovasculitides may be similar. In these
cases, clinical criteria help and the radiologist
should consider the patient’s laboratory values or
the other organ systems involved as a means to
narrow the differential diagnosis.
Imaging findings of neurovasculitis may be
similar to the more common noninflammatory
vasculopathies including atherosclerosis or
uncommon vasculopathies, including reversible
cerebral vasoconstrictive syndrome. In this cir-
cumstance, consideration of the anatomic distri-
bution, laboratory values, and other organs
A.J. Davis
involved, as well as the patient demographics,
clinical presentation, and clinical course, helps to
differentiate inflammatory from noninflammatory
vasculopathies.
Imaging plays an essential role in the diagnosis
of these diseases. All modalities contribute to the
analysis of neurovasculitis; conventional catheter
angiography is no longer the preeminent diagnos-
tic standard. The radiologist should be familiar
with the findings of a given disease process using
all the available imaging modalities as the infor-
mation is complimentary. When all imaging find-
ings are taken together, they provide a more
comprehensive analysis and a more certain diag-
nosis. Imaging may be divided into those studies
that investigate the vessel lumen, the vessel wall,
and the parenchyma. Parenchymal findings are the
least specific but are usually necessary to detect the
presence of disease as well as follow progression
or remission. Vessel lumen abnormalities are
highly suggestive for vasculitis when present but
are not specific and the sensitivity is low for the
small-vessel vasculitides. Demonstration of vessel
wall or perivascular inflammatory change is the
most specific and compelling radiographic find-
ing, although it is only inconsistently found likely
due to the limitations of routine imaging protocols.
Nonetheless, direct findings of mural inflamma-
tory change should be aggressively sought by the
radiologist. MRI and nuclear medicine imaging
provide the most sensitive evaluation for vessel
wall inflammation although extracranial large elas-
tic vessel inflammation is seen particularly well
with CT/CTA.
Biopsy is frequently required, not only to con-
firm the diagnosis but, in half the positive cases, to
provide an alternative diagnosis. Imaging should
be used to guide the site of surgical biopsy.
Despite the uncommon occurrence of
neurovasculitis, it is important to be familiar
with these diseases. They play a particularly
important role in the differential diagnosis of
stroke in young and middle-aged adults and typ-
ically have devastating clinical consequences. If
diagnosed early, they may be effectively treated,
relieving symptoms and avoiding permanent cen-
tral and peripheral neurologic damage.
31 Imaging of the Neurovasculitides
References
1. Giannini C, Salvarani C, Hunder G, Brown RD (2012)
Primary central nervous system vasculitis: pathology
and mechanisms. Acta Neuropathol 123:759–772
2. Marsh EB, Zeiler SR, Levy M, Llinas RH, Urrutia VC
(2012) Diagnosing CNS Vasculitis: The Case Against
Empiric Treatment. Neurologist 18:233–238
3. Weyand CM, Goronzy JJ (2003) Giant-Cell Arteritis
and Polymyalgia Rheumatica. Ann Intern Med
139:505–515
4. Hotchi M (1992) Pathological studies on Takayasu
Arteritis. Heart Vessels Suppl 7:11–17
5. Watts R, Al-Taiar A, Mooney J, Scott D, Macgregor A
(2009) The epidemiology of Takayasu Arteritis in the
UK. Rheumatology 48:1008–1011
6. Adams HP, Kapelle J, Biller J et al (1995) Ischemic
stroke in young adults. Iowa Registry of Stroke in
Young Adults. Arch Neurol 52:491–495
7. Gemmete JJ, Davagnanam I, Toma AK, Brew S,
Ganesan V (2013) Arterial ischemic stroke in children.
Neuroimaging Clin of N Am 23:781–798
8. Hunder GG, Arend WP, Bloch DA, Calabrese LH,
Fauci AS, Fries JF et al (1990) The American College
of Rheumatology 1990 criteria for the classification of
vasculitis. Introduction. Arthritis Rheum
33:1065–1067
9. Jennette JC, Falk RJ, Andrassy K, Bacon PA, Churg J,
Gross WL et al (1994) Nomenclature of systemic vas-
culitides. Proposal of an international consensus con-
ference. Arthritis Rheum 37:187–192
10. Kuker W (2007) Cerebral vasculitis: imaging signs
revisited. Neuroradiology 49:471–479
11. Kaufman TJ, Kallmes DF (2008) Diagnostic cerebral
angiography: archaic and complication-prone or here
to stay for another 80 years? AJR Am J Roentgenol
190:1435–1437
12. Hajj-Ali RA, Calabrese LH (2013) PACNS.
Autoimmun Rev 12:463–466
13. Lie JT (1996) Vasculitis associated with infectious
agents. Curr Opin Rheumat 8:26–29
14. Calabrese LH, Mallek JA (1988) PACNS. Report of
8 new cases, review of the literature and proposal for
diagnostic criteria. Medicine 67:20–39
15. Chu CT, Gray L, Goldstein LB, Hulette CM (1998)
Diagnosis of intracranial vasculitis: a multi-disciplinary
approach. J Neuropathol Exp Neurol 57:30–38
16. Miller DV, Salvarani C, Hunder GG, Brown RD, Parisi
JE, Christianson TJ, Giannini C (2009) Biopsy findings
in PACNS. Am J Surg Pathol 33:35–43
17. Hankey G (1991) Isolated angiitis/angiopathy of the
CNS. Prospective diagnostic and therapeutic experi-
ence. Cerebrovasc Dis 1:2–15
18. Alrawi A, Trobe J, Blaivas M, Musch DC (1999) Brain
biopsy in PACNS. Neurology 53:858–860
19. Moharir M, Shroff M, Benseler S (2013) Childhood
central nervous system vasculitis. Neuroimaging Clin
N Am 23:293–308
715
20. Joseph FG, Scolding NJ (2002) Cerebral vasculitis: a
practical approach. Pract Neurol 2:80–93
21. Neel A, Pangnoux C (2009) PACNS. Clin Exp
Rheumatol 27(Suppl 52):S95–S107
22. Lie JT (1997) Biopsy diagnosis of systemic vasculitis.
Bailliere Clin Rheumatol 11:219–236
23. Rossi CM, Comite G (2009) The clinical spectrum of
the neurological involvement in vasculitides. J Neurol
Sci 285:13–21
24. Adams HP (2014) Cerebral vasculitis. In: Biller J,
Ferro JM (eds) Neurologic aspects of systemic disease
part I. Handbook of clinical neurology. vol 119 (3rd
series). Elsevier BV
25. Kuker W (2007) Imaging of cerebral vasculitis. Int J
Stroke 2:184–190
26. Gomes LJ (2010) The role of imaging in the diagnosis
of central nervous system vasculitis. Curr Allergy
Asthma Rep 10:163–170
27. Zuccoli G, Pipitone N, Haldipur A, Brown RD Jr,
Hunder G, Salvarani C (2007) Imaging findings in
primary central nervous system vasculitis. Clin Exp
Rheumatol 29(Suppl 64):S104–S109
28. Salvarani C, Brown RD Jr, Calamia KT et al (2007)
Primary central nervous system vasculitis: analysis of
101 patients. Ann Neurol 62:442–451
29. Wasserman BA, Stone JH, Hellman PMG (2001) Reli-
ability of normal findings on MR imaging for exclud-
ing the diagnosis of vasculitis of the central nervous
system. AJR Am J Roentgenol 177:455–459
30. Pomper M, Miller T, Stone J, Tidmore W, Hellman D
(1999) CNS vasculitis in autoimmune disease: MR
imaging findings and correlation with angiography.
AJNR Am J Neuroradiol 20:75–85
31. Koçer N, Islak C, Siva A, Saip S, Akman C,
Kantarci O, Hamuryudan V (1999) CNS involvement
in Neuro-Behçet syndrome: an MR study. AJNR Am J
Neuroradiol 20:1015–1024
32. Akman-Demir G, Serdaroglu P, Tasçi B (1999) Clinical
patterns of neurological involvement in Behçet’s dis-
ease: evaluation of 200 patients. Brain 122:2171–2182
33. Spitzer C, Mull M, Rohde V, Kosinski CM (2005)
Non-traumatic cortical subarachnoid haemorrhage:
diagnostic work-up and etiological background. Neu-
roradiology 47:525–531
34. Ay H, Sahin G, Saatci I, Soylemezoglu F, Saribas O
(2002) PACNS and silent cortical hemorrhages. AJNR
Am J Neuroradiol 23:1561–1563
35. Salvarani C, Brown RD Jr, Gene G (2012) Adult pri-
mary central nervous system vasculitis. Lancet
380:767–777
36. K€uker W, Gaertner S, Nägele T, Dopfer C,
Schöninge M, Jens Fiehler J, Rothwell PM, Herrlinger
U (2008) Vessel wall contrast enhancement: a diagnos-
tic sign of cerebral vasculitis. Cerebrovasc Dis
26:23–29
37. Kumar RS, Singh A, Rathore C, Kesa-Vades C (2010)
Primary angiitis of central nervous system: tumor-like
lesion. Neurol India 58:147–149
716
38. Panchal NJ, Niku S, Imbesi SG (2005) Lymphocytic
vasculitis mimicking aggressive multifocal cerebral
neoplasm: MR imaging and MR spectroscopic appear-
ance. AJNR Am J Neuroradiol 26:642–645
39. Muccio CF, DiBlasi A, Espositio G, Brunese L,
D’Arcos F, Caranc F (2013) Perfusion and spectros-
copy magnetic resonance imaging in a case of lympho-
cytic vasculitis mimicking brain tumor. Pol J Radiol
78:66–69
40. Qu SB, Khan SW, Hua L (2009) Primary central ner-
vous system vasculitis mimicking brain tumour: case
report and literature review. Rheumatol Int 30:127–134
41. Park MS, Marlin AE, Gaskill SJ (2013) Angiography-
negative PACNS in childhood. J Neurosurg Pediatr.
doi:10.3171/2013.9.PEDS13159. Published online
25 Oct 2013
42. Sundgren PC, Jennings J, Attwood JT et al (2005) MRI
and 2D-CSI MR spectroscopy of the brain in the eval-
uation of patients with acute onset of neuropsychiatric
systemic lupus erythematosus. Neuroradiology
47:76–585
43. Woolfenden AR, Tong DC, Marks MP, Ali AO, Albers
GW (1998) Angiographically defined PACNS: is it
really benign? Neurology 51:183–188
44. Harris KG, Tran DD, Sickels WJ, Cornell SH, Yuh WT
(1994) Diagnosing intracranial vasculitis: the roles of
MR and angiography. AJNR Am J Neuroradiol
15:317–330
45. Greenan TJ, Grossman RI, Goldberg HI (1992) Cere-
bral vasculitis: MR imaging and angiographic correla-
tion. Radiology 182:65–72
46. Duna GF, Calabrese LH (1995) Limitations of invasive
modalities in the diagnosis of PACNS. J Rheumatol
22:662–667
47. Imbesi SG (1999) Diffuse cerebral vasculitis with nor-
mal results on brain MR imaging. AJR Am J
Roentgenol 173:1494–1496
48. Brouwer PA, Bosman T, van Walderveen MAA,
Krings T, Leroux AA, Willems PWA (2010) Dynamic
320-section CTA in cranial arteriovenous shunting
lesions. AJNR Am J Neuroradiol 31:767–770
49. Birnbaum J, Hellman DB (2009) PACNS. Arch Neurol
66:704–709
50. Cloft HJ, Phillips CD, Dix JE, McNulty BC, Zargado
MT, Kallmes DF (1999) Correlation of angiography
and MR imaging in cerebral vasculitis. Acta Radiol
40:83–87
51. Hellmann DB, Roubenoff R, Healy RA, Wang H
(1992) Central nervous system angiography: safety
and predictors of a positive result in 125 consecutive
patients evaluated for possible vasculitis. J Rheumatol
19:568–572
52. Khandelwal N, Kalra N, Garg MK, Kang M, Lal A,
Jain S, Suri S (2009) Multidetector CT angiography in
Takayasu’s arteritis. Eur J Radiol 77:369–374
53. Yamada I, Nakagawa T, Himeno Y, Numano F,
Shibuya H (1998) Takayasu’s arteritis: evaluation of
the thoracic aorta with CTA. Radiology 209:103–109
A.J. Davis
54. Demaerel P, De Ruyter N, Maes F, Velghe B, Wilms G
(2004) Magnetic resonance angiography in suspected
cerebral vasculitis. Eur Radiol 14:1005–1012
55. Park JH (1996) Conventional and CT angiographic
diagnosis of Takayasu’s arteritis. IntJ Cardiol 54:
SI65–SI71
56. Choe YH, Han BK, Koh EM, Kim DK, Do YS, Lee
WR (2000) Takayasu’s arteritis: assessment of disease
activity with contrast-enhanced MR imaging. AJR Am
J Roentgenol 175:505–551
57. Bley TA, Uhl M, Carew J, Markl M, Schmidt D, Peter
HH, Lander M, Wieben M (2007) Diagnostic value of
high-resolution MR imaging in giant cell arteritis.
AJNR Am J Neuroradiol 28:1722–1727
58. Cheng Y, Lv N, Wang Z, Chen B, Dang A (2013)
18F-FDG-PET in assessing disease activity in
Takayasu’s arteritis: a meta-analysis. Clin Exp
Rheumatol 31:S22–S27
59. Besson FL, Arienti JJ, Bienvenu B et al (2011) Diag-
nostic performance of F18 fluorodeoxyglucose posi-
tron emission tomography in giant cell arteritis: a
systematic review and meta-analysis. Eur J Nucl Med
Mol Imaging 38:1764–1772
60. Amezcua-Guerra LM, Pineda C (2007) Imaging stud-
ies in the diagnosis and management of vasculitis. Curr
Rheumatol Rep 9:320–327
61. Direskeneli H, Aydın SZ, Merkel PA (2011) Assess-
ment of disease activity and progression in Takayasu’s
arteritis. Clin Exp Rheumatol 29:S86–S91
62. Blockmans D, DeCeuninck L, Vander-Schueren S,
Knockaert D, Mortel-Mans L, Bobbaers H (2006)
Repetitive 18F-fluorodeoxyglucose positron emission
tomography in giant cell arteritis: a prospective study
of 35 patients. Arthritis Rheum 55:131–137
63. Karapolat I, Kalfa M, Keser G, Yalçin M, Inal V,
Kumanlioglu K, Pirildar T, Aksu K (2013) Comparison
of 18F-FDG PET/CT findings with current clinical
disease status in patients with Takayasu’s arteritis.
Clin Exp Rheumatol 31:S15–S21
64. Schmidt WA, Nerenheim A, Seipelt E, Poehls C,
Gromnica-Ihle E (2002) Diagnosis of early Takayasu’s
arteritis with sonography. Rheumatology 41:496–502
65. Gotway M, Araoz PA, Macedo TA, Stanson AW
et al (2004) Imaging findings in Takayasu’s arteritis.
AJR Am J Roentgenol 184:1945–1950
66. Kissin EY, Merkel PA (2004) Diagnostic imaging in
Takayasu’s arteritis. Curr Opin Rheumatol 16:31–37
67. Seth S, Goyal NK, Jagia P, Gulati G, Karthikeyan G,
Sharma S, Talwar KK (2006) Carotid intima–medial
thickness as a marker of disease activity in Takayasu’s
arteritis. Int J Cardiol 108:385–390
68. Schmidt WA, Kraft HE, Vorpahl K, Volker L, Erika J,
Gromnica-Ihle EJ (1997) Color duplex ultrasonogra-
phy in the diagnosis of temporal arteritis. New Engl J
Med 337:1336–1342
69. Pfadenhauer K, Weber H (2003) GCA of the occipital
arteries: a prospective color coded duplex sonography
study in 78 patients. J Neurol 250:844–849
31 Imaging of the Neurovasculitides
70. Karassa FB, Matsagas MI, Schmidt WA, Ioannidis JP
(2005) Meta-analysis: test performance of ultrasonog-
raphy for giant cell arteritis. Ann Intern Med
142:359–369
71. Néel A, Auffray-Calvier E, Guillon B, Fontenoy AM
et al (2012) Challenging the diagnosis of PACNS: a
single-center retrospective study. J Rheumatol
39:1026–1034
72. Orwat DE, Batalis NI (2012) Intravascular large B-cell
lymphoma. Arch Pathol Lab Med 136:333–338
73. Calabrese LH, Dodick DW, Schwedt TJ, Singhal AB
(2007) Narrative review: reversible cerebral vasocon-
striction syndromes. Ann Intern Med 146:34–44
74. Ducros A, Boukobza M, Porcher R, Sarov M,
Valade D, Bousser MG (2007) The clinical and radio-
logical spectrum of reversible cerebral vasoconstriction
717
syndrome. A prospective series of 67 patients. Brain
130:3091–3101
75. Ducros A, Bousser MG (2009) Reversible cerebral
vasoconstriction syndrome. Pract Neurol 9:256–267
76. Younger DS (2004) Vasculitis of the nervous system.
Curr Opin Neurol 17:317–336. Lippincott Williams &
Wilkins
77. Mohammad AJ, Jacobsson LT, Westman KW
et al (2009) Incidence and survival rates in Wegener’s
granulomatosis, microscopic polyangiitis,
Churg–Strauss syndrome and polyarteritis nodosa.
Rheumatology 48:1560–1565
78. Kobayashi Y, Ishii K, Oda K et al (2005) Aortic wall
inflammation due to TA imaged with 18F-FDG PET
co-registered with enhanced CT. J Nucl Med
46:917–922
 Miscellaneous Vascular Malformations
(Cavernous Malformations, 32
Developmental Venous Anomaly,
Capillary Telangiectasia, Sinus Pericranii)

Luiz Celso Hygino da Cruz and Cintia Elias Pires

Contents Capillary Telangiectasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . 741

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720
Cerebral Cavernous Malformations . . . . . . . . . . . . . . 720
Clinical Manifestations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723
Histological Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725
Neuroimaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 726
Differential Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 733
Treatment and Final Considerations . . . . . . . . . . . . . . . . . 733
Developmental Venous Anomaly . . . . . . . . . . . . . . . . . . 733
Clinical Manifestations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 736
Histological Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 739
Neuroimaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 739
Treatment and Final Considerations . . . . . . . . . . . . . . . . . 741
Clinical Manifestations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 742
Histological Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 742
Neuroimaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 742
Treatment, Differential Diagnosis, and Final
Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 743
Sinus Pericranii . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 744
Clinical Manifestations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 744
Neuroimaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745
Treatment, Differential Diagnosis, and Final
Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747

L.C.H. da Cruz (*)

Department of Radiology, Clinics CDPI/DASA and IRM,
Federal University of Rio de Janeiro, Rio de Janeiro, Brazil
e-mail: celsohygino@hotmail.com
C.E. Pires
Department of Radiology, Clinics CDPI/DASA, National
Institute of Cardiology, Federal University of Rio de
Janeiro, Rio de Janeiro, Brazil
e-mail: cintiaepires@yahoo.com.br

# Springer Science+Business Media New York 2016 719

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_8
720 L.C.H. da Cruz and C.E. Pires

malformations without arteriovenous shunting

Abstract
[1, 2]. This latter group includes conditions of
Vascular malformations involving the brain are
miscellaneous etiologies such as venous anoma-
uncommon lesions that are seen increasingly
lies (cavernous hemangioma or cavernoma and
more often in clinical practice, primarily
developmental venous anomaly), capillary anom-
because of advances in cross-sectional brain
alies (capillary telangiectasia), and sinus
imaging. They are typically divided into two
pericranii, a special condition considered as a
main categories: arteriovenous malformations
venous malformation. Some can be angiographic-
and vascular malformations without arteriove-
ally occult, like cavernomas and capillary telangi-
nous shunting, including cavernous hemangi-
ectasias [3]. All these lesions must be
oma, developmental venous anomaly, capillary
distinguished from one another due to their differ-
telangiectasia, and sinus pericranii. This latter
ent natural histories and various treatment strate-
group is the topic of interest of this chapter.
gies. As the main focus of this chapter, these
Although all these lesions are clinically rele-
conditions are and will be discussed in detail. In
vant, their natural history and prognosis
this chapter, we attempt to provide a concise
depend on several factors, including lesion
review of the imaging methods utilized in the
type, location, size, and associated conditions.
clinical diagnosis and treatment of the cerebral
Understanding the associated imaging findings
malformations listed above.
and potential complications of these lesions
Combined with digital subtraction angiogra-
assists in determining the appropriate treatment
phy (DSA), cross-sectional imaging, including
options. In this chapter, we attempt to provide a
multiplanar computed tomography (CT) and mag-
concise review of the imaging methods used in
netic resonance imaging (MRI), plays an essential
the clinical diagnosis and treatment of the cere-
role not only in the differential diagnosis of brain
bral malformations listed above.
malformations but also in the long-term follow-up
monitoring of patients with lesions that do not
Keywords
require immediate intervention. It is well
Cerebral vascular malformations • Cavernous
established that MRI is the best noninvasive imag-
hemangioma • Cavernoma • Capillary telangi-
ing modality for the detection of a vascular mal-
ectasia • Developmental venous anomaly •
formation. In certain clinical settings and
Sinus pericranii • SWI
pathologies, it has become apparent that MRI
alone could be sufficient to evaluate some types
of vascular malformations and play an important
Introduction role in treatment management. In addition, it is
extremely helpful in presurgical planning to
Vascular malformations of the brain are uncom- assess the extent of the lesion, define borders,
mon lesions that manifest by imaging tests as and plan the surgical approach, based on the
abnormal vessels in the brain; they are clinically extent of the lesion and its relationship with sur-
relevant since clinical outcome may include hem- rounding areas and localization of eloquent corti-
orrhagic stroke, epilepsy, and focal neurologic cal areas.
deficits. Their pathogenesis involves abnormal
vasculogenesis and/or angiogenesis. The evalua-
tion of a vascular malformation of the brain may Cerebral Cavernous Malformations
be a diagnostic challenge to radiologists due to
their similar clinical manifestations and imaging Cerebral cavernous malformations (CCMs), also
features. known as cavernous angiomas, cavernous heman-
Vascular brain lesions are typically divided giomas, or simply cavernomas, are common cere-
into two main categories: brain arteriovenous bral vascular malformations consisting of clusters
malformations (AVMs) and vascular of deformed and dilated vessels, with little or no
32 Miscellaneous Vascular Malformations (Cavernous Malformations, Developmental Venous. . .
Fig. 1 Right temporal lobe cavernous malformation. (a)
T1-weighted MR. (b) Postcontrast T1-weighted MR. (c)
FLAIR image. (d) Gradient-echo (GRE) T2*-weighted
MR. (e) Diffusion-weighted MR. The lesion presents the
characteristic “popcorn” magnetic resonance appearance,
secondary to blood components in different stages of deg-
radation (a–c). A hypointense rim is well demonstrated on
intervening brain parenchyma, lined by endothe-
lium and filled with thrombosed blood at various
stages [4, 5]. These lesions usually have a charac-
teristic appearance on MRI. Modern MR imaging
sequences are highly sensitive for detecting
CCMs as well as associated hemorrhage at vari-
ous stages of thrombosis and reorganization
(Fig. 1). CCM is combined with capillary telangi-
ectasia and thrombosed arteriovenous malforma-
tion, under the term occult cerebrovascular
malformation, since none of these lesions is dem-
onstrated by conventional angiography [3].
CCMs are considered one of the four most
common types of vascular malformations,
accounting for approximately 8–16 % of all cere-
brovascular malformations. Their estimated prev-
alence on the basis of MR imaging findings is up
to 0.4–0.5 % of the population [6]. Two forms
have been described: a sporadic form, in which
patients usually have an isolated lesion, and a
721
FLAIR images (c). A subtle contrast enhancement can be
seen after intravenous gadolinium administration (b). GRE
T2* weighted confirms the hemorrhagic component of the
lesion. Subacute hemorrhagic foci within the lesion can
demonstrate restricted diffusion, due to the presence of
methemoglobin (e)
familial form, usually characterized by multiple
lesions, with an autosomal dominant pattern of
inheritance [7]. Typically, patients with a single
lesion have a sporadic form of the disease, while
those with multiple lesions (10–30 % of all cases)
often have an autosomal dominant form localiz-
able to the CCM1/KRIT1, CCM2/MGC4607, or
CCM3/PDCD10 gene loci. Genetic linkage ana-
lyses mapped three CCM loci to chromosome 7q
(CCM1), 7p (CCM2), and 3q (CCM3) [8]. The
hallmark of familial CCM is the presence of
multifocal lesions throughout the brain with the
appearance of new lesions over time (Fig. 2).
However, in about 20 % of individuals with mul-
tiple lesions, no mutation is observed [8]. The
sporadic form of CCM is often characterized by
a solitary lesion (or a cluster of lesions). There is a
high association between CCM and DVA (Fig. 3).
This association with a DVA is more likely to
define the nongenetic familial form of the disease.
722
Fig. 2 Multiple cavernous malformations in a patient with
familial form. Axial T2-weighted (a) and GRE T2*-
weighted (b) images demonstrate a CCM lesion in the
Familial CCMs are unlikely to be associated with
DVA, while sporadic forms have a high associa-
tion rate with this vascular malformation [9]. Even
though the multiplicity of lesions is characteristic
of the familial form, it has been reported in 10 %
to up to 33 % of supposed sporadic cases. Addi-
tionally, they are commonly seen after therapeutic
irradiation of the brain (Fig. 4), along with capil-
lary telangiectasias in the radiotherapy bed [10].
L.C.H. da Cruz and C.E. Pires
left frontal lobe. Note multiple small foci hypointense
lesions on GRE T2*-weighted images (b–d, arrows)
consisting with small cavernous angiomas
Although CCM can occur in virtually all parts
of the central nervous system, they are much more
common in the cerebral hemispheres, followed by
the infratentorial compartment. A superficial cere-
bral hemisphere location with proximity to the
subarachnoid space (Fig. 5) and the ventricle is
very common. Infratentorial involvement corre-
sponds roughly to one fourth of the cases.
Although the brain stem and the cerebellum
32 Miscellaneous Vascular Malformations (Cavernous Malformations, Developmental Venous. . .
Fig. 3 Cavernous malformation with associated develop-
mental venous anomaly. Axial postcontrast T1-weighted
(a) and FLAIR images (b) and coronal T2-weighted (c)
image demonstrate an heterogeneous signal intensity
hemispheres are equally affected, the pons is the
most common location of the brain stem (Fig. 6).
CCM can also rarely occur in the spinal cord
(Fig. 7), where it frequently coexists with multiple
brain lesions (Fig. 8) [11, 12].
Extra-axial lesions are extremely atypical and
most lesions arise from the leptomeninges or the
cranial nerves (Fig. 9). They have been described
in the cerebellopontine angle, intraventricularly,
in the pituitary fossa, optic chiasm, and cavernous
sinus. Extra-axial cavernous lesions usually dem-
onstrate atypical findings on cross-sectional imag-
ing, including hyperintense signal intensity on
723
lesion, hypointense on GRE T2* weighted, without mass
effect, in the right cerebellum peduncle, adjacent to the
fourth ventricle. There is a developmental venous anomaly
adjacent to the CCM (a)
T2-weighted sequences and intense homogenous
enhancement. Dural-based cavernomas in associ-
ation with dural sinuses have also been described
[13–15].
Clinical Manifestations
Most lesions are found incidentally in neuroimag-
ing studies and may remain asymptomatic
throughout life. They can be discovered by an
episode of symptomatic hemorrhage or gradual
and nonspecific symptoms can be seen. Usually,
724
Fig. 4 A 15-year-old male with germinoma was treated
with surgery followed by radiation therapy and chemother-
apy. FLAIR (a) and SWI (b) obtained 2 years after the
treatment show postsurgical changes in the right frontal
lobe, and no lesion is seen in the left occipital lobe. Sagittal
patients clinically present with CCMs between the
fourth and sixth decades of life, but symptoms can
start early in life. Common clinical manifestations
are seizures, focal neurological deficits, and hem-
orrhage (Fig. 10) [11, 12]. Symptoms related to
mass effect and headache can also occur.
When clinically apparent, seizures are reported
as the most common symptom, occurring in
40–60 % of the symptomatic patients, and are
the most frequent manifestation of the
supratentorial lesions. Seizures are most often
focal, can be refractory to treatment, and are
thought to be related to cortical involvement asso-
ciated with irritative effect of hemorrhage and the
L.C.H. da Cruz and C.E. Pires
T1-weighted (c), axial FLAIR (d), SWI (e), and SWI phase
(f) images obtained 4 years after the radiation therapy show
a round lesion of mixed signal intensity with peripheral
hypointense rim in the left occipital lobe, suggestive of
cavernous angioma
inflammatory reaction, as well as to some degree
of cortex compression. A long-standing seizure
without any acute episode of hemorrhage can be
frequently observed.
Focal neurologic deficits can occur in 10–45 %
of the symptomatic patients. It is a rare condition,
mostly seen in infratentorial involvement.
Progressive deficits are more related to lesion
growth and less commonly to an episode of hem-
orrhage. The enlargement of the lesion is more
likely to be caused by chronic or recurrent extrav-
asation of blood and thrombosis.
It is well known that small and subclinical
hemorrhages are a common occurrence.
32 Miscellaneous Vascular Malformations (Cavernous Malformations, Developmental Venous. . .
Fig. 5 Cavernous malformation with associated develop-
mental venous anomaly located at the subcortical region of
the left frontal lobe. (a) Postcontrast T1-weighted MR. (b)
FLAIR MR. (c) GRE T2*-weighted MR. The lesion
Symptomatic hemorrhage is a rare condition. The
risk of a hemorrhage from a CCM is thought to
range from 0.1 % to 0.7 % per year and can reach
up to 25 % in patients with a prior history of
hemorrhage. The risk of bleeding is increased in
young and female patients with infratentorial
lesions. Seizures and acute neurological deficits
are the most frequent symptoms reported. Exten-
sion of hemorrhage to the subarachnoid space and
to the ventricles is uncommon, as well as superfi-
cial siderosis, secondary to multiple hemorrhagic
episodes [11, 12, 16].
CCMs can exhibit a wide range of behaviors,
including enlargement, regression, and de novo
formation. These lesions have also been described
to be associated with prior brain radiotherapy.
CCMs take about 1–26 years to develop after
radiation [10]. The diagnosis of cavernoma
should be considered when a new hemorrhagic
lesion appears in the radiation bed, particularly if
the radiation therapy was performed in childhood
(Fig. 4).
Histological Findings
By gross pathology, CCM presents as a well-
circumscribed, red-colored spherical or lobulated
mass lesion, with a grapelike configuration,
containing multiple thin-walled vascular
725
demonstrates the heterogeneous signal characteristics of
hemorrhage in various stages of resolution. A developmen-
tal venous anomaly is seen after the intravenous contrast
administration
structures. A rim of gliotic brain tissue usually
delimits the periphery of the CCM with a variable
degree of extralesional hemosiderin pigment from
prior hemorrhages. There is no capsule involving
the lesion [4, 5].
Upon microscopic examination, CCM has a
classic appearance of “popcorn” or a “mulberry-
like” cluster of sinusoidal blood cavities
surrounded by a single layer of endothelium,
with little or no intervening neural tissue. The
vascular structures are filled with blood, thrombi
in different stages, and organized connective tis-
sue. Adjacent brain parenchyma is usually gliotic.
Almost all CCMs have peripheral deposition of
hemosiderin, suggesting the occurrence of multi-
ple episodes of asymptomatic bleeding. Calcifica-
tions and cholesterol deposit are common findings
within them. Arteriovenous shunting is absent,
and irrigation, as well as drainage, is performed
by normal vasculature. Different from arteriove-
nous malformations, no muscular or elastic tissue
is present in vessel walls, no matter their thickness
[4, 5, 17].
Although the exact mechanism underlying the
pathogenesis of CCMs is unknown, the impor-
tance of the mechanism involved in the prolifera-
tion and differentiation of angiogenic precursors
and of members of the apoptotic mechanism has
been pointed out as a key regulator of the patho-
genesis of CCM [6].
726
Fig. 6 Cavernous malformation with associated develop-
mental venous anomaly located in the left aspect of the
pons. Axial FLAIR (a) and coronal T2-weighted MR (b)
images show a round heterogeneous lesion, with a
hypointense rim. On the axial postcontrast T1-weighted
Neuroimaging
Approximately 80 % of the CCM cases are
supratentorial intra-axial lesions. However, they
can be found anywhere in the central nervous
system, including the brain stem and spinal cord.
Extra-axial lesions have also been described.
They are usually solitary, although up to one
third of patients with sporadic lesions have more
than one lesion. In the familial form, multiple
lesions are usually seen in supratentorial and
L.C.H. da Cruz and C.E. Pires
MR (c, d) images, a developmental venous anomaly is
identified with a drainage vein at the inferior aspect of the
fourth ventricle, the left ambiens cistern, and the
quadrigeminal cistern
infratentorial compartments. Lesions vary in size
from few millimeters to several centimeters
across, with an average size of around 1.5
cm. They usually lack mass effect, unless
there have been recent episodes of bleeding. Typ-
ical lesions show little or no enhancement [7, 11,
12, 14].
Computed tomography (CT) can detect CCM
as a spontaneous high-attenuation lesion, with a
variable presence of calcification. However, CT
can be normal in around 30–50 % of the cases.
32 Miscellaneous Vascular Malformations (Cavernous Malformations, Developmental Venous. . . 727

Fig. 7 Cavernous malformation in the cervical spine. A T1-weighted image (b) is seen. On T1-weighted MR after
heterogeneous cervical spine lesion with peripheral intravenous contrast administration (c), a venous structure
hypointense rim on the T2-weighted (a) and GRE T2*- is demonstrated inferiorly to the described cavernoma
weighted (d) images and spontaneous hyperintense on (arrow)

Although CCM may be diagnosed by using CT microhemorrhage, CCM contains large amounts
scans, MR imaging is much more sensitive for of deoxyhemoglobin or hemosiderin, which gen-
demonstrating these lesions. Owing to the erate the susceptibility effect, causing marked sig-
blood stagnation phenomenon, or chronic nal loss in specific MR sequences, including
728
Fig. 8 Multiple cavernous malformations. Axial FLAIR
image (a) demonstrates a subependymal CCM in the
atrium of the right lateral ventricle. On susceptibility-
weighted images (SWI), multiple cavernomas can be
detected as hypointensity lesions (b, c). Sagittal
L.C.H. da Cruz and C.E. Pires
T2-weighted (d) and axial GRE T2*-weighted (e) images
also demonstrate a cavernoma in the posterior and right
aspect of the cervical spine, with an exophytic component
(d, e)
32 Miscellaneous Vascular Malformations (Cavernous Malformations, Developmental Venous. . .
Fig. 9 Cavernous malformation of the left trigeminal
nerve. An expansive, extra-axial lesion of mixed signal
intensity on both T2-weighted (a, b) and T1-weighted (c)
images, involving the cisternal component of the left fifth
susceptibility-weighted imaging (SWI) and GRE
sequences, particularly a T2*-weighted GRE
sequence. These sequences are highly sensitive
for the susceptibility artifact, which make them
especially useful in the detection of smaller and
concomitant cavernous lesions that may not be
detected with traditional sequences.
Thus, the sensitivity of MRI to flowing blood
and blood products of varying ages, combined with
the greater contrast resolution of MRIs, greatly
increases the specificity of MRI compared to any
729
cranial nerve, invading the ipsilateral Meckel cave. On
GRE T2*-weighted (d) image, the lesion presents areas
of low signal intensity, as well as a hypointense rim
other modality and has largely eliminated
misdiagnosis of CCM. MRI is also used in the
follow-up monitoring of patients with known
CCM and for the assessment of family members
in whom similar lesions are suspected. In addition,
it is extremely helpful in presurgical planning to
assess the extent of the lesion, define borders, and
plan the surgical approach and exposure.
Computed Tomography: CT scanning is not
sensitive for the diagnosis of CCM and can be
normal in 30–50 % of patients. When this lesion is
730
Fig. 10 Supratentorial cavernous malformation with a
recent hemorrhage. An 80-year-old man with mental con-
fusion, seizures, and right hemiparesis. Axial T1-weighted
(a) and T2-weighted (c) images demonstrate an expansive,
round, and hemorrhagic lesion in the left postcentral gyrus,
detected on CTscans, it usually appears as a rounded
slightly hyperdense lesion that may or may not have
interspersed calcifications. If there has been a recent
episode of hemorrhage, then it is more conspicuous
and may be surrounded by a mantle of edema.
Although most lesions do not enhance, faint
enhancement can occasionally be seen. CTA does
L.C.H. da Cruz and C.E. Pires
surrounded by moderate amount of edema, which also
involves the precentral gyrus. The postcontrast
T1-weighted image (b) shows mild contrast enhancement.
The susceptibility artifacts generated by the blood products
can easily be seen on GRE T2*-weighted image (d)
not bring additional information, unless there are
associated malformations, such as DVAs.
Magnetic Resonance Imaging: MRI is the
method of choice, demonstrating a characteristic
“popcorn” or “mulberry” appearance with a rim of
signal loss due to hemosiderin, better depicted on
susceptibility-weighted sequences [12, 14].
32 Miscellaneous Vascular Malformations (Cavernous Malformations, Developmental Venous. . .
Fig. 11 Cavernous malformation associated to a develop-
mental venous anomaly. (a) Axial T1-weighted MR. (b)
Coronal T2-weighted MR. (c) Coronal postcontrast
T1-weighted MR. Axial GRE T2*-weighted MR. A left
pontine and middle cerebellar peduncle CCM with the
Recently, more advanced imaging techniques
such as high-field and susceptibility-weighted
magnetic resonance imaging (SWI) have been
employed for the evaluation of CCMs. Further-
more, diffusion tensor imaging (DTI) and func-
tional magnetic resonance imaging (fMRI) have
been applied to the preoperative and
intraoperative management of these lesions.
– T1-weighted sequences can demonstrate a
round heterogeneous lesion, with a “popcorn”
or “mulberry” appearance, containing blood
components in different stages of degradation.
Deoxyhemoglobin related to acute bleeding
has intermediate signal intensity on T1W
images, whereas methemoglobin related to
731
classic “popcorn” MRI appearance, surrounded by a
hypointense rim. The lesion causes compression of the
fourth ventricle. After the intravenous contrast administra-
tion, a DVA is observed adjacent to the inferior aspect of
the CCM
subacute bleeding shows hyperintense signal.
Hypointense areas correspond to hemosiderin
deposition, related to chronic hemorrhage.
– T2-weighted sequences reveal a heterogeneous
lesion surrounded by a marked hypointense
rim, reflecting hemosiderin deposition.
– FLAIR sequences can demonstrate
perilesional edema in cases with recent epi-
sodes of bleeding.
– Since cavernomas contain large amounts of
deoxyhemoglobin (blood stagnation) or hemo-
siderin (chronic microhemorrhage), sequences
that are highly sensitive for susceptibility artifact
such as SWI and GRE T2*-weighted sequences
are extremely useful for detecting marked signal
loss within the lesions (Fig. 11). Especially for
732
Fig. 12 A 74-year-old male patient with familial CCM.
Axial T2-weighted image (a) shows areas of white matter
high signal intensity and a small focus of low signal inten-
sity on the left (arrow). T2-weighted GRE image (b) dem-
onstrates multiple foci of low signal intensity on the
multiple lesions, frequently associated with the
familial form of the disease, these techniques are
of great use. SWI seems to be more sensitive
than T2* to detect cavernomas (Fig. 12) [18].
The use of the blood pool agent ferumoxytol in
susceptibility-weighted MRI has been recently
discussed and seems to enhance the visibility of
CCM, among other vascular malformations
[19]. Since ferumoxytol has a good biocompat-
ibility profile, has no known long-term toxicity,
and, in contrast to the gadolinium-based contrast
agents, can be safely given to patients with
decreased renal function, it seems to be a prom-
ising tool for the diagnosis and follow-up of
cavernomas. However, further studies are neces-
sary to support the practical use of this agent.
– Diffusion-weighted images (DWI) usually do
not demonstrate restricted diffusion unless
there is subacute hemorrhage (Fig. 1).
– Contrast-enhanced T1-weighted sequences
can demonstrate minimal or absent gadolinium
enhancement. Associated DVAs can also be
detected in T1-weighted postcontrast images.
Digital Subtraction Angiography: On DSA,
most CCMs are angiographically occult and do
L.C.H. da Cruz and C.E. Pires
frontoparietal regions. SWI (c) shows a higher number of
lesions and confirms the presence of larger lesions
(arrows), which are not seen on the T2-weighted GRE
image (Reprinted with permission from Ref. [18])
not demonstrate arteriovenous shunting. When
they are evident on angiograms, the findings are
nonspecific and include minimal vascular blush in
the late venous phase. When they occur in com-
bination with other complex vascular
malformations, angiography can be helpful in
defining other lesions.
A classification system based on imaging and
pathologic features has been reported to stratify
these lesions [20]. Type I lesions are characterized
by hyperintensity on both T1WI and T2WI, which
is consistent with subacute hemorrhage. In Type II,
loculated regions of hemorrhage are surrounded by
gliosis and hemosiderin-stained brain parenchyma.
These CCMs exhibit a mixed signal intensity core
on both T1WI and T2WI, with a well-
circumscribed hypointense rim on T2WI, and are
the classic CCM with a “popcorn” appearance.
Type III lesions demonstrate a core that is iso- or
hypointense on T1WI and hypointense on T2WI,
surrounded by a hypointense rim on T2WI, consis-
tent with chronic resolved hemorrhage or hemosid-
erin within and surrounding the lesion. Type IV
corresponds to tiny lesions often seen as
punctate hypointense foci on susceptibility
sequences [21].
32 Miscellaneous Vascular Malformations (Cavernous Malformations, Developmental Venous. . . 733

Differential Diagnosis vascularized lesions. The association with DVA

should be considered before performing a surgical
Although cavernomas show a classic aspect on removal, due to the risk of causing a venous infarct
MRI, some conditions can be misdiagnosed as in the surrounding parenchyma. Multiple sequence
the multiple familial forms, giant cavernomas, MRIs play an important role in presurgical plan-
and CCM associated with brain parenchyma hem- ning to assess the extent of the lesion, define bor-
orrhage. Multiple foci lesions with hypointense ders, and plan the surgical approach and exposure.
lesions on susceptibility sequences can be a variety
of pathologic conditions that may mimic CCMs Cerebral Cavernous Malformation
and have to be considered in the differential diag-
General Imaging Neuroimaging
nosis. This variety of conditions include residual features modalities characteristics Hot spot
hyperintense hemorrhage; infections, specially Common MRI is the Characteristic Look for
treated toxoplasmosis, and fungus infection, like cerebral modality of “popcorn” or rounded
vascular choice “mulberry” areas of
aspergillosis and candidiasis; prior radiation ther- malformation appearance on marked
apy; disseminated intravascular coagulopathy; with classic MRI with signal loss in
amyloid angiopathy; diffuse axonal injury and appearance heterogeneous susceptibility-
on MRI signal intensity weighted MR
metastatic lesions secondary to melanoma and cho- within the sequences for
riocarcinoma; and metastasis from the lung, kid- lesion and a detection of
ney, and thyroid neoplasms (Fig. 13) [22]. rim of signal smaller and
loss due to concomitant
The distinction of CCM from hemorrhagic hemosiderin, cavernous
neoplasm may be difficult (Fig. 14). Some imag- better depicted lesions that
on may not be
ing features can be useful and can allow differen- susceptibility- detected with
tiation from a neoplasm. Some imaging signs are weighted traditional
more likely to be observed in CCMs, such as sequences sequences
(SWI and GRE
absence or little surrounding edema, absence of T2*)
identifiable nonhemorrhagic neoplasm, complete Isolated or CT scans can be Lack of mass
hemosiderin ring in the adjacent parenchyma, and multiple normal in effect, unless
lesions. In the 30–50 % of cases there has been
expected temporal evolution of the hematoma. latter type, recent bleeding
The presence of T1 hyperintensity in the they are often
familial, with
perilesional edema around a hemorrhagic lesion autosomal
can be able to differentiate CCM from other hem- dominant
orrhagic masses, like hemorrhagic tumors and transmission
Occur in all Angiographically Little or no
intracerebral hemorrhages (Fig. 15) [23].
parts of the occult enhancement
central
nervous
system but
Treatment and Final Considerations most
commonly in
Although most CCMs can simply be followed over the cerebral
hemispheres
time, surgical removal is an option in lesions caus-
ing significant morbidity related to mass effect,
epileptic activity, or repeated hemorrhage. Since
cavernomas are well circumscribed and Developmental Venous Anomaly
surrounded by a gliotic rim, surgical removal is
relatively simple. Control of hemorrhage is usually Developmental venous anomaly (DVA), also
easier because blood flow through the lesions is known as venous angioma or venous vascular
slower than that expected in more highly malformation, is the most frequently encountered
734
Fig. 13 A 77-year-old woman with lung cancer. Multiple
hemorrhagic metastatic lesions are better demonstrated on
SWI (b) image than on T2-weighted image (a).
Postcontrast axial T1-weighted image (c) shows multiple
cerebral vascular malformation. Usually asymp-
tomatic, it is frequently reported as a fortuitous
finding in imaging studies. DVAs represent a
purely venous entity, with no arterial component.
They are mostly considered extreme anatomical
variations of the transmedullary veins that are
necessary for the drainage of normal brain tissue.
Although DVA was thought to be rare before the
advent of CT and MR imaging, it is now consid-
ered the most common anomaly of the intracranial
vasculature, accounting for 63 % of vascular
malformations, with an overall incidence of
around 2.5 % [24].
DVAs are believed to be adaptations to acci-
dents occurring during embryogenesis between the
fourth and seventh stages of embryologic develop-
ment, resulting in occlusion or maldevelopment of
either the superficial or deep veins. Due to the
plasticity of the vascular system at this stage,
DVAs are formed as compensatory pathways,
recruiting and dilating preexisting transmedullary
veins [25, 26]. Since the involved vessels are not
abnormally formed, but apparently merely dilated
without any proliferative potential, the term DVA
seems to be more appropriate than “venous angio-
mas” or “venous vascular malformations” [26].
Although DVAs can be seen anywhere in infra-
or supratentorial compartments, draining either to
superficial or deep vein systems, the most
L.C.H. da Cruz and C.E. Pires
enhancing small lesions in the cortical-subcortical junction
of the hemispheres and at the subependyma of the right
lateral ventricle
common locations are the frontoparietal regions,
accounting for 36–64 %, usually draining toward
the frontal horn of the lateral ventricle. The cere-
bellar hemispheres are also a very common loca-
tion, estimated to represent 14–27 % of the cases,
draining toward the fourth ventricle [27]. They
may be isolated, draining one territory, or may
compose a complex venous draining system of a
whole cerebral or cerebellar hemisphere.
DVAs frequently coexist with other types of
CNS vascular malformations such as arteriove-
nous malformation, capillary telangiectasia, and,
especially, cavernous angiomas (15–30 % of
patients) (Fig. 16) [28]. Indeed, the association
between DVAs and cavernous malformations is
so common that the presence of a DVA on an
imaging study should promptly demand a search
for a cavernoma, which is more clinically impor-
tant. They are also associated with head and neck
venous malformations (Fig. 17) and hemangi-
omas and, more rarely, with cortical dysplasia,
since both occur at the same developmental
stage. DVA is thought to be a manifestation of
the underlying abnormal neuronal migration and
not its cause [29]. In a prior publication, the fre-
quency of DVAs in patients with cervicofacial
venous malformations was described as around
20 %. Most of these cervicofacial malformations
were superficial and extensive [30].
32 Miscellaneous Vascular Malformations (Cavernous Malformations, Developmental Venous. . .
Fig. 14 A solitary breast metastatic lesion on the right
lateral portion of the pons in a 65-year-old woman can be
confused with CCM. Axial T1-weighted (a), T2-weighted
(c), and GRE T2*-weighted (d) images demonstrate an
expansive mixed signal intensity lesion, with hemorrhagic
component, associated with a hypointense rim, surrounded
Arterialized DVAs are an atypical form
described when an early angiographic
opacification occurs during the mid- or late arte-
rial phases, typically found in large supratentorial
DVAs. In some particular examples, these sub-
types can be associated with an arteriovenous
malformation (AVM) [24].
DVAs can also be associated with regional
cerebral abnormalities other than CCMs within
the drainage territory in about 65 % of the cases.
The most frequent abnormality is locoregional
cerebral atrophy, accounting for almost one third
of the cases, followed by white matter lesions
(ranging from 8 % to 30 %) and dystrophic calci-
fication (nearly 10 %), which occurs typically in
735
by edema. The lesion causes compression to the anterior
and lateral portion of the fourth ventricle. Enhancement of
the central portion of the lesion was seen after gadolinium
administration on T1-weighted image (b). The histological
diagnosis was made after biopsy
the supratentorial and cerebellar white matter, as
well as in the basal ganglia and the caudate
nucleus. This calcification may represent chronic
hemorrhage, long-standing cerebral ischemia, or
venous hypertension [31]. White matter abnor-
malities usually present as a slight hyperintense
area that characteristically respects the boundaries
of the DVA drainage territory (Fig. 18). They are
stable over time, do not enhance after contrast
administration, and do not have mass effect. The
etiology of signal intensity changes is unclear but
may be related to edema, gliosis, or leukoaraiosis
secondary to focal vascular narrowing within the
common venous stem that can alter the hemody-
namics of the drainage area [32].
736
Fig. 15 Axial FLAIR (a), coronal T2-weighted (b), and
SWI (c) images demonstrate a hemorrhagic mass with
heterogeneous signal intensity, surrounded by edema in
the right frontal lobe. Sagittal T1-weighted image (d) dem-
onstrates hyperintensity within the vasogenic edema. Note
Clinical Manifestations
DVAs represent a physiologically competent
venous drainage pathway of normally functioning
brain tissue. Only rarely have neurologic symp-
toms been attributed to them. Although DVAs are
usually found incidentally in neuroimaging stud-
ies performed for other reasons and are asymp-
tomatic in most cases, their diagnosis often
generates concern among physicians who are
L.C.H. da Cruz and C.E. Pires
the centripetal pattern of the T1 hyperintense perilesional
signal intensity sign, observed only at the deep area around
the hemorrhagic lesion, but not observed at the peripheral
margin. The histological diagnosis of CCM was confirmed
after surgical excision
less familiar with the field of intracranial vascular
malformations. Also, their management requires
thorough understanding of the nature of DVAs,
including their frequent coexistence with other
types of vascular malformations.
DVAs are typically benign, asymptomatic, and
uncomplicated. However, they can rarely hemor-
rhage and become symptomatic, most notably after
venous ischemia or infarctions from spontaneous
thrombosis of the venous collector. On the other
32 Miscellaneous Vascular Malformations (Cavernous Malformations, Developmental Venous. . . 737

Fig. 16 Developmental venous anomaly with associated angioma associated to a hemorrhagic lesion, suggestive
cavernous malformation located at the left cerebellum of cavernous malformation. A drainage vein is better seen
hemisphere. Axial FLAIR (a), postcontrast T1-weighted at SWI (c, arrow)
MR (b), and SWI (c) images show a typical venous

Fig. 17 Postcontrast T1-weighted images demonstrate a right frontotemporal lymphangioma (arrows) associated with a
right cerebral hemisphere developmental venous anomaly, which has a deep drainage vein

hand, based on current literature, most instances of
hemorrhage with DVAs have been in patients with
combined vascular malformations, especially
CCM. It is controversial to attribute clinical symp-
toms to DVAs. In the vast majority of symptomatic
cases, hemorrhage and clinical manifestations such
as seizures probably originated from the accompa-
nying vascular malformation, rather than from the
DVA alone. The association between DVAs and
CCMs has been widely described. DVA is
associated with one or more cavernomas in about
13–40 % of cases [28]. These CCMs are thought to
738 L.C.H. da Cruz and C.E. Pires

Fig. 18 Axial FLAIR (a) and SWI (b) images demonstrated a venous angioma with deep venous drainage, associated to
marked signal intensity abnormality, extending to the ventricular margin

Fig. 19 Axial postcontrast T1-weighted (a–c) images Axial FLAIR (d) and SWI (e, f) images demonstrate a left
demonstrate the characteristic enhancement and caput cerebellum hemisphere hematoma associated with devel-
medusae distribution of the developmental venous anom- opmental venous anomaly. No cavernoma was observed
aly, which drains anteriorly via an enlarged medullary vein. associated to the DVA

be responsible for the vast majority of symptomatic
cases, previously attributed to the DVAs [24]. It is
interesting that CCMs associated to DVAs have a
more aggressive clinical course than CCMs alone
and are more likely to present with symptomatic
hemorrhage [33]. Although DVAs have a very low
risk of hemorrhage (0.22–0.68 %) [34], DVAs can
very rarely cause intracerebral hemorrhage when
they are not associated with CCM. The bleeding
risk associated with DVAs is greater in the poste-
rior fossa (Fig. 19). However, it can very rarely
occur in other regions of the brain parenchyma
[35]. Thrombosis of the collector vein of a DVA
is a rare complication that can lead to cerebral
32 Miscellaneous Vascular Malformations (Cavernous Malformations, Developmental Venous. . .
venous infarction in about 53 % of the cases, which
can be associated with hemorrhage (37 %), and,
less frequently, subarachnoid and intraventricular
hemorrhage (5 %) [36].
DVAs may also rarely cause symptoms by their
mechanical relationship with adjacent brain struc-
tures. Some reports have described clinical symp-
toms related to DVAs, such as cranial nerve
compression [37] and hydrocephalus, secondary
to cerebral aqueduct obstruction by DVA [38].
Histological Findings
A cluster of venous radicles that converge into a
collecting vein characterizes DVA, resulting in the
typical appearance of caput medusae. The
collecting vein has a variable length and joins
the superficial or the deep venous system after
crossing the brain parenchyma. In some particular
cases, the drainage pathway can be done by both
the superficial and deep venous systems, with one
of those as predominant [24].
Macroscopically, DVAs are characterized by
multiple medullary or subcortical dilated veins,
which are radially disposed with a centripetal
drainage system into one venous collector that is
also dilated. This collector drains into either a
superficial (subcortical) or a deep (pial) venous
system. The usual drainage vein of the specific
location is commonly absent.
Microscopically, DVAs appear as a conglom-
erate of dilated veins with minimal parietal thick-
ening and hyalinization, with the absence of
elastic lamina, and loosely arranged smooth mus-
cle layers, intermingled with normal brain paren-
chyma [24]. The intervening brain tissue
associated with DVA is usually normal, with no
evidence of hemorrhage or gliosis [22]. It can be
extended, compromising a whole cerebral hemi-
sphere, or limited to a small territory.
Neuroimaging
DVAs are best depicted in contrast-enhanced
imaging studies. The venous collector is usually
identified as a linear or curvilinear enhancing
739
structure, and enhancement of the dilated
transmedullary veins with a radial distribution
converging to the venous collector (caput medu-
sae) can also be observed.
Computed Tomography: DVAs are usually not
identified by non-enhanced brain CT. However,
they can rarely appear as focal hyper-attenuations,
especially when associated with a cavernoma.
After intravenous iodinated contrast administra-
tion, DVAs can be identified as radially distributed
vascular structures converging into a unique
venous collector, which in turn drains into a corti-
cal or deep vein, or even into a dural venous sinus.
When the anomaly is extremely small, a linear
enhancing structure (venous collector) running in
the direction of the ventricular or pial surface may
be the only finding suggestive of DVA.
Although it is usually unnecessary, a more
detailed evaluation of a DVA’s venous drainage
can be obtained with angio-CT with 3D MIP
reformation.
Magnetic Resonance:
– T1-weighted sequences: Although DVAs are
usually not depicted on non-enhanced T1WI,
especially when they are small, large lesions
can eventually demonstrate flow voids.
– T2-weighted sequences: With T2WI, flow
voids possibly associated with degradation
blood products can be seen.
– FLAIR: Hyperintense foci on FLAIR sequences
related to venous ischemia or hemorrhage in
different phases can also be demonstrated.
– Magnetic susceptibility sequences:
Susceptibility-weighted imaging (SWI) and
T2* gradient echo demonstrate that the DVA
has a marked hypointensity secondary to the
paramagnetic deoxyhemoglobin with venous
blood (Fig. 20). These sequences can also dem-
onstrate hypointense foci if DVAs are associ-
ated with old bleeding or a cavernoma (Fig. 21).
– As described for cavernomas, the use of the
blood pool agent, ferumoxytol, seems to demon-
strate additional tributary veins and improve the
detection of DVAs in susceptibility-weighted
MRI [19]. However, further studies are neces-
sary to support practical use of this agent.
– Contrast-enhanced T1WI sequences: A T1WI
sequence after gadolinium administration is
740
Fig. 20 Venous angioma with intravascular hypointensity
on susceptibility image. Right frontal venous angioma is
demonstrated on axial FLAIR (a) and postcontrast
Fig. 21 Axial T1-weighted image after contrast adminis-
tration (a, b) and FLAIR images (c) demonstrate a DVA
with deep venous drainage. Associated signal intensity
abnormality surrounds the DVA, extending into the white
matter (c). On SWI images (d, e), linear areas of
the best sequence to identify a DVA, which
appears as a tubular-enhanced structure directed
to a pial or ventricular surface. The small
transmedullary veins draining radially into the
collector are also seen. Although venous angio-
MR can easily demonstrate the classic caput
L.C.H. da Cruz and C.E. Pires
T1-weighted MR (b) images. A marked hypointense signal
intensity on SWI (c) is seen, probably secondary to
deoxyhemoglobin in venous blood
hypointensity can also be observed and can be secondary
to deoxyhemoglobin in venous blood (d, arrow). Round
hypointense areas within the lesion can be also demon-
strated (e, arrows), which may correspond to cavernomas
or even previous hemorrhage
medusae pattern, this sequence is usually not
necessary for the diagnosis of DVAs.
Digital Subtraction Angiography: DSA is
only rarely indicated if CT or MR interpretations
32 Miscellaneous Vascular Malformations (Cavernous Malformations, Developmental Venous. . . 741

are doubtful. The classic angiographic feature is Developmental Venous Anomaly

the typical caput medusae appearance of the med-
General Imaging Neuroimaging
ullary veins seen only in the late venous phase of features modalities characteristics Hot spot
angiography. Larger lesions can demonstrate a Usually Best depicted Radially Always look
dense capillary stain; however, the lack of arterial incidental by contrast- distributed for other
finding in enhanced CT small vascular associated
enlargement suggests a DVA [39]. neuroimaging or MRI structures vascular
In rare cases, the transmedullary veins can studies (caput malformations,
be seen earlier in the late capillary phase performed for medusae especially
other reasons pattern) cavernomas,
due to dilatation of capillary spaces, which pro- converging in cases of
vides a more rapid transit time, a finding that into a unique symptomatic
should not be misinterpreted as an arteriovenous venous DVAs
collector,
shunt. which in turn
Angiography ideally details the venous drain- drains into a
cortical or
age to the cortical or ventricular surface or to the deep vein, or
dural sinus and is the gold standard method to even into a
evaluate permeability of the whole extension of venous dural
sinus
the collector vein. Asymptomatic Angiography:
in most cases rarely
indicated
No age or SWI:
Treatment and Final Considerations gender important for
predilection the
DVAs should be considered as simple varia- identification
of associated
tions of the brain’s venous system and usually vascular
do not require intervention. Since the vast malformations,
especially
majority of DVAs are clinically asymptomatic,
cavernomas
most experts question the benefit of surgical
treatment, since the resulting risk of an iatro-
genic venous infarct and hemorrhagic compli-
cations would far exceed the risk of irreversible Capillary Telangiectasia
damage from the DVA itself during a patient’s
lifetime. Additionally, DVAs that become Brain capillary telangiectasias (BCTs) are a dis-
symptomatic by bleeding frequently coexist tinct category of vascular malformations that are
with other types of vascular malformations, occult on angiography but are being increasingly
suggesting that the symptoms are actually discovered incidentally on MRIs when the brain is
caused by the associated vascular malformation imaged for other reasons [40]. They consist of
other than by the DVA itself. Therefore, any localized collections of multiple thin-walled vas-
signal intensity anomaly of the adjacent paren- cular channels interposed between normal brain
chyma should prompt the search for other vas- parenchyma.
cular lesions. Most capillary telangiectasias are considered
However, it is always important to mention benign entities, generally asymptomatic, and
the existence of a DVA in a medical report, without further implications for patient manage-
especially if it is an incidental finding in a ment. The true incidence of these malformations
neuroimaging test performed for other surgical is difficult to estimate because most are likely to
reasons, such as tumors. In this case, the neu- be clinically asymptomatic, but autopsy series
rosurgeon must be alerted to the fact that struc- suggest they are not uncommon, representing
tural damage of the venous anomaly could approximately 16–20 % of all CNS vascular
compromise the drainage of normal brain malformations [41, 42]. They are also considered
parenchyma. the second most common vascular anomaly after
742
developmental venous anomalies on imaging
studies [43].
The area of involvement of the brain is typically
small, ranging from several millimeters to 2 cm in
size, and common sites of involvement include the
pons, cerebral hemispheres, and spinal cord. The
pons is the classic location of these lesions. BCT is
inconspicuous on conventional pre-contrast MR
images and in most cases is detected due to faint
enhancement after intravenous contrast administra-
tion [41]. It can be difficult to differentiate CCMs
from BCTs when the lesion is in the pons. The
association with hemorrhage implies the diagnosis
of CCM rather than BCT. Typically, BCT is occult
on angiography studies.
Clinical Manifestations
The vast majority of capillary telangiectasias are
incidental findings, as these malformations are
nearly always asymptomatic, without further
implications for patient management
[40]. Although headaches, confusion, weakness,
dizziness, visual changes, vertigo, tinnitus, and
seizures can be rarely associated with capillary
telangiectasias, confidently attributing such gen-
eralized symptoms to these lesions is difficult.
Furthermore, the precise cause of these sup-
posed symptoms is usually not known because
reports of apoplectic hemorrhage developing
from capillary telangiectasia are extremely rare
[44]. In fact, similar to DVAs, the detection of
hemorrhage on imaging studies of patients with
capillary malformations should prompt consider-
ation of coexisting lesions, such as cavernomas, or
may indicate a “mixed” form of lesion.
To avoid the diagnostic dilemma, a character-
istic signal intensity loss on susceptibility-
weighted sequences has been described, which,
in conjunction with postcontrast enhancement, is
virtually diagnostic of BCT [45].
Histological Findings
Macroscopically, capillary telangiectasias can be
undetectable or present merely as petechial
L.C.H. da Cruz and C.E. Pires
hemorrhages or small areas of altered coloration
of the brain tissue, far more common in the pons.
Microscopically, it is composed of numerous
thin-walled vessels without smooth muscle or
elastic fibers that are separated by neural tissue.
This is in contrast to cavernous malformations,
which have no normal brain tissue within their
confines. There is little or no gliosis, calcification,
or gross hemorrhage [40].
Neuroimaging
Capillary telangiectasias are mostly located
within the brain stem and pontine regions,
although they can be seen virtually anywhere in
the brain parenchyma or spinal cord [41]. They
are usually seen as subtle lesions without mass
effect or circumjacent edema. As they are almost
always invisible both on CT and DSA, introduc-
tion of MRI was definitive for the wise recogni-
tion of this entity by the radiology community.
Computed Tomography: Unenhanced CT can
almost never depict capillary malformations,
although punctate calcifications can be seen in
exceptional cases. Enhanced CT can rarely dem-
onstrate capillary malformations as small areas of
subtle enhancement in typical locations (pontine
region).
Magnetic Resonance Imaging: Contrast-
enhanced MRI is the best imaging modality for
the diagnosis of capillary telangiectasia.
– T2 or FLAIR sequences can be normal or
demonstrate a discreet area of slightly
increased signal intensity, without mass effect
or adjacent edema.
– Susceptibility-weighted magnetic resonance
sequences (SWI/T2*-weighted GRE sequence)
can demonstrate typical moderate hypointense
signal in the whole malformation or just in the
center of the lesion and therefore play an impor-
tant role in the differential diagnosis with other
brain stem lesions, such as neoplastic or sub-
acute ischemic lesions. This hypointensity may
be explained by the low-velocity blood flow
in ectatic venous channels, resulting in a drop
of oxygen saturation and thus increasing
32 Miscellaneous Vascular Malformations (Cavernous Malformations, Developmental Venous. . .
Fig. 22 Typical case of a central pontine capillary telan-
giectasia. The lesion is slightly hyperintense on
T2-weighted image (a), faint homogeneous enhancement
on postcontrast T1-weighted image (b), and considerable
deoxyhemoglobin, causing susceptibility effect
[40] or presumably representing residual sub-
clinical bleeding. SWI is more sensitive to sus-
ceptibility changes than T2* gradient echo,
being able to demonstrate small BCT lesions
[43, 45].
– On T1-weighted sequences, capillary
malformations are typically iso- to hypointense
compared with the brain parenchyma. After IV
gadolinium administration, the lesion appears
as an ill-defined focal enhancement, sometimes
with a lineariform and mottled pattern. A prom-
inent draining vein can be depicted on both
postcontrast T1WI and SWI (Fig. 22) [43].
743
signal intensity drop on SWI (c). Note that SWI (d) more
clearly depicts a prominent vessel related to the BCT
(arrow), which may correspond to a drainage vein or an
associated developmental venous anomaly
– Diffusion-weighted imaging is usually isointense
or hypointense, which is helpful in discerning
BCTs from subacute ischemic disease [46].
Digital Subtraction Angiography: Catheter
angiograms are usually normal but can rarely
demonstrate a faint blush.
Treatment, Differential Diagnosis,
and Final Considerations
Since capillary telangiectasias are benign lesions
that are almost always asymptomatic, have
744 L.C.H. da Cruz and C.E. Pires

interspersed normal brain parenchyma, and are Sinus Pericranii

most frequently located in the pons, surgical treat-
ment becomes impractical and unnecessary. No Sinus pericranii (SP) is an unusual venous mal-
follow-up is required if the imaging appearances formation characterized by anomalous commu-
are typical. nication between the intracranial dural sinuses
BCT is a benign entity and should be differ- and dilated epicranial venous blood vessels
entiated from some conditions to avoid unneces- through a bony defect. It consists of a collection
sary follow-up imaging evaluation, surgical of nonmuscular venous structures adhered
biopsy, or even instigation of treatment for pre- tightly to the outer surface of the skull, which
sumed pathology. Its MRI appearance should be communicates with an intracranial venous sinus
distinguished from primary neoplasms like astro- through diploic veins. The varicosities, inti-
cytomas, lymphomas, metastases, inflammatory mately associated with periosteum, are distensi-
processes, active demyelination lesions, and sub- ble and vary in size with changes in intracranial
acute ischemia. In the vast majority of cases, pressure [47].
demonstration of typical signal intensity loss on Sinus pericranii can appear at any age. The
susceptibility sequences (SWI, T2*GE) in an vast majority of cases are congenital in nature,
enhancing focal brain lesion, otherwise but they can also be acquired, traumatic, or spon-
unremarkable on conventional MR images, is taneous. In congenital cases, there is often associ-
highly specific for BCT, and no differential diag- ation with other venous anomalies and
nosis needs to be considered in this scenario syndromes, especially blue rubber bleb nevus
[45]. In cases of atypical features, the lack of syndrome.
mass effect or edema should favor the diagnosis Traumatic causes can be related to a skull frac-
of capillary malformations instead of glioma or ture, a tear in the dural sinuses or emissary veins, a
metastasis. Additionally, the lack of water depressed bone fracture leading to a dural sinus
restriction in DWI sequences is useful to differ- tear, or an epidural venous hematoma caused by a
entiate these lesions from subacute ischemic sinus tear. SP can be associated with other con-
disease [46]. genital vascular anomalies, such as an aneurysmal
internal cerebral vein, cavernous hemangioma,
General Imaging Neuroimaging
features modalities characteristics Hot spots
and systemic angioma [48].
Usually Best Poorly Demonstration Three different pathophysiological types of
incidental depicted demarcated of signal sinus pericranii have been described: circulation
finding in by area of slightly intensity loss
neuroimaging contrast- increased signal on SWI/T2* in
in a closed system in which the blood comes from
studies enhanced intensity on T2 an enhancing the sinus and also returns to it, circulation origi-
performed for MRI and FLAIR focal brain nating in the sinus with peripheral passage of the
other reasons sequences, with lesion,
a faint otherwise venous drainage, and a connection between an
ill-defined unremarkable angioma and the sinus [48].
enhancement on
on T1WI conventional
MR images, is
highly specific Clinical Manifestations
for BCT
Almost always Almost Moderately No mass effect,
asymptomatic always hypointense on edema, or In congenital cases, it usually manifests in the first
invisible SWI/T2* restricted months or years of life as a compressible
both on diffusion
CT and
nonpulsatile soft tissue scalp mass, most com-
DSA monly located in the frontal region (40 %) along
No age or No mass effect or close to the midline. Other possible locations
gender or edema
predilection
include the parietal (34 %), occipital (23 %), and,
more rarely, temporal regions (4 %) [49, 50].
32 Miscellaneous Vascular Malformations (Cavernous Malformations, Developmental Venous. . .
Patients typically present with a nontender,
nonpulsatile fluctuant soft tissue mass in the
scalp. The varicosities typically reduce in upright
position and distend when prone or with the
Valsalva maneuver, crying, coughing, and jugular
vein compression. They are usually stable or show
discreet enlargement with time, although there
have been rare reports of spontaneous
regression [51].
Since it is usually asymptomatic and painless,
the primary concern is often cosmetic, but occa-
sionally vague symptoms may occur, including
headache, vertigo, feelings of fullness, and local
pain [52]. A sudden clinical alteration of the
lesion from soft and painless to firm and painful
is a concern and should raise suspicion for
complications such as thrombosis or, more
rarely, hemorrhage, emboli, and infection. The
occurrence of partial thrombosis is not unex-
pected because of the typical slow flow
of blood within these lesions. Other reported
complications include hemorrhage and air
embolisms [53].
Neuroimaging
The diagnosis of SP is made from a combination
of clinical and radiologic features. It is based on
the demonstration of an extracranial venous com-
ponent communicating with a dural sinus through
a bony defect. The most commonly involved sinus
is the sagittal superior, followed distantly by the
transverse sinus. The extracranial venous compo-
nent may include a venous varix, venous malfor-
mation, or multiple veins.
Radiography: Plain skull radiography
shows bony defects, cortical thinning, or focal
bony erosions, in close relation to a soft
tissue mass.
Computed Tomography: Unenhanced CT
scans reveal a paramedian extracranial soft tissue
mass overlying a calvarial defect, usually in close
relation to the sagittal superior sinus. A
transosseous defect within the skull and connec-
tions between the epicranial varicosities with an
underlying dural sinus can be demonstrated. CT
745
angiography is useful for excluding arterial
abnormalities and can demonstrate heteroge-
neous enhancement of the extracranial venous
component because of the typical slow flow sta-
tus of some of the connections. CT venography
(CTV) or delayed imaging usually shows
homogenous enhancement throughout the vascu-
lar mass. In cases of thrombosis, CTV may
demonstrate filling defects within the
varicosities [47].
Magnetic Resonance Imaging: MR images
demonstrate a paramedian extracranial mass
with a T2 hyperintense signal in most cases.
Signal intensity in T1WI sequences is variable.
Flow voids may be present in a rapid-flowing
varix or venous malformation. Contrast-
enhanced MRI demonstrates homogenous “vas-
cular” enhancement in cases of varix and hetero-
geneous enhancement in cases of venous
malformations. MR venography (MRV) delin-
eates sharply demarcated vascular components
of the lesion. In cases of thrombosis, a subacute
thrombus shows hyperintensity on T1-weighted
sequence, in correspondence to filling defects
(Fig. 23) [47].
Doppler Sonography: Ultrasound imaging
demonstrates anechoic or hypoechoic extracra-
nial lesions with lobulating and bulging con-
tours. Doppler flowmetry shows intracranial
and extracranial venous communication through
a calvarial defect. SP shows pulsatile Doppler
wave patterns and high-flow/low-flow vascular
shunting independent of changes in intracranial
pressure [54].
Digital Subtraction Angiography: The arterial
phase of DSA usually shows no arterial abnormal-
ities with exception for an eventual arterial blush.
DSA during the venous phase confirms the
transosseous connection between the pericranial
varicosities and the dural sinus [50].
Treatment, Differential Diagnosis,
and Final Considerations
The differential diagnosis is extensive and
includes most of scalp lesions that appear as soft
746
Fig. 23 Sinus pericranii. Midsagittal postcontrast
T1-weighted MR images (a–c) and a tridimensional con-
structive interference in steady-state (3D CISS) MR
sequence (d). An emissary vein extends from a
compressible masses such as meningocele,
encephalocele, skin or skull lesions, and other
vascular abnormalities like hemangiomas, arterio-
venous malformations, lipoma, dermoid cysts,
and epidermoid cysts [52]. In general, neuroim-
aging modalities such as CT and MR are able to
differentiate these conditions. In case of atretic
cephalocele, absence of enhancement usually
helps to discern this entity from sinus pericranii.
L.C.H. da Cruz and C.E. Pires
subcutaneous part of the venous anomaly through a small
bone defect at the parietal region (c, d arrows) to the
posterior part of the superior sagittal sinus
Although SP is usually asymptomatic, it may
rarely cause fatal complications such as thrombo-
sis, traumatic air embolism, or hemorrhage. Thus,
prompt treatment after diagnosis is required for
prophylactic purposes to prevent complications,
as well as for aesthetic purposes.
Treatment options include surgery in most
cases and endovascular therapy for smaller lesions
[50]. In surgical treatment, the key point is to
32 Miscellaneous Vascular Malformations (Cavernous Malformations, Developmental Venous. . . 747

resect the extracranial venous package and ligate arteriovenous malformations) from those that do
the emissary-communicating vein. not have shunting. The latter group is the main
topic of our chapter and includes miscellaneous
General Imaging Neuroimaging
features modalities characteristics Hot spots
etiology conditions, such as venous anomalies
Most cases Diagnosis is Extracranial Think of SP (cavernous hemangioma or cavernoma and devel-
are based on the paramedian in extracranial opmental venous anomalies), capillary anomalies
congenital demonstration soft tissue masses in
and become of an mass with close relation
(capillary telangiectasia), and sinus pericranii, a
clinically extracranial vascular to the sagittal special condition characterized by an abnormal
apparent in venous enhancement superior sinus communication between the dural venous sinus
the first years component in the
or months as communicating venography and extracranial venous system. All these lesions
a to dural sinus phase must be distinguished from one another by their
compressible through a bony overlying a different natural histories and the various treat-
nonpulsatile defect bony defect
soft tissue in close ment strategies. Cross-sectional imaging, includ-
scalp mass relation to a ing multiplanar computed tomography and
dural sinus
magnetic resonance imaging, is important not
Traumatic Best depicted Sagittal Think of
etiology on CT or MR superior sinus thrombosis in only in the differential diagnosis of brain
more venography much more painful and malformations but also in the long-term follow-
uncommon common noncompressible
lesions with
up monitoring of patients with lesions that do not
filling defects in require immediate intervention. It is also of great
delayed contrast- importance in presurgical planning to assess the
enhanced
imaging (CTV, extent of the lesion, define borders, and plan the
MRV, or DSA) surgical approach. In this chapter, we attempt to
Doppler provide a concise review of the imaging methods
ultrasound is
useful for utilized in the clinical diagnosis and management
neonates of the vascular lesions listed above.
DSA during the
venous phase
confirms the
diagnosis
References
1. Chaloupka JC, Huddle DC (1998) Classification of
vascular malformations of the central nervous system.
Summary Neuroimaging Clin N Am 8(2):295–321
2. MacDonald RL, Stoodley M, Weir B (2001) Vascular
malformations of the central nervous system.
Cerebral vascular malformations are uncommon Neurosurg Q 11(4):231–247
lesions of great clinical relevance, since clinical 3. Sigal R, Krief O, Houtteville J, Halimi P, Doyon D,
outcome may include serious neurological dis- Pariente D (1990) Occult cerebrovascular
malformations: follow-up with MR imaging. Radiol-
ability in a significant proportion of individuals
ogy 176:815–819
bearing them. These lesions are seen increasingly 4. Tagle P, Huete I, Méndez J, del Villar S (1986) Intra-
in clinical practice, primarily because of advances cranial cavernous angioma: presentation and manage-
in cross-sectional brain imaging, and their evalu- ment. J Neurosurg 64(5):720–723
5. Russel DS, Rubistein LJ (1989) Pathology of tumors of
ation provides a diagnostic challenge to radiolo-
the nervous system, 5th edn. Williams & Wilkins,
gists due to their similar clinical manifestations Baltimore
and imaging features. Different classification sys- 6. Cavalcanti DD, Kalani MY, Martirosyan NL, Eales J,
tems have been put forward. The most commonly Spetzler RF, Preul MC (2012) Cerebral cavernous
malformations: from genes to proteins to disease. J
used classification system separates vascular
Neurosurg 116:122–132
lesions that demonstrate shunting from arterial 7. Brunereau L, Labauge P, Tournier-Lasserve E
to venous systems malformations (brain et al (2000) Familial form of intracranial cavernous
748
angioma: MR imaging findings in 51 families. French
Society of Neurosurgery. Radiology 214(1):209–216
8. Riant F, Bergametti F, Ayrignac X, Boulday G,
Tournier-Lasserve E (2010) Recent insights into cere-
bral cavernous malformations: the molecular genetics
of CCM. FEBS J 277(5):1070–1075
9. Petersen TA, Morrison LA, Schrader RM, Hart BL
(2010) Familial versus sporadic cavernous
malformations: differences in developmental venous
anomaly association and lesion phenotype. AJNR Am
J Neuroradiol 31:377–382
10. Jain R, Robertson PL, Gandhi D, Gujar SK, Muraszko
KM, Gebarski S (2005) Radiation-induced cavernomas
of the brain. AJNR Am J Neuroradiol 26:1158–1162
11. Robinson JR, Awad IA, Little JR (1991) Natural his-
tory of the cavernous angioma. J Neurosurg 75
(5):709–714
12. Batra S, Lin D, Recinos PF, Zhang J, Rigamonti D
(2009) Cavernous malformations: natural history, diag-
nosis and treatment. Nat Rev Neurol 5(12):659–670
13. Sohn CH, Kim SP, Kim IM et al (2003) Characteristic
MR imaging findings of cavernous hemangiomas in
the cavernous sinus. AJNR Am J Neuroradiol
24:1148–1151
14. Ide C, De Coene B, Baudrez V (2000) MR features of
cavernous angioma. JBR-BTR 83(6):320
15. Meyer FB, Lombardi D, Scheithauer B, Nichols DA
(1990) Extra-axial cavernous hemangiomas involving
the dural sinuses. J Neurosurg 73(2):187–192
16. Kondziolka D, Monaco EA, Lunsford LD (2013) Cav-
ernous malformations and hemorrhage risk. Prog
Neurol Surg 27:141–146
17. Pinker K, Stavrou I, Knosp E et al (2006) Are cerebral
cavernomas truly nonenhancing lesions and thereby
distinguishable from arteriovenous malformations?
MRI findings and histopathological correlation. Magn
Reson Imaging 24(5):631–637
18. de Souza JM, Domingues RC, Cruz LC Jr, Domingues
FS, Iasbeck T, Gasparetto EL (2008) Susceptibility-
weighted imaging for the evaluation of patients with
familial cerebral cavernous malformations: a compari-
son with T2-weighted fast spin-echo and gradient-echo
sequences. AJNR Am J Neuroradiol 29(1):154–158
19. Dósa E, Tuladhar S, Muldoon LL, Hamilton BE,
Rooney WD, Neuwelt EA (2011) MRI using
ferumoxytol improves the visualization of central ner-
vous system vascular malformations. Stroke 42
(6):1581–1588
20. Zabramski JM, Wascher TM, Spetzler RF et al (1994)
The natural history of familial cavernous
malformations: results of an ongoing study. J
Neurosurg 80:422–432
21. Campbell PG, Jabbour P, Yadla S, Awad IA (2010)
Emerging clinical imaging techniques for cerebral cav-
ernous malformations: a systematic review. Neurosurg
Focus 29(3):E6. 1–15
22. Atlas SW (ed) (2008) Magnetic resonance imaging of
the brain and spine, 4th edn. Lippincott Williams &
Wilkins, Philadelphia
L.C.H. da Cruz and C.E. Pires
23. Yun TJ, Na DG, Kwon BJ et al (2008) A T1
hyperintense perilesional signal aids in the differentia-
tion of a cavernous angioma from other hemorrhagic
masses. AJNR Am J Neuroradiol 29(3):494–500
24. Ruiz DS, Yilmaz H, Gailloud P (2009) Cerebral devel-
opmental venous anomalies: current concepts. Ann
Neurol 66:271–283
25. Truwit CL (1992) Venous angioma of the brain: his-
tory, significance, and imaging findings. AJR Am J
Roentgenol 159:1299–1307
26. Cheong WY, Tan KP (1993) Cerebral venous
angioma – a misnomer? Ann Acad Med Singapore 22
(5):736–741
27. Lee C, Pennington MA, Kenney CM (1996) MR eval-
uation of developmental venous anomalies: medullary
venous anatomy of venous angiomas. AJNR Am J
Neuroradiol 17(1):61–70
28. Huber G, Henkes H, Hermes M et al (1996) Regional
association of developmental venous anomalies with
angiographically occult vascular malformations. Eur
Radiol 6:30–37
29. Desai K, Bhayani R, Nadkarni T et al (2002) Develop-
mental deep venous system anomaly associated with
congenital malformation of the brain. Pediatr
Neurosurg 36:37–39
30. Boukobza M, Enjolras O, Guichard JP et al (1996)
Cerebral developmental venous anomalies associated
with head and neck venous malformations. AJNR Am
J Neuroradiol 17(5):987–994
31. San Millán Ruíz D, Delavelle J, Yilmaz H et al (2007)
Parenchymal abnormalities associated with develop-
mental venous anomalies. Neuroradiology 49:987–995
32. Santucci GM, Leach JL, Ying J, Leach SD, Tomsick
TA (2008) Brain parenchymal signal abnormalities
associated with developmental venous anomalies:
detailed MR imaging assessment. AJNR Am J
Neuroradiol 29(7):1317–1323
33. Wurm G, Schnizer M, Fellner FA (2005) Cerebral
cavernous malformations associated with venous
anomalies: surgical considerations. Neurosurgery 57
(1 Suppl):42–58; discussion 42–58
34. McLaughlin MR, Kondziolka D, Flickinger JC
et al (1998) The prospective natural history of cerebral
venous malformations. Neurosurgery 43:195–200; dis-
cussion 200–191
35. Ku MG, Rhee DY, Park HS, Kim DN (2009) Repeated
intracerebral hemorrhage from developmental venous
anomaly alone. J Korean Neurosurg Soc 45:46–49
36. Gama RL, Nakayama M, Tavora DGF, Bomfim RC,
Carneiro TC, Pimentel LHC (2008) Thrombosed
developmental venous anomaly associated with cere-
bral venous infarct. Arq Neuropsiquiatr 66
(3-A):560–562
37. Peterson AM, Williams RL, Fukui MB, Meltzer CC
(2002) Venous angioma adjacent to the root entry zone
of the trigeminal nerve: implications for management
of trigeminal neuralgia. Neuroradiology 44:342–346
38. Paulson D, Hwang SW, Whitehead WE, Curry DJ,
Luerssen TG, Jea A (2012) Aqueductal developmental
32 Miscellaneous Vascular Malformations (Cavernous Malformations, Developmental Venous. . .
venous anomaly as an unusual cause of congenital
hydrocephalus: a case report and review of the litera-
ture. J Med Case Rep 6:7
39. Lee M, Kim MS (2012) Image findings in brain devel-
opmental venous anomalies. J Cerebrovasc Endovasc
Neurosurg 14(1):37–43
40. Lee RR, Becher MW, Benson ML et al (1997) Brain
capillary telangiectasia: MR imaging appearance and
clinic histopathologic findings. Radiology
205:797–805
41. Castillo M, Morrison T, Shaw J, Bouldin T (2001) MR
imaging and histologic features of capillary telangiec-
tasia of the basal ganglia. AJNR Am J Neuroradiol
22:1553–1555
42. Huddle DC, Chaloupka JC, Sebgal V (1999) Clinically
aggressive diffuse capillary telangiectasia of the brain
stem: a clinical radiologic-pathologic case study.
AJNR Am J Neuroradiol 20:1674–1677
43. El-Koussy M, Schroth G, Gralla J, Brenkenfeld C,
Andres RH, Jung S et al (2012) Susceptibility-
weighted MR imaging for diagnosis of capillary telan-
giectasia of the brain. AJNR Am J Neuroradiol
33:715–720
44. Bland LI, Lapham LW, Ketonen L, Okawara SH
(1994) Acute cerebellar hemorrhage secondary to cap-
illary telangiectasia in an infant: a case report. Arch
Neurol 51:1151–1154
45. Barr RM, Dillon WP, Wilson CB (1996) Slow-flow
vascular malformations of the pons: capillary telangi-
ectasias? AJNR Am J Neuroradiol 17:71–78
749
46. Finkenzeller T, Fellner FA, Trenkler J et al (2010) Cap-
illary telangiectasias of the pons: does diffusion-
weighted MR increase diagnostic accuracy? Eur J
Radiol 74:e112–e116
47. Carpenter JS, Rosen CL, Bailes JE, Gailloud P (2004)
Sinus pericranii: clinical and imaging findings in two
cases of spontaneous partial thrombosis. AJNR Am J
Neuroradiol 25:121–125
48. Wen C, Chang Y, Wang H, Kuo M, Tu Y (2005) Sinus
pericranii: from gross and neuroimaging findings to
different pathophysiological changes. Childs Nerv
Syst 21:482–488
49. Vaquero J, de Sola RG, Martinez R (1983) Lateral
sinus pericranii. Case report. J Neurosurg 58:139–140
50. Lee CH, Lee YS, Lee JH, Lee HG et al (2009) Sinus
pericranii: a case report and the literature review. Kor J
Cerebrovasc Surg 11:174–178
51. Hayakawa I, Fujiwara K, Sasaki A et al (1978) Spon-
taneous regression of sinus pericranii: report of a case
(in Japanese). No Shinkei Geka 6:91–95
52. Sheu M, Fauteux G, Chang H et al (2002) Sinus
pericranii: dermatologic considerations and literature
review. J Am Acad Dermatol 46:934–941
53. Vinas FC, Valenzuela S, Zuleta A (1994) Literature
review: sinus pericranii. Neurol Res 16:471–474
54. Kim YJ, Kim IO, Cheon JE, Lim YJ, Kim WS, Yeon
KM (2011) Sonographic features of sinus pericranii in
4 pediatric patients. J Ultrasound Med 30:411–417
 Miscellaneous Vascular Diseases
33
Matylda H. Machnowska and Richard I. Aviv

Contents Pathology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 763

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752
Fabry’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752
Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752
Clinical Presentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752
Pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 753
CT Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 753
MRI Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 753
PRES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 753
Pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 754
Pathology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 754
Imaging Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 755
RCVS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 756
Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 756
Clinical . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 756
Moyamoya . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 758
Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 758
Genetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 759
Clinical Presentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 759
Pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 759
Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 760
Cerebral Autosomal Dominant
Arteriopathy with Subcortical Infarcts
and Leukoencephalopathy (CADASIL) . . . . . . . . . . 762
Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 762
Genetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 762
MRI Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 763
CARASIL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 763
Vasculitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 764
PACNS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 764
Takayasu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 765
Giant Cell Arteritis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 768
Polyarteritis Nodosa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 768
Kawasaki Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 769
Churg-Strauss Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . 769
Wegener’s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 770
Systemic Lupus Erythematosus . . . . . . . . . . . . . . . . . . . 771
Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771
Clinical Presentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 772
Pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 772
Pathology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 772
Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 772
Special Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 773
Infectious Vasculitides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 773
Varicella Zoster Vasculitis . . . . . . . . . . . . . . . . . . . . . . . . . . . 773
HIV Vasculitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 773
Bacterial Vasculitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 774
Tuberculous Vasculitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 774
Fungal Vasculitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 774
Neuro-Behcet’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 775
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 776
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 777

M.H. Machnowska (*) • R.I. Aviv

Division of Neuroradiology, Department of Medical
Imaging, School of Medicine, Sunnybrook Health
Sciences Centre, University of Toronto, Toronto,
ON, Canada
e-mail: Matylda.machnowska@sunnybrook.ca;
Richard.aviv@sunnybrook.ca

# Springer Science+Business Media New York 2016 751

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_23
752 M.H. Machnowska and R.I. Aviv

Abstract
Fabry’s Disease
This chapter discusses a variety of inflamma-
Fabry’s disease is a multisystemic X-linked lyso-
tory, infectious, and not otherwise easily cate-
somal storage disorder which results from muta-
gorized diseases that can affect the intracranial
tions in the α-galactosidase A gene (GLA) leading
vessels. These include diseases with known
to deficiencies of α-galactosidase A.
genetic mutations such as Fabry’s disease,
cerebral autosomal-dominant arteriopathy
with subcortical infarcts and leukoence-
Epidemiology
phalopathy (CADASIL), and cerebral
autosomal recessive arteriopathy with
Estimates of the prevalence of Fabry’s disease in
subcortical infarcts and leukoencephalopathy
younger stroke patients (<55 years) range up to
(CARASIL); those arising as a result of genetic
4 %. A recent prospective study including the
and environmental factors such as moyamoya;
largest cohort of patients to date reported a lower
and those with yet unknown etiologies such as
prevalence of 0.5 % overall (0.8 % in females and
Takayasu’s, posterior reversible encephalopa-
0.4 % in males) [1]. A more conservative estimate
thy syndrome (PRES) and Behcet’s disease. In
of disease prevalence is also reflected in a recent
each case, the preferred imaging methods and
study of young patients with expected neurologic
characteristic imaging findings are discussed.
hallmarks of the disease, including not only stroke
We concentrate on the standard imaging
but also transient ischemic attack, intracranial
methods of MRI, cerebral angiography, and
hemorrhage, unexplained white matter lesions,
noninterventional angiography. Where appli-
and vertebrobasilar dolichoectasia [2].
cable, advanced imaging techniques including
Although an X-linked disorder, lyonization in
cerebrovascular reactivity, vessel wall imag-
heterozygous females includes presentations
ing, perfusion imaging, and MR spectroscopy
ranging from asymptomatic to symptom severity
are discussed.
similar to males. Generally, disease in females is
characterized by less kidney involvement but sim-
ilar prevalence of stroke or TIA. Patients usually
Keywords
have their first stroke between 20 and 50 years of
Angiography • Functional imaging • Vessel
age, with males presenting a decade earlier than
wall imaging • Autoimmune vasculitis •
females.
RCVS • PRES • Infectious vasculitis •
Takayasu’s • Moyamoya • CADASIL
Clinical Presentation

Introduction Early clinical symptoms include acroparesthesia,

This chapter discusses a plethora of autoimmune,
infectious, and not easily categorized CNS and
systemic diseases, which result in neurological
symptoms primarily as a result of their effects on
the intracranial vasculature. We aim to review the
current clinical and pathological understanding of
the diseases and to discuss standard, and, where
applicable, advanced imaging methods currently
employed in the clinical setting.
present in 60–80 % of classically affected chil-
dren, waning in adulthood; anhidrosis leading to
heat and exercise intolerance; angiokeratoma;
corneal opacities; gastrointestinal symptoms
including abdominal pain, diarrhea, and vomiting;
and tinnitus. Cardiac, renal, and central nervous
system abnormalities often present outside of
childhood. Affected adolescent males have
shown decreased heart rate variability suggesting
autonomic dysfunction. Cardiac findings in adults
33 Miscellaneous Vascular Diseases
include left ventricular hypertrophy, arrhythmia,
and angina. Renal manifestations include
microalbuminuria and proteinuria in the second
and third decades progressing to azotemia in the
fifth to sixth decades. Symptoms due to CNS
involvement, including headache, vertigo, cogni-
tive impairment, and ischemic strokes, usually
present in adulthood.
Pathophysiology
Artery-associated complications give rise to
nephropathy and CNS. The etiology of the
vasculopathy is not entirely clear, although sev-
eral mechanisms have been postulated. These
include deposition of glycolipid moieties, such
as globotriaosylceramide (Gb3) in the lysosomes
of smooth muscle and endothelial cells; an
increase in prothrombotic factors and inflamma-
tory mediators; and altered cerebrovascular flow.
In addition to lysosomal deposition, Gb3 is also
found in Schwann cells, in dorsal root ganglia,
and in CNS neurons.
Fabry patients show increased soluble
intercellular adhesion molecule-1 (sICAM-1),
soluble vascular cell adhesion molecule-1
(sVCAM-1), and macrophage 1 antigen MAC-1
levels suggesting an enhanced inflammatory state.
Increased levels of plasminogen activator inhibi-
tor, and reduced thrombomodulin concentrations
compared to controls, suggest an enhanced coag-
ulant state [3].
Several PET studies have shown evidence of
cerebral hyperperfusion, particularly in the poste-
rior cerebral circulation [4]. This is in contradis-
tinction to classic leukoaraiosis, which is usually
caused by hypoperfusion. Studies have also
shown disordered autoregulation which may con-
tribute to the white matter disease [5].
CT Findings
Hypodensities, corresponding to white matter
lesions, can be seen in the deep and
753
periventricular white matter. There is evidence of
end-arterial ischemic lesions, more prominent in
the posterior territories. Dystrophic calcification is
seen in the basal ganglia, pulvinar of the thalamus,
and subcortical arcuate fibers of the white
matter [6].
MRI Findings
Structural MRI findings show T2 hyperintense
white matter lesions corresponding to the distri-
bution seen on CT. Ischemic lesions, predomi-
nantly in the vertebrobasilar territory, are also
seen. Pulvinar T1 hyperintensity has been
suggested as a very specific finding for Fabry’s
disease, present in about 30 % of patients over the
age of 50 [6]. This pulvinar T1 hyperintensity
does not show signal loss on fat-saturated imag-
ing. Additionally, signal drop seen in the pulvinar
on susceptibility weighted imaging corresponds
with the degree of calcification on CT suggesting
that mineralization of the pulvinar is responsible
for pulvinar T1 hyperintensity (Fig. 1). The intra-
cranial vessels are dolichoectatic [7].
Several studies have shown increased white
matter diffusivity on diffusion-weighted and dif-
fusion tensor imaging, predating structural
changes. A recent MR study comparing white
matter lesion load, white matter diffusivity
changes, and arterial diameters, but not the T1
hyperintense pulvinar sign, has shown that
increased basilar artery diameter is the best dis-
criminator between patients with Fabry’s disease
and normal controls [8].
PRES
The term posterior reversible encephalopathy syn-
drome is a clinical and radiologic syndrome with
specific imaging findings, characterized by
reversible, mostly posterior territory, vasogenic
edema, which is the end point of a myriad of
different etiologies.
754
Fig. 1 Sagittal T1-WI and
SWI images (a, b) showing
pulvinar T1 hyperintensity
and SWI hypointensity
(arrowheads) in a 30-year-
old male with Fabry’s
disease (Images courtesy of
D. Mandell)
Other names used to describe the clinical or
radiologic syndrome include hypertensive
encephalopathy and reversible posterior
leukoencephalopathy. Most authors prefer the
term posterior reversible encephalopathy because
up to 30 % of patients are not hypertensive at
presentation, and imaging changes are often also
present in the cortex and not limited to the white
matter. Symptoms include headaches, decreased
alertness, altered mental functioning, seizures,
and visual loss occasionally attributable to cortical
blindness.
The clinical and imaging findings have
been associated with preeclampsia/eclampsia,
posttransplant disorders, renal failure, immune
suppression, infection and sepsis, autoimmune
disease, chemotherapeutic agents, and numerous
other miscellaneous conditions.
Pathophysiology
Although several theories of PRES etiology
exist, two main hypotheses are likely.
The original theory suggests that hypertension
leads to autoregulatory vasoconstriction and
hypoperfusion, ischemia, and edema. A more
popular theory suggests that severe hypertension
results in breakthrough of the autoregulatory
mechanism leading to edema.
There has been a recent resurgence of interest
in the original theory, as a result of several studies
M.H. Machnowska and R.I. Aviv
showing that about 30 % of patients who develop
symptoms and imaging findings compatible with
PRES are not hypertensive or demonstrate cere-
bral hypoperfusion. Further, several of the under-
lying diseases or agents associated with the
development of PRES affect endothelial cell
function [9].
Pathology
The white matter and cortical edema tends to
concentrate in the parieto-occipital regions.
Pathology reports of biopsy specimens have
shown findings of white matter vacuolation, con-
sistent with white matter edema often with mild
inflammatory reaction and, less commonly, find-
ings of neuronal necrosis and petechial hemor-
rhages [10]. Autopsy specimens have shown
fibrinoid necrosis of arterioles with thrombosis
of arterioles and capillaries, as well as parenchy-
mal microinfarcts and petechial hemorrhages
thought to be secondary to vascular changes.
Occasionally, pathology specimens have revealed
evidence of white matter pallor, attributed to
demyelination and astrocytosis [11].
More recent pathology reports have revealed
evidence of perivascular T-cell trafficking, endo-
thelial activation, endothelial immunoreactivity to
tumor necrosis factor-alpha, robust endothelial
VEGF mRNA transcription, and VEGF protein
expression [9].
33 Miscellaneous Vascular Diseases
Imaging Findings
CT
CT typically demonstrates focal regions of sym-
metric, predominantly subcortical white matter
parieto-occipital hypodensity. The frontal lobes,
cerebellum, and inferior temporal-occipital lobes
may also be affected. The distribution does not
conform to a vascular territory but is often seen in
watershed regions. The cortex may be involved.
MRI
MRI shows T2 prolongation in the subcortical
white matter and cortex almost always involving
the parieto-occipital lobes. Isolation solely to the
posterior brain regions is rare, however, and
involvement of other brain regions including the
frontal lobes, temporal lobes, and cerebellum is
Fig. 2 PRES encephalopathy. (a, b) Hyperintensity is
seen on T2-WI images with subcortical and cortical
involvement in the parieto-occipital regions (arrowheads).
(c, d) DWI is normal due to T2 wash out. (e) ADC map
755
often seen. Thalamic, basal ganglia, and
brainstem involvement has been described.
Findings are most commonly associated with
hypointensity or isointensity on diffusion-
weighted imaging with concomitantly increased
ADC values. Foci of predominantly cortical dif-
fusion restriction have been described. These may
represent ischemia and irreversible injury. Revers-
ible cortical diffusion restriction has also been
described in the setting of seizure activity, pre-
sumably related to excitotoxic edema [12] (Fig. 2)
Contrast enhancement has been shown in
16–30 %. Petechial and subarachnoid hemor-
rhages (SAH) occur in 15 % of lesions
[10]. FLAIR sulcal hyperintensity may be second-
ary to SAH but can also occur secondary to pro-
teinaceous exudate. Diffusion tensor imaging has
revealed regions of reversible increased
anisotropy [13].
confirms vasogenic edema in the distribution of T2
hyperintensity (black arrow). (f) MRV is normal (Images
courtesy of M. Shroff)
756
Catheter Angiography
Angiographic studies range from normal to find-
ings of vasospasm evidenced by arterial
narrowing that can be peripheral, central, or dif-
fuse. Arterial beading has also been described.
Peripheral vasospasm is often difficult to charac-
terize on MRA/CTA imaging and is better visual-
ized with catheter angiography [14].
Perfusion Studies
CTP studies, performed in hypertensive patients,
have shown increased cerebral blood flow,
increased cerebral blood volume, and decreased
mean transit time allowing for differentiation
between PRES and cerebrovascular accident [15].
Hypoperfusion, however, has been demon-
strated using Tc99m-HMPAO SPECT in women
with eclampsia, performed within 48 h of
delivery [16].
Several recent MR perfusion studies have also
shown decreased perfusion in PRES [14]. Find-
ings have generally shown decreased cerebral
blood flow and volume in affected regions, with
some variability of mean transit time.
The discrepancy of these perfusion findings is
not well understood. It may relate to the timing of
the perfusion study with respect to disease onset.
Animal experiments have shown that vasocon-
striction occurs within the limits of
autoregulation, at which point there is break-
through of the autoregulatory mechanism
resulting in hyperperfusion and breakdown of
the blood-brain barrier. In animal studies, this
has been followed by vasodilation and
hypoperfusion. Alternatively, authors postulate
that abnormal radiopharmaceutical uptake could
represent luxury perfusion in the setting of loss of
autoregulation and vascular tone, rather than true
hyperperfusion.
RCVS
Reversible cerebral vasoconstriction syndrome
(RCVS) is defined as a group of disorders that
result from prolonged but reversible vasoconstric-
tion of the cerebral arteries. It is usually associated
M.H. Machnowska and R.I. Aviv
with acute onset, severe headaches and may be
accompanied by other neurological signs. Spon-
taneous resolution of the vasoconstriction is usu-
ally seen in 1–3 months [17]. In its broadest
definition, the term encompasses disorders of
numerous etiologies characterized by reversible
vasospasm which do not show histologic hall-
marks of vasculitis. RCVS has been known in
the literature by multiple names including Call-
Fleming syndrome, benign angiopathy of the
central nervous system, thunderclap headache
with reversible vasospasm, and migrainous
angiopathy. Reversible vasospasm in the setting
of known etiologic factors such as drug-induced
vasculopathy, hypercalcemia, unruptured saccular
aneurysm, dissection, postcarotid endarterectomy,
and postpartum angiopathy is also included in the
umbrella term of RCVS [17]. Known etiologic
factors account for up to 25–60 % of cases of
RCVS [18].
Epidemiology
The age of onset of RCVS peaks at around
45 years of age and is more common in females,
with a 2.6:1 preponderance in a French study to
10:1 in Taiwanese study [18].
Clinical
Presentation is similar to aneurismal rupture and
characterized by rapid onset thunderclap head-
ache, peaking in less than 1 min and recurring
over 1–2 weeks.
Diagnostic criteria were defined by Calabrese
and include: [17]
1. Transfemoral angiography or indirect (CT or
MRI) angiography documenting segmental
cerebral artery vasoconstriction.
2. No evidence of aneurismal subarachnoid
hemorrhage.
3. Normal or near-normal cerebrospinal fluid
analysis (protein level <80 mg/dL, white
blood cell count <10/μL, normal glucose
level).
33 Miscellaneous Vascular Diseases
Fig. 3 Reversible cerebral vasoconstriction syndrome. A
36-year-old female presented with an acute thunderclap
headache. (a) CT shows occipital parenchymal (arrow)
and subarachnoid hemorrhage. (b, c) CTA performed the
same day shows beading within middle cerebral artery
(MCA) and anterior cerebral artery (ACA) territories
(arrowheads). Lumbar puncture and serum biochemistry
4. Severe, acute headache, with or without addi-
tional neurological signs or symptoms.
5. The diagnosis cannot be confirmed until revers-
ibility of the angiographic abnormalities is
documented within 12 weeks of onset. If death
occurs before the follow-up studies are com-
pleted, autopsy should rule out conditions such
as vasculitis, intracranial atherosclerosis, and
aneurismal subarachnoid hemorrhage, which
can also manifest with headache and stroke.
CSF analysis must be performed to rule out
subarachnoid hemorrhage in patients presenting
with thunderclap headache, with normal imaging
findings. CT and MRI are usually normal.
757
results revealed no evidence of an inflammatory process.
The patient was diagnosed with RCVS, and steroids were
withheld. (d) MRI T2-WI study performed 3 months later
shows residual hemosiderin deposition in the left occipital
lobe. (e) Concurrent MRA shows complete resolution of
arterial narrowing
Occasionally, there may be a small amount of
subarachnoid hemorrhage overlying the cortical
surface or in the parenchyma. In 10–39 % of
patients, findings may result in ischemic or hem-
orrhagic parenchymal strokes [19] (Fig. 3a, d).
Reversible brain edema, compatible with PRES,
is seen in about 10 % of patients [19].
CTA and MRA usually show segmental
beading and intermittent dilatation of vessels
(string of beads) [17] (Fig. 3b, c). Initial
noninvasive angiography may be negative or
show distal vessel irregularities. Maximum
vessel constriction involving the proximal
arterial beds usually occurs 18–22 days following
headache onset, often at the time of headache
758
resolution, and may persist for several weeks.
This has led some authors to speculate that it is
actually the vascular dilation, and not constric-
tion, which is responsible for pain and microvas-
cular rupture resulting in hemorrhage [20]. The
gold standard in diagnosis remains catheter angi-
ography, which has shown a sensitivity of 100 %
compared to noninvasive methods (MRA and
CTA) which showed sensitivities of around
75 % [19].
Recent studies have shown an association
between RCVS and vertebral or carotid artery
dissection, particularly in postpartum females,
and it may be prudent to screen the neck vessels,
particularly in patients who present with neck
pain [19].
The foremost mimicker of RCVS is primary
angiitis of the CNS (PACNS), while the clinical
symptoms are usually more gradual in onset in
angiitis. Headache is usually subacute and pro-
gressive, rather than thunderclap type, and may
be accompanied by transient neurological deficits
eventually progressing to stroke. Asymptomatic
infarcts are often seen on initial MR studies.
Angiographic findings may lead to clinical confu-
sion due to their similarities in the two
pathologies.
Distinguishing features are that in at least 70 %
of cases, MRI will be normal in RCVS, while it
will be abnormal in PACNS in up to 90 % of
cases showing multiple, deep, or superficial
infarcts of different ages. CSF analysis should be
normal in RCVS and abnormal in over 95 % of
PACNS cases. Irregular, eccentric, and asymmet-
ric occlusions without interspersed dilations
are more suggestive of a vasculitis rather than
RCVS [18]. However, such subtle angiographic
differences may be difficult to appreciate,
particularly due to the limits in spatial resolution
of noninvasive angiography. Vessel wall
imaging, which utilizes a single inversion
recovery-prepared fast spin echo black blood
T1-WI sequence prior to and following
contrast administration, has shown promise in
distinguishing the two entities [21]. Vessel wall
enhancement is seen with inflammatory
processes affecting the vessel lumen and is absent
in RCVS.
M.H. Machnowska and R.I. Aviv
Moyamoya
Moyamoya syndrome is an occlusive cerebrovas-
cular disease which affects the terminal internal
carotid arteries and proximal circle of Willis ves-
sels. Associated engorgement of the lenticulostriate
and thalamoperforating leads to the moyamoya or
“puff of smoke” appearance on angiography.
Moyamoya disease refers to the idiopathic form
of the arteriopathy; moyamoya syndrome refers to
an association with other pathologic processes
including neurofibromatosis, sickle cell anemia,
Down syndrome, and prior cranial irradiation.
Epidemiology
Incidence of moyamoya disease in Japan has
range up to 0.94/100,000 of the population, with
increasing incidence reported in recent years
attributed to better diagnostic measures
[22]. Rates of moyamoya syndrome in the United
States and Europe are generally one tenth those in
Japan [23]. The disease has a bimodal distribution
with peaks in the first and fourth decades of life
[23]. Both symptomatic and asymptomatic forms
of the disease have a slight female to male pre-
dominance of about 2:1 [22].
Children generally present with ischemic
events such as TIAs or strokes, while adult
patients more commonly present with hemor-
rhage. The disease accounts for 6 % of strokes
in children [24].
Clinical associations with moyamoya include
skull base, sellar, and suprasellar tumors and optic
nerve gliomas either idiopathically or as a result of
radiation treatment, Fanconi anemia, sickle cell
anemia and other hemoglobinopathies, neurofi-
bromatosis type I, progeria, inflammatory factors
such as anticardiolipin syndrome, protein S eleva-
tion, SLE, neuro-Behcet’s, Kawasaki disease;
collagen vascular disorders such as Marfan, con-
genital cardiac disease, renal artery stenosis,
infectious diseases such as tuberculosis,
leptospirosis, tonsillitis, atherosclerosis, and
fibromuscular dysplasia (Fig. 4). These clinical
associations can be seen in the setting of unilateral
disease.
33 Miscellaneous Vascular Diseases
Fig. 4 Conventional angiogram showing left MCA steno-
sis in a child with neurofibromatosis representing
moyamoya syndrome (Courtesy of M. Shroff, G. DeVeber)
Genetics
The familial form of disease accounts for about
15 % of patients.
To date one gene locus has been implicated in
moyamoya: the ring finger protein 213 (RNF213)
gene located on chromosome 17q has been iden-
tified as a susceptibility gene for moyamoya dis-
ease. Other gene foci identified by chromosomal
linkage analyses include 3p24.2-26, 6q25, 8q23,
and 12p12. The inheritance pattern is thought to
be autosomal dominant with incomplete
penetrance [25].
Clinical Presentation
Children with moyamoya typically present with
evidence of TIAs (40 %) or infarction (40 %)
[26]. Motor disturbances vary from weakness to
paralysis of the extremities. Other presentations
include headaches, seizures, and decreasing
cognitive function. Paroxysmal symptoms have
been seen in the setting of elevated body temper-
atures, exercise or episodes of crying, and hyper-
ventilation leading to hypocapnia and
vasoconstriction.
759
Forty to 65 % of adults present with hemor-
rhage usually localized to the basal ganglia and
ventricular system. Hemorrhages are often recur-
rent and more common in females [27].
Hemodynamic studies, discussed below, have
shown a decreased cerebral blood flow in children
with moyamoya in comparison to age-matched
controls and adults with the disease, an increased
cerebral blood volume suggesting a chronic state
of vasodilation, and an increased oxygen extrac-
tion fraction and significantly impaired cerebro-
vascular reactivity [28]. These findings may help
to explain why ischemia is the most common
presenting symptom in children.
Pathophysiology
Moyamoya disease is characterized by complex
channels of medium- and small-sized muscular
arteries branching off the circle of Willis, anterior
choroidal arteries, internal carotid arteries, and
posterior cerebral arteries and connecting to the
distal portions of the anterior and middle cerebral
arteries. These channels give rise to numerous,
dilated or stenotic, tortuous perforating vessels
that correspond to the lenticulostriate and
thalamoperforating vessels. Dilated vessels are
more common in younger patients, while stenotic
vessels in older patients.
Microscopic findings in stenotic vessels are
characterized by intimal thickening, thickening
and duplication of the elastic lamina, fibrosis of
the tunica media, and fibroelastosis of the wall.
The media is thinned, and there is a decrease in the
external diameter of the vessel [29]. Dilated ves-
sels show fibrosis of the media, attenuation of the
wall, and moderate increase in elastic fibers
[29]. Lipid deposits have been found, but are
thought to relate to atherosclerosis.
Hemorrhage is attributed to rupture of the per-
forating vessels, aneurismal rupture, or very rarely
rupture of dilated collateral arteries on the brain
surface. Frequency of intracranial aneurysms is
increased in patients with moyamoya. Aneurysms
can be saccular or dissecting and occur centrally,
most frequently in the basilar arteries or
peripherally [29].
760
Imaging
Angiography
Catheter angiography remains the gold standard
in moyamoya diagnosis. While MRA is an accept-
able alternative in diagnosing moyamoya, cathe-
ter angiography is recommended in adult surgical
candidates, with pediatric cases considered on an
individual basis [30]. The vascular picture is char-
acterized by stenosis of the supraclinoid distal
ICA, the proximal MCA and ACA branches, and
the development of dilated striate arteries.
Involvement of the posterior cerebral artery
(PCA) is seen in about 30 % of patients, usually
at the late stage of disease.
Six angiographic stages are described: [31]
1. Narrowing of the carotid termination.
2. Dilation of the proximal ACA and MCA with
initial moyamoya blush and dilation of the
perforating moyamoya vessels.
3. Stenosis of the ACA and MCA with intensifi-
cation of the moyamoya blush.
4. Minimization of the moyamoya blush with exten-
sion of internal carotid artery (ICA) occlusion to
the junction of the posterior communicating artery
(PComm). The constituent vessels are thinned,
and there is a concomitant increase in the collat-
eral vessels from the extracranial carotid arteries.
5. Further reduction of the moyamoya and com-
plete disappearance of the main arteries arising
from the ICA. Occlusion of the ICA can reach
as far as C2. Persisting increase in ECA collat-
eral circulation.
6. Disappearance of the moyamoya, with intra-
cranial supply from the ECA and potentially
vertebrobasilar system.
Angiography can be used to identify aneu-
rysms, define a specific transdural anastomotic ves-
sel, and, in very rare cases, identify other vascular
anomalies. Investigation of the ECA system for
stenoses is also recommended and may have impli-
cations for treatment success rates [30] (Fig. 5).
CT
Unenhanced CT may be completely normal or
show subarachnoid and intraparenchymal
M.H. Machnowska and R.I. Aviv
hemorrhage or atrophy and infarction in the ante-
rior circulation territory.
CTA
CTA is a powerful tool in the diagnosis of
suspected vascular abnormalities in the setting of
intracerebral hemorrhage. It can be used for the
assessment of intracranial stenoses and the identi-
fication of aneurysms and areas of contrast
enhancement. It has also been used as a method
for neurosurgical planning and assessment of
EC-IC bypass.
MRI
MRI findings include atrophy, particularly in the
anterior circulation territory. Evidence of prior
infarcts or hemorrhage can be seen on FLAIR and
gradient-recalled echo imaging. Asymptomatic
microbleeds, in the periventricular deep white mat-
ter, have been seen in up to 44 % of patients [32].
Diffusion-weighted imaging will detect acute
infarcts. The flow voids at the skull base should be
carefully scrutinized for evidence of ICA and
proximal ACA/MCA diminution. Increased flow
voids can also be seen within the basal cisterns,
basal ganglia, and around the circle of Willis
consistent with increased collateral circulation
(Fig. 5a).
Increased intersulcal FLAIR signal
corresponding to increased contrast in the sub-
arachnoid space on enhanced T1-WI images has
been termed the “ivy sign” and is felt to represent
slow flow in dilated pial collaterals (Fig. 5b).
MRA
According to the guidelines for diagnosis and
treatment of moyamoya disease, MRA can pro-
vide a definitive diagnosis when the following
findings are fulfilled on time of flight (TOF) imag-
ing using a scanner of at least 1.5-T.
1. Stenosis or occlusion of the terminal portion of
the intracranial internal carotid artery or prox-
imal portion of the anterior and/or middle cere-
bral arteries.
2. Abnormal vascular networks in the basal
ganglia. When two or more visible flow voids
in the basal ganglia are present at least
33 Miscellaneous Vascular Diseases 761

Fig. 5 Moyamoya. (a, b)

Axial FLAIR images show
multiple flow voids in the
basal ganglia (white
arrowheads), white matter
leukoaraiosis in the
watershed regions, and the
ivy sign in the right
hemispheric sulci (arrow).
(c) MRA shows numerous
leptomeningeal and
lenticulostriate collaterals
(black arrowheads), which
show retrograde filling of
the distal right MCA on
conventional angiography,
(d) (Courtesy of M. Shroff,
G. DeVeber)

unilaterally on MRI, they can be deemed as of cerebrovascular reactivity which is defined as a

representing an abnormal vascular network.
3. Bilaterality of findings 1 and 2.
In cases of unilaterality, the guideline authors
consider a quasi-moyamoya or moyamoya
syndrome [33].
Functional and Hemodynamic Studies
In general, hemodynamic studies show decreased
cerebral blood flow, increased mean transit time,
and increased cerebral blood volume in the
affected hemisphere, coupled with an increased
oxygen extraction fraction [30]. The cerebral met-
abolic rate of oxygen is often decreased, reflecting
the influence of limited cerebral blood flow, par-
ticularly in ischemic cases [34]. Many perfusion
studies are performed at baseline and following a
vasodilatory challenge using acetazolamide or
end-tidal CO2 manipulation. The use of a
vasodilatory challenge allows for the calculation
change in blood flow per unit change in stimulus,
such as carbon dioxide. In the setting of a down-
stream stenosis, maintenance of normal cerebral
blood flow is possible due to a relaxation of vas-
cular tone, resulting in normal baseline perfusion
parameters. Analogous to the myocardial stress
test, a cerebrovascular reactivity study uncovers
the compensatory vasodilatory state by stressing
its limits. The strengths and limitations of the
currently utilized functional and hemodynamic
studies are discussed below.
PET is the only imaging modality that can
measure oxygen extraction fraction and the cere-
bral metabolic rate of oxygen consumption. Its
limitations include the need for an on-site cyclo-
tron due to the rapid decay of 15O (2 min half-life)
and a PET scanner, the availability of which is less
than MR, CT, or SPECT. The whole body radia-
tion exposure is 0.5–2.0 mSv, depending on the
quality of the PET camera and protocol. Increased
762
oxygen extraction fraction is linked to progression
of disease and reversed with EC-IC bypass.
CT and MR are relatively easily accessible, but
technically difficult in moyamoya patients due to
the lack of availability of a reliable arterial input
function in the diseased anterior circulation vessels,
occasionally necessitating the placement of ROIs
in posterior circulation vessels [35]. The presence
of multiple diseased vessels precludes the accurate
comparison of concentration time curves derived
from one major vessel to those derived in the
microvascular territories supplied by several differ-
ent diseased vessels. Flow maps will only be accu-
rate for the territory of the vessel in which the
arterial input function was positioned, but not for
those tissues fed by other diseased vessels [36]. A
recent study of 16 patients showed poor angio-
graphic stage correlation with normalized baseline
CT perfusion parameters (with respect to ipsilat-
eral cerebellar hemisphere), but significant corre-
lation with percent changes of CBF during an
acetazolamide challenge [37].
SPECT uses the single-pass lipophilic isotopes
99m
technetium and 123iodine that cross the blood-
brain barrier. The use of single-pass isotopes permits
an estimation of blood flow, due to first pass depo-
sition through the microcirculation, without provid-
ing quantitative measurements. Measurements are
weighted to tracer delivery/deposition, but immune
to the input function problem of CT and MR perfu-
sion studies. While technetium isotope use alone
requires a 24 h interval between baseline and acet-
azolamide challenge, the combination of technetium
and iodine allows the study to be performed in a
single sitting using separate windowing for the
99m
technetium and 123iodine keV [27].
A recently developed MRI technique uses
blood oxygenation level-dependent contrast
(BOLD) imaging while manipulating the
end-tidal CO2 (ETCO2) through a rebreathing
circuit [38]. This method does not rely on input
functions or pharmacologic agents and provides
spatial and quantitative maps of percent BOLD
signal change per mm Hg change in ETCO2. It can
be performed on all 1.5 and 3-T MRI systems with
echoplanar capability. In normal tissues, BOLD
signal increases with hypercapnea due to a wash-
out of deoxyhemoglobin. The paramagnetic
M.H. Machnowska and R.I. Aviv
effects of deoxyhemoglobin normally cause
dephasing of protons in local tissues, reducing
the returned BOLD signal. This method not only
shows patients in whom CVR is exhausted but
also those with a paradoxical response or steal
phenomenon in patients with uncompensated
moyamoya disease. CVR defects may also reverse
with EC-IC bypass.
Proton MR spectroscopy has shown reduced
levels of choline, creatine, and N-acetylaspartate
in the affected brain of moyamoya patients with
relative increase 6 months following revasculari-
zation surgery [39].
Cerebral Autosomal Dominant
Arteriopathy with Subcortical Infarcts
and Leukoencephalopathy (CADASIL)
Cerebral autosomal-dominant arteriopathy with
subcortical infarcts and leukoencephalopathy is a
genetic small artery disease that leads to dementia
and early death.
Its main clinical features include migraine with
aura, ischemic events, mood disturbances, and
cognitive impairment. Other presentations such
as apathy or seizures have also been described.
Dementia and dependence are prominent features
of the end stages of the disease [40].
Epidemiology
The prevalence of the disease is reported as 2–4
per 100,000. The mean age of disease onset is in
the mid third decade, with earlier discovery of the
disease in patients presenting with migraine
(mean age 28 years) than those with stroke
(mean age 41 years) [41]. The median age of
death is reported as 65 years for males and
71 years for females, with 80 % of patients depen-
dent by the time of death [40].
Genetics
CADASIL is a genetic disorder due to mutations
in the extracellular domain of the Notch 3 gene
33 Miscellaneous Vascular Diseases
Fig. 6 CADASIL Axial
FLAIR images show
widespread T2 signal
hyperintensity in the
periventricular, deep, and
subcortical white matter,
volume loss, temporal pole
involvement (arrows), and
subcortical lacunar infarcts
(arrowhead)
located on chromosome 19. The gene contains
33 exons, but only mutations in exons 2–24
occur in CADASIL, 60 % occurring in exons
3 and 4.
Pathology
Macroscopic examination shows diffuse myelin
pallor sparing the U-fibers. Lacunar infarcts are
seen in the white matter and basal ganglia. Pontine
involvement is most common in the brainstem,
with relative sparing of the medulla. Microscopic
examination reveals a classic non-amyloid granu-
lar osmiophilic material in the media extending
into the adventitia of small- and medium-sized
leptomeningeal arteries.
Fibrosis of the adventitia and sometimes dupli-
cation and hyalinosis of the lamina are described.
Loss of vascular smooth muscle cells occurs but to
a lesser extent than in Binswanger’s dementia
[42]. Apoptosis in layers 3 and 5 of the cortex is
associated with increased severity of white matter
lesions [43].
MRI Imaging
MR imaging shows areas of FLAIR
hyperintensity in the white matter, pons, and
external capsules (Fig. 6).
763
Microbleeds are seen in 25–69 % of patients
and were present most frequently in the thalami,
cortical/subcortical regions, white matter, pons,
and basal ganglia [44, 45]. Number does not cor-
relate with white matter lesion burden, but is
increased with age. Similarly, dilated Virchow-
Robin spaces, mainly in the lentiform nuclei and
subcortical white matter of the temporal lobes, are
seen significantly more frequently in CADASIL
patients than age-matched controls and show a
correlation with increasing age, not white matter
lesion burden [46].
Recent studies have shown a correlation with
the disease burden at diagnosis and severity of
progression [44].
CARASIL
Cerebral autosomal recessive arteriopathy with
subcortical infarcts and leukoencephalopathy is a
recently recognized single gene disorder caused
by mutations in the HTRA1 gene. Its prevalence is
currently unknown, with approximately
50 patients described in the literature to date,
mostly from Japan with a 3:1 male to female ratio.
Its main clinical manifestations are ischemic
stroke, progressive dementia, early alopecia, and
spondylosis deformans. Ischemic stroke occurs in
about half of patients and the onset of dementia is
764
in the fourth decade (mean 32 years). Alopecia,
seen in 90 % of patients, usually presents in the
third decade, but may be seen in adolescence.
Low back pain is seen in 80 % of patients, with
an onset between 20 and 40 years.
CT shows diffuse white matter hypodensity,
volume loss, and corticospinal tract Wallerian
degeneration. Similarly, MRI studies show T2
hyperintense periventricular and deep white mat-
ter lesions, with some sparing of U-fibers. Lesions
subsequently extend into basal ganglia, thalami,
brainstem, and cerebellum. Temporal lobe
involvement may be seen, similar to
CADASIL [47].
Vasculitis
PACNS
Primary angiitis of the CNS (PACNS) is a rare,
potentially severe, form of vasculitis that involves
the small- (200–300 μm in diameter) and medium-
sized (500 μm) vessels of the brain and spinal
cord. PACNS is also referred to as granulomatous
angiitis, granulomatous giant cell angiitis,
noninfectious granulomatous angiitis, and iso-
lated angiitis. Its current name reflects the fact
that many cases have a non-granulomatous
histology.
Epidemiology
PACNS has an estimated incidence of 2.4 per
1,000,000 inhabitants, with a slight female pre-
dominance [48, 49]. Although recent large cohort
studies state a slight female predominance, cau-
tion must be taken in interpreting this data, as
several studies include patients with angio-
graphic, but not pathologic, proof of disease and,
therefore, may occasionally include patients with
benign angiopathy of the CNS, who were once
characterized as a subset of PACNS with a more
favorable prognosis and have since been
reclassified as RCVS. Earlier studies showed
either a slight male predominance of 4:3 or no
sex predilection. The median age of onset is about
50 years [49].
M.H. Machnowska and R.I. Aviv
Clinical Symptoms
Most frequent symptoms include headache,
altered cognition, focal neurological manifesta-
tions, stroke, and visual symptoms [48]. Onset is
usually gradual and progressive.
The largest series includes 101 patients, show-
ing no statistically significant correlation with
male gender and increased mortality rate or higher
disability scores (modified Rankin scale 4–6).
Instead, presentation with a focal neurological
deficit rather than headache, cognitive impairment
versus headache or constitutional symptoms,
cerebral infarction versus no infarction, and
large-vessel versus small-vessel involvement sig-
nificantly correlated with increased mortality.
The disease in children may more often present
with seizures [49].
Pathology
Biopsy of the brain or spinal cord is considered
the gold standard. Biopsy results are often nega-
tive in 25–50 % of cases [48]. Including
leptomeninges in the biopsy results in slightly
higher diagnostic yields [49].
Three patterns are recognized: granulomatous,
lymphocytic, and necrotizing. Findings include
inflammatory infiltrates of giant cells, plasma
cells, lymphocytes in the vessel wall accompanied
by necrosis, or granuloma formation. A sparse
perivascular mononuclear cell infiltrate is
insufficient for diagnosis as it can be seen in
many other conditions. Necrotizing vasculitis is
associated with the destruction of the internal
elastic lamina.
CSF analysis is abnormal in 80–90 % of cases.
ESR and CRP are more likely to be elevated in
children [49].
Imaging
CT imaging will show nonspecific findings of
vasculitis such as infarcts or hemorrhages and is
less sensitive than MRI.
MRI imaging is abnormal in nearly all cases,
even those that are angiography or biopsy nega-
tive, and most commonly shows cortical and sub-
cortical infarcts, which are frequently multiple
and frequently involve both cerebral hemispheres.
Infarctions may occur in a large or perforator
33 Miscellaneous Vascular Diseases 765

Fig. 7 PACNS. Superior

cerebellar artery (SCA)
infarct shown on DWI (a)
secondary to stenosis of the
right SCA shown on the
right vertebral artery
injection (b) (Courtesy of
M. Shroff, G. DeVeber)

Fig. 8 PACNS. (a) DWI showing right MCA lenticulostriate territory infarction. (b, c) MR and digital subtraction
angiography show beading and marked stenosis of the right M1 and M2 vessels

artery distribution. Intracranial hemorrhage Angiography

occurs in about one tenth of patients [48]. Contrast
enhancement of the meninges, perivascular
spaces, and parenchymal mass lesions may occur
[48]. Although nonspecific, negative MRI and
negative CSF sampling are thought to have a
high negative predictive value for the presence
of vasculitis. The specificity of both tests is esti-
mated at 36 % (Figs. 7 and 8).
High-resolution pre- and postcontrast T1-WI
black blood imaging of the vessel walls has been
used to distinguish vasculitic processes from ath-
erosclerosis and RCVS (Fig. 9). Although this is a
novel technique, preliminary results have shown
concentric luminal enhancement in vasculitis,
eccentric enhancement in atherosclerosis, and
lack of enhancement in RCVS.
The classic angiographic signs of vasculitis are
alternating areas of stenosis and dilation (beading)
involving both large and small vessels. A recent
study showed angiographic abnormalities in 90 %
of patients with PACNS [48]. MRA and CTA
show similar findings, although with a lower sen-
sitivity owing to the limited spatial resolution of
500 μm, below the threshold of detecting small-
vessel vasculitides.
Takayasu
Epidemiology – 1–2/1,000,000, with more reports
initially in South East Asia, but recent reports also
in the western world.
766 M.H. Machnowska and R.I. Aviv

Fig. 9 Vessel wall imaging shows thickening (a, white arrowheads) and resultant collateralization through ECA
arrow) and enhancement (b, white arrowhead) of the cav- branches (black arrows). The patient presented with left
ernous ICA walls in a patient with rheumatoid arthritis and temporal lobe hemorrhage on SWI (e)
SLE. Marked bilateral ICA stenosis (c, d) (black

Pathology Clinical
This is a large-vessel vasculitis with predominant In addition to systemic symptoms of fever
involvement of the aorta and great vessels. Histo- and weight loss, the main neurological
logically, findings are those of a chronic large- symptoms attributable to Takayasu include
vessel vasculitis characterized by lymphocytic, dizziness, headache, vertigo, vision loss,
NK cell, neutrophil, and macrophage infiltration memory loss, TIAs, strokes, hypertensive
in the outer thirds of the media and adventitia. encephalopathy, and increased intracranial
Giant cells and granulomas have been described. pressure [50].
The elastic fibers are fragmented, and there can be CNS findings have also included
loss of smooth muscle cells, medial weakening, subarachnoid hemorrhage, most commonly due
and vascular dilation. Reactive fibrosis and to aneurismal rupture, but also non-aneurismal in
aneurismal formation is seen. etiology.
33 Miscellaneous Vascular Diseases
Fig. 10 A 6-year-old female with Takayasu’s arteritis.
Coronal MR angiogram shows aneurismal dilation of the
right ICA (a, arrowhead). Axial T1-WI, contrast-enhanced
image shows wall enhancement and a crescentic
intraluminal filling defect in the right ICA (b) corroborated
MRI
Arterial stenosis is the most common finding,
although dilation, occlusion, and aneurysm for-
mation can be seen. In the head and neck, the
subclavian and common carotid arteries are most
often involved. There may be rare involvement of
intracranial vessels. Initial experience using seg-
mented k-space inversion recovery prepared gra-
dient echo MR pulse sequences to obtain delayed
enhancement images of the aortic wall, suggests a
correlation between wall enhancement and active
inflammation [51]. MRI may also show evidence
of ischemia, white matter lesions, and findings
767
on the transverse ultrasound image (c). The patient
presented with concurrent posterior reversible leukoence-
phalopathy felt to be a result of renal artery involvement
(d) (Courtesy of M. Shroff)
compatible with posterior reversible encephalop-
athy [50] (Fig. 10).
Special Procedures
Panarteritis is seen in two thirds of patients with
Takayasu’s. Angiographic classification is based
on the segment involved.
Type I: aortic arch and its branches
Type II: descending thoracic and abdominal aorta
Type III: aortic arch and thoracoabdominal aorta
Type IV: pulmonary artery in addition to the afore-
mentioned types
768
Type V: coronary artery in addition to the other
types[52].
Angiography cannot distinguish between
active and chronic inflammation.
PET labeled with fluorodeoxyglucose (FDG) is
more sensitive than MRI for demonstrating the early
stages of Takayasu’s, with increased uptake in the
vessel wall seen during active inflammation [53].
Giant Cell Arteritis
The disease occurs almost exclusively in those
over 50, with highest prevalence in the seventh
decade. Women are affected two to three times
more commonly than men. The disease incidence
is around 10–29/100,000 [54].
Genetics
Many small cohort genetic studies have looked at
associations between genes encoding cytokines
and their receptors, adhesion molecules, and
chemokines; however, the most consistent associ-
ation is between GCA and the MHC class II
molecules, specifically carriage of the
HLA-DRB1*04 alleles. The latter has also been
associated with resistance to corticosteroid treat-
ment and greater risk of visual manifestation.
Pathophysiology
The disease is a panarteritis, involving all layers of
the vessel wall, the basis of which is an antigen-
driven inflammatory response [55]. The disease
stimulus is currently unknown. It is thought to
activate adventitial dendritic cells, which release
chemoattractants to recruit CD4+ T cells to the
adventitia-media border. T cells release INF-γ
and thereby lead to attraction and differentiation
of macrophages resulting in giant cell infiltration
and granuloma formation in the intima-media
junction. Inflammatory cytokine release by mac-
rophages promotes systemic inflammation [56].
Clinical
Because of the predominant involvement of exter-
nal carotid arteries and their branches, symptoms
M.H. Machnowska and R.I. Aviv
include bitemporal headaches, jaw claudication,
scalp tenderness, and most severely vision loss.
The latter occurs irreversibly in 10–15 % of
patients, due to acute anterior optic neuritis, reti-
nal artery occlusion, or retrobulbar optic neuritis
[57]. Systemic symptoms such as fever, weight
loss, and polymyalgia also occur.
Imaging
Rates of positive temporal artery biopsy are
50–90 % owing to the segmental nature of the
disease. Some authors claim the sensitivity of
imaging, if it includes not only the temporal arter-
ies but also the aorta and supra-aortic arteries, is
superior to biopsy [58].
Ultrasound
Color Doppler ultrasound was the first imaging
modality used to assess temporal arteritis and
currently has the highest imaging resolution of
all imaging techniques used for diagnosis.
T axial and lateral resolution is 0.1 mm using
high-frequency probes. Doppler ultrasound can
assess arterial stenoses, wall edema, and blood
flow velocities through the affected segments [59].
CTA
CTA may be helpful as a quick way of evaluating
the aorta and great vessels in patients with known
temporal artery involvement. A recent study
showed a 67 % occurrence of aortitis in newly
diagnosed GCA patients [60]. Similarly to con-
ventional angiography, its role in evaluating tem-
poral arteritis is limited.
MRI/MRA
3-T MRI has been quite helpful in the assessment
of the temporal artery walls using contrast-
enhanced fat-saturated spin echo T1-WI images,
achieving a sensitivity of 80.6 % and specificity of
97 % in a recent study [61].
Polyarteritis Nodosa
Polyarteritis nodosa (PAN) is a necrotizing vascu-
litis, affecting small- and medium-sized arteries,
never arterioles, capillaries, or venules. This is in
33 Miscellaneous Vascular Diseases
contradistinction to microscopic polyangiitis
(MAP), which has recently emerged in the litera-
ture, as a separate entity affecting mainly small
vessels. MAP has been shown to be an
antineutrophil cytoplasmic antibody (ANCA)-
associated vasculitis, associated more commonly
with myeloperoxidase (MPO) ANCA, but
manifesting as a pure vasculitis without granu-
loma formation.
PAN is characterized as a hyaline-like necrosis
of the media and internal elastic lamina, with exten-
sion of an inflammatory reaction to the arterial wall
and adjacent tissues. A substantial portion of PAN
patients are carriers of the HBV virus with recent
reports describing HCV association. HBV-PAN is
shown to have greater gastrointestinal organ
involvement, with some improvement in symp-
tomatology following antiretroviral therapy [62].
Clinical Manifestations
As with other systemic vasculitides, the disease is
characterized by constitutional symptoms of fevers
and weight loss, but also livedo reticularis, myal-
gias, weakness or leg tenderness, and peripheral
neuropathies. The most common organ of involve-
ment is the peripheral nervous system. CNS man-
ifestations have been reported in 20–45 % of
patients and manifest mainly as encephalopathy,
focal neurological deficits, and seizures [63].
Epidemiology
PAN generally presents in the fourth to sixth
decades of life without gender preference. Its inci-
dence is estimated at 2–3/100,000.
Imaging
CT and MRI have shown cortical and subcortical
infarcts. Angiography has shown segmental
nonspecific narrowing of middle-sized and small-
sized arteries, consistent with vasculitis, although
may be normal. Subarachnoid hemorrhage and intra-
cranial aneurysms are also reported although rare.
Kawasaki Disease
Kawasaki disease is an acute febrile mucocutane-
ous lymph node syndrome with classic findings of
769
fever, conjunctivitis, rash, lymphadenopathy, irri-
tability, oromucosal erythema, limb edema, and
skin peeling classically over the tips of the fingers
and toes. Coronary artery aneurysms are present
in 25 %. It generally presents in children less than
5 years of age. CNS involvement has been
described in 0.4–4 % of cases and includes facial
palsy, seizures, cerebral artery stenosis, and
moyamoya syndrome. In a small recent study,
seven out of ten patients with Kawasaki disease
showed CSF pleocytosis [64].
Churg-Strauss Syndrome
Like MAP and Wegener’s, Churg-Strauss is an
ANCA-associated small to medium vessel necro-
tizing vasculitis, often preceded by asthma and
eosinophilia manifesting similar to Wegener’s
with granuloma formation. Unlike Wegener’s,
however, ANCA positivity is present in only
about 40 % compared to 90 % of cases and is
associated more commonly with MPO-ANCA
(pANCA). It has also been referred to as allergic
granulomatosis or allergic granulomatosis with
polyangiitis.
Epidemiology
The prevalence rates of Churg-Strauss range
between 2 and 22/million, suggesting the disease
is about two to eight times less common than
MPA or Wegener’s [65].
Genetics
Positive associations are found with the human
leukocyte antigen HLA-DRB1*07 genotype.
A strong association with IL10 promoter polymor-
phisms has been detected in Churg-Strauss [66].
Clinical
Asthma is present in about 90 % of patients with
Churg-Strauss syndrome [67]. Chest findings are
characterized by migratory infiltrates and symp-
tomatic cardiac involvement which occurs in up to
47 % of patients. Eosinophilic myocardial infil-
tration may progress to fibrosis and left ventricular
dysfunction resulting in myocardial infarction and
arrhythmia as the main clinical features.
770
Peripheral neuropathy is seen in about half of
patients [67]. CNS involvement is seen in
10–30 % of patients and manifests as seizures,
stroke, and encephalopathy.
Pathophysiology
Multiple cell types contribute to the pathogenesis
of Churg-Strauss. These include predominantly
CD4+ Th2 (helper T cells) and eosinophils.
Circulating levels of eosinophils parallel disease
activity. Recent research has also shown the pres-
ence of Th1 and Th17 T-cell responses and indi-
rect evidence for the role of B cells [68].
Pathology
The small-vessel necrotizing vasculitis with possi-
ble granuloma formation is nonspecific and often
indistinguishable from other forms of ANCA-
associated vasculitides. Eosinophilic extravascular
tissue infiltration of virtually any organ character-
izes the early phase of disease. Nerve biopsy may
demonstrate epineural necrotizing vasculitis, eosin-
ophilic infiltrates, and lymphocytes.
MRI Findings
Granulomatous involvement of the parenchyma
and meninges can be seen and is usually T2
hypointense and enhancing and indistinguishable
from other granulomatous processes such as
Wegener’s, sarcoid, and tuberculosis. Patients
may present with infarct and hemorrhage. There
are a few case reports of aneurismal subarachnoid
hemorrhage attributed to vasculitic changes of the
disease.
Wegener’s
Wegener’s disease is a c-ANCA-associated nec-
rotizing small-vessel vasculitis with predomi-
nantly airway, lung, and kidney involvement.
The disease is currently referred to as necrotizing
granulomatosis with polyangiitis or necrotizing
granulomatosis.
Epidemiology
Wegener’s granulomatosis is rare with reported
prevalence rates of 23.7–156.5/million, with
M.H. Machnowska and R.I. Aviv
highest incidences reported in Northern Europe.
The usual age of diagnosis is in the fifth and sixth
decades, but the disease can occur in the pediatric
population or present as late as the eighth
decade [69]. Two forms exist: the proteinase
3-antineutrophil cytoplasmic autoantibody
(PR3-ANCA) (cANCA)-associated form pre-
dominates in Caucasians in the Northern Hemi-
sphere and the perinuclear myeloperoxidase
(MPO-ANCA)-associated form predominates in
Caucasian and non-Caucasians of the Southern
Hemisphere.
ANCA positivity is seen in over 90 % of
Wegener’s patients compared to 40 % of Churg-
Strauss patients [66].
Clinical
Pulmonary and kidney involvement is most char-
acteristic of Wegener’s. Pulmonary symptoms
include cough, dyspnea, thoracic pain, and
hemoptysis and reflect the involvement of the
entire respiratory tract. Kidney involvement usu-
ally manifests as microscopic hematuria, elevated
serum creatinine, and proteinuria ultimately
resulting in renal failure. Peripheral nervous sys-
tem involvement and cranial neuropathies are
more common than central nervous system
involvement and are seen in 20–50 % [70].
Sinonasal and orbital involvement is seen in
about 50 % of patients. CNS involvement is seen
in about 8 % of patients [70]. CNS manifestations
are due to sino-orbital spread of disease, CNS
vasculitis, or granulomatous lesions. CNS symp-
toms result from pachymeningeal thickening, vas-
culitis, and least likely pituitary involvement
thought to result from contiguous spread. Symp-
toms include headaches, meningismus, seizures,
encephalopathy, and stroke [71].
Genetics
The key candidate genes are the MHC complex
II-associated HLA-DRB1 genes on chromosome
6, particularly the extended DPB1*0401/
RXRB03 haplotype. Another association has
been made with the PTPN22 polymorphism
620 W in ANCA positive Wegener’s patients,
which results in disruption of the PTPN22 protein
responsible for T-cell receptor signaling. This may
33 Miscellaneous Vascular Diseases 771

result in enhanced T-cell activation with subse-

quent greater humoral abnormalities [72].

Pathophysiology
The disease is thought to result from activation of
cytokine primed neutrophils by ANCA resulting in
the release of proteolytic enzymes, reactive oxygen
species, and proinflammatory mediators [73]. The
disease is characterized by vasculitis and granu-
loma formation suggesting the presence of T-cell
hyperactivity. The causative agents resulting in
granuloma formation are unknown [73].
Several etiologic environmental factors have
been implicated in the development of WG and
include crystalline silica, organic solvents, cad-
mium, volatile hydrocarbons, and pesticides [73].

Imaging Findings
CNS imaging findings include pachymeningeal
thickening, which is nonspecific and cannot be
differentiated from granulomatous infections,
neoplastic involvement by benign and malignant
entities, hypertrophic pachymeningitis, and
noninfectious granulomatous diseases such as sar-
coid (Fig. 11).
Pituitary involvement is rare, most often affect-
ing the posterior with absence of the neurohypo-
physeal “bright spot,” but can lead to pituitary gland
enlargement with homogenous or heterogenous
enhancement and infundibular thickening. The
most common clinical finding is diabetes insipidus.
Less common involvement of the anterior pituitary
may manifest as hyperprolactinemia or panhypopi-
tuitarism [74]. Vasculitis is not well characterized
by angiography due to small-vessel involvement.
Orbital involvement, better characterized with MRI
than CT, can be seen in 50 % of patients and present
as both intra- and extraconal masses.
Sinonasal involvement is characterized by
bone involvement which may manifest as bone
destruction, usually of the septum and sinus walls,
sclerosing osteitis, bony thickening, and nodular
mucosal thickening [75]. Septal destruction may
ultimately result in a “saddle-nose” deformity
[70]. Sinonasal findings have to be differentiated
from other diseases. The bony destruction can be
mimicked by T-cell lymphoma, toxic etiologies
(cocaine), sinonasal squamous cell carcinoma,
Fig. 11 Pachymeningeal thickening along the right ante-
rior convexity in a 10-year-old patient with Wegener’s
granulomatosis (Courtesy of M. Shroff)
sinonasal undifferentiated carcinoma (SNUC),
melanoma, esthesioneuroblastoma, and NUT
midline carcinoma. Osteitis can be mimicked by
ossifying fibroma, osteoma, Paget disease, fibrous
dysplasia, and sinonasal sarcoid [75] (Fig. 12).
Systemic Lupus Erythematosus
Systemic lupus erythematosus (SLE) is a heterog-
enous autoimmune disease with widespread
systemic involvement and well-recognized neuro-
psychiatric syndromes.
Epidemiology
Recent large cohort studies of the disease suggest
the prevalence of neuropsychiatric lupus
(NPSLE) including both major and minor criteria
to be somewhere between 30 % and 40 % [76, 77].
NPSLE prevalence in children is similar to that in
adults. CNS disease accounts for about 5–7 %
mortality in SLE patients.
772
Fig. 12 Sinonasal
manifestations of
Wegener’s including
erosion of the nasal bone (a)
necessitating nasal bone
prosthesis (b)
Clinical Presentation
Recognized neuropsychiatric syndromes include
major manifestations such as seizure, cerebrovas-
cular disease, acute confusional state, psychosis,
and myelopathy. Minor CNS complaints also rec-
ognized as part of the neuropsychiatric syndrome
spectrum by the American College of Rheumatol-
ogy (ACR) include headache, anxiety, mild forms
of depression, and cognitive dysfunction. About
16 % of patients develop at least one of the
major criteria: seizures, cerebrovascular disease,
myelopathy, optic neuritis, aseptic meningitis,
and psychosis. About 50–60 % of the
19 ACR-recognized neuropsychiatric syndromes
are thought to be a result of metabolic distur-
bances, infections, or drug effects, including pos-
terior reversible encephalopathy syndrome, rather
than direct disease activity on the CNS [78].
Pathophysiology
The etiology of CNS disease in SLE is multifac-
torial and includes cardioembolism from left-
sided valvular disease, a prothrombotic state due
to antiphospholipid antibody (aPL), thrombotic
thrombocytopenic purpura, antibody-mediated
injury, and vasculitis. Thromboembolism is the
predominant cause of cerebrovascular events
[79]. Antiphospholipid is associated with ische-
mic stroke, venous thrombosis, and TIA.
M.H. Machnowska and R.I. Aviv
Pathology
Of the cases of lupus vasculitis, more than 60 %
demonstrate leukocytoclastic inflammation, 30 %
vasculitis with cryoglobulinemia, and about 6 %
systemic vasculitis resembling polyarteritis nodosa.
Imaging
The most common but also most nonspecific man-
ifestation of SLE is focal T2 hyperintensities in
the subcortical and periventricular white matter.
These usually do not contrast enhance and may
occasionally show diffusion restriction. The
hyperintensities are poorly differentiated from
microangiopathic disease seen with increasing
age or with increased vascular risk factors, such
as hypertension and diabetes mellitus, multiple
sclerosis, migraine, sarcoidosis, and HIV enceph-
alopathy. In 12 % of SLE patients, MRI has shown
cortical T2 hyperintensities, associated with dif-
fusion restriction, similar to those seen with sei-
zures [80]. Large-vessel infarction is usually
embolic in origin and associated with bland endo-
carditis. Bilateral middle cerebellar peduncle
infarcts have been reported as a presenting feature
of SLE [81].
Forty to 50 % of SLE patients with neuropsy-
chiatric symptom show normal MRI studies
[80]. Several metabolic and functional
imaging techniques have been used to identify
33 Miscellaneous Vascular Diseases
brain changes undetectable by conventional
imaging. These include magnetization transfer
imaging (MTI), proton magnetic resonance
spectroscopy (MRS), and PET. MTI shows
reduced magnetization transfer ratios in
patients with NPSLE indicative of subclinical
damage, likely due to axonal damage [82]. MRS
shows lower N-acetyl-aspartate (NAA) to creatine
(Cr) ratios reflecting axonal damage and reduced
neuronal-axonal density. Increased choline (Cho)
to creatine (Cr) ratios reflect inflammation and are
associated with gliosis, vasculopathy, and edema.
Ratios are elevated in the active phase of disease
and chronically increased in SLE patients with
moderate to severe cognitive impairment
[83]. MRS changes have been shown to herald
the future appearance of white matter lesions in
normal appearing brain matter [84].
FDG-PET studies show hypometabolism in the
parieto-occipital white matter in 60–80 % of
patients [85].
Special Procedures
Lupus vasculitis occurs in about 10 % of patients
and in most cases affects only small-caliber ves-
sels [86]. Nonspecific small-vessel occlusion may
be seen. Large-vessel occlusion can be seen in
active disease and has been associated with
moyamoya disease. Patients with systemic
middle-sized vessel involvement have the worst
outcomes. Large-vessel occlusion carries a 13 %
recurrent stroke risk. Focal ectasias and intracra-
nial aneurysm formation have also been reported.
Infectious Vasculitides
Varicella Zoster Vasculitis
Varicella zoster virus (VZV) is a human herpes
virus manifesting as a primary infection
(chickenpox) or secondary reactivation (shingles).
Primary infection usually occurs in children and is
responsible for about 31 % of strokes in the young
[87], while reactivation is more common in immu-
nocompromised or older individuals, due to a
773
decrease of cell-mediated immunity to VZV.
Both primary and secondary CNS vasculitis may
be preceded by a rash in about one half to two
thirds of patients, although the onset of CNS find-
ings may occur several months following cutane-
ous findings [88].
While a pure VZV-related encephalitis exists,
it is far less common than VZV-associated vascu-
litis. The vasculitis most commonly involves large
and small arteries leading to large-vessel and per-
forator territory infarcts and well-demarcated
ischemic foci at the gray-white matter junctions.
In a recent study, a combination of large- and
small-vessel involvement was seen in 50 % of
patients, followed by small arterial involvement
(37 %) and large-vessel involvement (13 %)
[88]. Based on feline studies, the disease spread
is thought to be through ipsilateral trigeminal gan-
glion afferents to the anterior circulation arteries,
i.e., the carotid terminus and MCA, and from the
cervical dorsal root ganglia to the vertebrobasilar
system [89]. PCR techniques have detected
viral DNA in the media of cerebral arteries,
while histopathologic evaluation has shown
multinucleated giant cells and herpes virions in
smooth muscle cells of the MCA. The small-
vessel vasculopathy often yields lesions with
intranuclear Cowdry A inclusions in glia
(a hallmark of herpes infection), central cavita-
tion, and macrophage influx [90].
MRI appearances show gray matter
hyperintensity, swelling, and cortical and subcor-
tical white matter ischemic foci or hemorrhages.
The predominant angiographic abnormalities
reveal segmental stenosis or occlusion of proxi-
mal vessels. Vessel wall imaging may show arte-
rial wall enhancement in the acute phase. There
are rare reports of aneurysm formation, subarach-
noid hemorrhage, vascular ectasia, and dissection
[90] (Fig. 13).
HIV Vasculitis
HIV causes predominantly a large-vessel vasculi-
tis, although cases of penetrating, small, and
medium arterial involvement are reported. Vessels
show disruption of the internal elastic lamina and
774
Fig. 13 Varicella zoster vasculitis in a 6-year-old child
resulting in left lenticulostriate territory infarcts shown on
DWI (a, b). Postcontrast T1-WI vessel wall imaging shows
subsequent media layer disruption [91]. Vessel
narrowing, occlusion, fusiform dilation, and aneu-
rysm formation have been described. The diagno-
sis should only be made after exclusion of
concomitant VZV, CMV, and syphilitic- or
toxoplasma-associated vasculitis.
Bacterial Vasculitis
Bacterial infection can cause vascular injury
through several mechanisms. Direct extension
from adjacent mastoiditis or sinusitis with or
without associated osteomyelitis can result in
vessel irregularity, narrowing, occlusion, or fusi-
form aneurysm formation. Basal vessels, includ-
ing the carotid arteries, are most often affected
with less common involvement of peripheral
vessels.
Meningitis, cerebritis, abscess, empyema, and
septic emboli can also result in vessel irregularity.
A retrospective study of 87 pneumococcal men-
ingitis patients found arterial and venous infarc-
tion in 22 % and 10 % of cases, respectively [92].
Proposed mechanisms for large and medium
arterial narrowing in meningitis include encroach-
ment of large vessels by subarachnoid inflamma-
tory exudate; invasion of the vessel wall by
inflammatory cells with resulting intimal thicken-
ing, focal stenosis, dilatation, and edema; and
bland vasospasm [93].
M.H. Machnowska and R.I. Aviv
wall enhancement in the left carotid terminus (c) (Courtesy
of M. Shroff, G. DeVeber)
Tuberculous Vasculitis
Tuberculous meningitis classically manifests with
exudative basilar infiltrates, resulting in an infil-
trative vasculitis, which proceeds from the adven-
titia of vessels to their inner layers, and affecting
predominantly the terminal carotids; proximal
MCA, ACA, and PCA; and basilar arteries. Pro-
liferative vasculitis, characterized by intimal pro-
liferation, and necrotizing vasculitis have also
been described. Vascular arterial narrowing has
been reported in the inferior aspect of vessels
and may be limited to regions surrounding
tuberculomas (Fig. 14).
Fungal Vasculitis
Fungal infection can be directly spread through
concomitant sinonasal infection, as in the case of
zygomycoses, or hematogeneously spread, partic-
ularly in the immunocompromised host, as in the
case of aspergillus infection in Western countries.
Vasculitis can occur in two patterns: the tubercu-
lous type, indistinguishable from tuberculosis, or
the hyphal type. The latter results from thrombosis
and inflammation caused by angioinvasive hyphal
elements and results in hemorrhagic infarctions,
subarachnoid hemorrhage, abscesses, or mycotic
aneurysms. A high index of suspicion for fungal
infection is required, particularly in the case of
33 Miscellaneous Vascular Diseases
Fig. 14 A 15-year-old male with TB meningitis. (a, b)
DWI images shows multiple basal ganglia and posterior
fossa infarcts. (c) Abnormal T2 signal is noted at the
margins of the interpeduncular cistern and in the cerebellar
patients with hematologic malignancies and bone
marrow transplant recipients who present with
hemorrhagic parenchymal CNS lesions [93].
Neuro-Behcet’s Disease
Behcet’s disease is a multisystem vasculitis which
classically presents a triad of oral and genital
ulcerations with uveitis. CNS involvement is
seen in 5–10 % of patients.
Epidemiology
Behcet’s disease follows the “Silk Road,” an
ancient trading route dating back to the second
century BC, spanning from the Mediterranean to
Japan. Its prevalence in the Mediterranean is
775
vermis suggestive of a basal meningitis. (d) MRA shows
narrowing of the right A1 segment (Courtesy of M Shroff,
G. DeVeber)
reported between 20 and 421/100,000 [94]. The
age of onset is usually in the third decade, with no
gender predilection for the systemic disease, and a
slight (2.6:1) male predominance seen in CNS
disease [95].
Pediatric involvement is rare, but can occur,
and a more severe course of disease is seen in men
younger than 25.
Etiology
The etiology of Behcet’s disease is
currently unknown. An association has been
made with a higher prevalence of HLA B5
(51) in Behcet’s patients, particularly in highly
endemic areas, but viral, bacterial, genetic, and
other autoimmune etiologies have also been
implicated [95].
776
Fig. 15 A 6-year-old male with Behcet’s disease. (a) CT
shows left middle cerebellar peduncle hypodensity and
associated mass effect. (b, c) FLAIR and contrast-
Clinical
CNS involvement is usually preceded by the clas-
sic triad of uveitis and genital and oral ulceration.
Neuro-Behcet’s syndrome (NBS) can present as
focal or multifocal CNS dysfunction resembling
multiple sclerosis. A subset of NBS patients dis-
plays symptoms of intracranial hypertension
attributable to dural venous sinus thrombosis.
Peripheral nervous system involvement is rare,
and greater than half of the patients are
moderately to severely disabled 10 years
following symptom onset.
Pathology
Large-vessel arterial and venous involvement is
seen in up to 35 % of cases, with venous involve-
ment seen about eight times more commonly than
arterial involvement in the CNS [96]. Pathologi-
cally, a mononuclear vascular and perivascular
infiltrate is noted. There is disruption of the inter-
nal elastic lamina and absence of intimal and
medial layers. Fibrinoid necrosis has been
observed. Venous occlusion, thrombophlebitis,
and thrombosis are the most common findings;
however, arterial occlusion and cerebral aneurysm
formation have been described.
Imaging
Parenchymal NBS occurs in about 75 % of cases
and carries a worse prognosis than venous throm-
bosis, the other NBS manifestation. The two
M.H. Machnowska and R.I. Aviv
enhanced T1-WI images show an ill-defined T2
hyperintense, ring enhancing lesion with peripheral
edema, and mild mass effect (Courtesy of M. Shroff)
manifestations seldom coexist. Parenchymal
lesions are characterized by T1 hypointensity
and T2 hyperintensity on MRI. Contrast enhance-
ment and greater cranial and caudal extent of T2
signal abnormality are observed in the acute and
subacute phases. The mesencephalic-thalamic
junction is the most common site of CNS involve-
ment. Other regions of involvement include the
pontobulbar region; the hypothalamus-thalamus,
basal ganglia; and least commonly the
periventricular, deep white matter and subcortical
white matter. Pontobulbar lesions tend to involve
the tegmentum and superior cerebellar peduncles
and always spare the red nuclei. Cortical lesions
are seldom seen [95] (Fig. 15).
Large-vessel involvement includes stenosis
and occlusion and rare aneurismal formation.
Venous sinus thrombosis is seen in 30 % of
patients. In addition to the dural venous sinuses,
the superior vena cava, subclavian veins, and jug-
ular veins can be involved.
Summary
The diseases affecting intracranial vessels
discussed in this chapter often result in transient
or permanent deficits not infrequently affecting
younger patients. When faced with work up for
stroke in the young, imaging can prove invaluable
in aiding diagnosis and treatment. Specific
33 Miscellaneous Vascular Diseases
findings such as extensive temporal white matter
involvement in CADASIL, the angiographic
“puff of smoke” in moyamoya, or pulvinar
hyperintensity in Fabry’s disease can lead to
rapid diagnosis of disease. Emerging new
methods, such as vessel wall imaging, can help
detect the presence of inflammation either to
guide diagnosis, biopsy, or treatment. Functional
imaging methods are becoming significantly more
important in determining the optimal time for
treatment, particularly in the case of moyamoya,
allowing for earlier detection of tissue at risk than
standard anatomical imaging methods.
References
1. Rolfs A, Fazekas F, Grittner U et al (2013) Acute
cerebrovascular disease in the young: the stroke in
young fabry patients study. Stroke 44(2):340–349
2. Brouns R, Thijs V, Eyskens F et al (2010) Belgian fabry
study: prevalence of fabry disease in a cohort of 1000
young patients with cerebrovascular disease. Stroke
41(5):863–868
3. DeGraba T, Azhar S, Dignat-George F et al (2000)
Profile of endothelial and leukocyte activation in
fabry patients. Ann Neurol 47(2):229–233
4. Moore DF, Kaneski CR, Askari H, Schiffmann R
(2007) The cerebral vasculopathy of fabry disease.
J Neurol Sci 257(1–2):258–263
5. Hilz MJ, Kolodny EH, Brys M, Stemper B, Haendl T,
Marthol H (2004) Reduced cerebral blood flow
velocity and impaired cerebral autoregulation
in patients with fabry disease. J Neurol 251(5):
564–570
6. Moore DF, Ye F, Schiffmann R, Butman JA (2003)
Increased signal intensity in the pulvinar on
T1-weighted images: a pathognomonic MR imaging
sign of fabry disease. AJNR Am J Neuroradiol
24(6):1096–1101
7. Fellgiebel A, Mazanek M, Whybra C et al (2006) Pat-
tern of microstructural brain tissue alterations in fabry
disease: a diffusion-tensor imaging study. J Neurol
253(6):780–787
8. Fellgiebel A, Keller I, Marin D et al (2009) Diagnostic
utility of different MRI and MR angiography measures
in fabry disease. Neurology 72(1):63–68
9. Bartynski WS (2008) Posterior reversible encephalop-
athy syndrome, part 2: controversies surrounding path-
ophysiology of vasogenic edema. AJNR Am J
Neuroradiol 29(6):1043–1049
10. Bartynski WS (2008) Posterior reversible encephalop-
athy syndrome, part 1: fundamental imaging and clin-
ical features. AJNR Am J Neuroradiol 29(6):
1036–1042
777
11. Wilson SE, de Groen PC, Aksamit AJ, Wiesner RH,
Garrity JA, Krom RA (1988) Cyclosporin A-induced
reversible cortical blindness. J Clin Neuroophthalmol
8(4):215–220
12. Kassem-Moussa H, Provenzale JM, Petrella JR, Lewis
DV (2000) Early diffusion-weighted MR imaging
abnormalities in sustained seizure activity. AJR Am J
Roentgenol 174(5):1304–1306
13. Mukherjee P, McKinstry RC (2001) Reversible poste-
rior leukoencephalopathy syndrome: evaluation with
diffusion-tensor MR imaging. Radiology 219(3):
756–765
14. Bartynski WS, Boardman JF (2008) Catheter angiog-
raphy, MR angiography, and MR perfusion in posterior
reversible encephalopathy syndrome. AJNR Am J
Neuroradiol 29(3):447–455
15. Wartenberg KE, Parra A (2006) CT and CT-perfusion
findings of reversible leukoencephalopathy during
triple-H therapy for symptomatic subarachnoid
hemorrhage-related vasospasm. J Neuroimaging
16(2):170–175
16. Naidu K, Moodley J, Corr P, Hoffmann M (1997)
Single photon emission and cerebral computerised
tomographic scan and transcranial doppler sono-
graphic findings in eclampsia. Br J Obstet Gynaecol
104(10):1165–1172
17. Calabrese LH, Dodick DW, Schwedt TJ, Singhal AB
(2007) Narrative review: reversible cerebral vasocon-
striction syndromes. Ann Intern Med 146(1):34–44
18. Ducros A (2014) Reversible cerebral vasoconstriction
syndrome. Handb Clin Neurol 121:1725–1741
19. Ducros A, Boukobza M, Porcher R, Sarov M,
Valade D, Bousser MG (2007) The clinical and radio-
logical spectrum of reversible cerebral vasoconstriction
syndrome. A prospective series of 67 patients. Brain
130(Pt 12):3091–3101
20. Ducros A, Fiedler U, Porcher R, Boukobza M, Stapf C,
Bousser MG (2010) Hemorrhagic manifestations of
reversible cerebral vasoconstriction syndrome: fre-
quency, features, and risk factors. Stroke 41(11):
2505–2511
21. Mandell DM, Matouk CC, Farb RI et al (2012) Vessel
wall MRI to differentiate between reversible
cerebral vasoconstriction syndrome and central ner-
vous system vasculitis: preliminary results. Stroke
43(3):860–862
22. Baba T, Houkin K, Kuroda S (2008) Novel epidemio-
logical features of moyamoya disease. J Neurol
Neurosurg Psychiatry 79(8):900–904
23. Smith ER, Scott RM (2012) Spontaneous occlusion of
the circle of willis in children: pediatric moyamoya
summary with proposed evidence-based practice
guidelines. A review. J Neurosurg Pediatr 9(4):
353–360
24. Nagaraja D, Verma A, Taly AB, Kumar MV,
Jayakumar PN (1994) Cerebrovascular disease in chil-
dren. Acta Neurol Scand 90(4):251–255
25. Mineharu Y, Takenaka K, Yamakawa H et al (2006)
Inheritance pattern of familial moyamoya disease:
778
autosomal dominant mode and genomic imprinting.
J Neurol Neurosurg Psychiatry 77(9):1025–1029
26. Smith ER, Scott RM (2008) Progression of disease in
unilateral moyamoya syndrome. Neurosurg Focus
24(2):E17
27. Burke GM, Burke AM, Sherma AK, Hurley MC,
Batjer HH, Bendok BR (2009) Moyamoya disease:
a summary. Neurosurg Focus 26(4):E11
28. Lee M, Zaharchuk G, Guzman R, Achrol A, Bell-
Stephens T, Steinberg GK (2009) Quantitative hemo-
dynamic studies in moyamoya disease: a review.
Neurosurg Focus 26(4):E5
29. Yamashita M, Oka K, Tanaka K (1983) Histopathology
of the brain vascular network in moyamoya disease.
Stroke 14(1):50–58
30. Bacigaluppi S, Dehdashti AR, Agid R, Krings T,
Tymianski M, Mikulis DJ (2009) The contribution of
imaging in diagnosis, preoperative assessment, and
follow-up of moyamoya disease: a review. Neurosurg
Focus 26(4):E3
31. Suzuki J, Takaku A (1969) Cerebrovascular
“moyamoya” disease. Disease showing abnormal
net-like vessels in base of brain. Arch Neurol 20(3):
288–299
32. Kikuta K, Takagi Y, Nozaki K et al (2005) Asymptom-
atic microbleeds in moyamoya disease: T2*-weighted
gradient-echo magnetic resonance imaging study.
J Neurosurg 102(3):470–475
33. Research Committee on the Pathology and Treatment
of Spontaneous Occlusion of the Circle of Willis,
Health Labour Sciences Research Grant for
Research on Measures for Infractable Diseases (2012)
Guidelines for diagnosis and treatment of
moyamoya disease (spontaneous occlusion of the cir-
cle of Willis). Neurol Med Chir (Tokyo) 52(5):
245–266
34. Nariai T, Matsushima Y, Imae S et al (2005) Severe
haemodynamic stress in selected subtypes of patients
with moyamoya disease: a positron emission tomogra-
phy study. J Neurol Neurosurg Psychiatry 76(5):
663–669
35. Sakamoto S, Ohba S, Shibukawa M, Kiura Y, Arita K,
Kurisu K (2006) CT perfusion imaging for childhood
moyamoya disease before and after surgical revascu-
larization. Acta Neurochir (Wien) 148(1):77–81,
discussion 81
36. Aviv R, Letourneau-Guillon L, Symons S, Fox A
(2013) Other arteriopathies. In: Naidich T, Castillo M,
Cha S, Smirniotopoulos J (eds) Imaging of the brain.
Saunders Elsevier, Philadelphia, pp 451–469
37. Kang KH, Kim HS, Kim SY (2008) Quantitative cere-
brovascular reserve measured by acetazolamide-
challenged dynamic CT perfusion in ischemic adult
moyamoya disease: initial experience with angio-
graphic correlation. AJNR Am J Neuroradiol 29(8):
1487–1493
38. Mikulis DJ, Krolczyk G, Desal H et al (2005) Preop-
erative and postoperative mapping of cerebrovascular
reactivity in moyamoya disease by using blood oxygen
M.H. Machnowska and R.I. Aviv
level-dependent magnetic resonance imaging.
J Neurosurg 103(2):347–355
39. Shimizu H, Shirane R, Fujiwara S, Takahashi A,
Yoshimoto T (1997) Proton magnetic resonance spec-
troscopy in children with moyamoya disease. Clin
Neurol Neurosurg 99(Suppl 2):S64–S67
40. Opherk C, Peters N, Herzog J, Luedtke R, Dichgans M
(2004) Long-term prognosis and causes of death in
CADASIL: a retrospective study in 411 patients.
Brain 127(Pt 11):2533–2539
41. Desmond DW, Moroney JT, Lynch T, Chan S, Chin SS,
Mohr JP (1999) The natural history of CADASIL: a
pooled analysis of previously published cases. Stroke
30(6):1230–1233
42. Okeda R, Arima K, Kawai M (2002) Arterial changes
in cerebral autosomal dominant arteriopathy with sub-
cortical infarcts and leukoencephalopathy (CADASIL)
in relation to pathogenesis of diffuse myelin loss of
cerebral white matter: examination of cerebral medul-
lary arteries by reconstruction of serial sections of an
autopsy case. Stroke 33(11):2565–2569
43. Viswanathan A, Gray F, Bousser MG, Baudrimont M,
Chabriat H (2006) Cortical neuronal apoptosis in
CADASIL. Stroke 37(11):2690–2695
44. van den Boom R, Lesnik Oberstein SA, Ferrari MD,
Haan J, van Buchem MA (2003) Cerebral autosomal
dominant arteriopathy with subcortical infarcts and
leukoencephalopathy: MR imaging findings at differ-
ent ages–3rd-6th decades. Radiology 229(3):683–690
45. Dichgans M, Holtmannspotter M, Herzog J, Peters N,
Bergmann M, Yousry TA (2002) Cerebral microbleeds
in CADASIL: a gradient-echo magnetic resonance
imaging and autopsy study. Stroke 33(1):67–71
46. Cumurciuc R, Guichard JP, Reizine D, Gray F, Bousser
MG, Chabriat H (2006) Dilation of virchow-robin
spaces in CADASIL. Eur J Neurol 13(2):187–190
47. Fukutake T (2011) Carasil. Brain Nerve 63(2):99–108
48. Salvarani C, Brown RD Jr, Calamia KT et al (2007)
Primary central nervous system vasculitis: analysis of
101 patients. Ann Neurol 62(5):442–451
49. Giannini C, Salvarani C, Hunder G, Brown RD (2012)
Primary central nervous system vasculitis: pathology
and mechanisms. Acta Neuropathol 123(6):759–772
50. Ringleb PA, Strittmatter EI, Loewer M et al (2005)
Cerebrovascular manifestations of Takayasu arteritis
in Europe. Rheumatology (Oxford) 44(8):1012–1015
51. Desai MY, Stone JH, Foo TK, Hellmann DB, Lima JA,
Bluemke DA (2005) Delayed contrast-enhanced MRI
of the aortic wall in Takayasu’s arteritis: initial experi-
ence. AJR Am J Roentgenol 184(5):1427–1431
52. Gulati A, Bagga A (2010) Large vessel vasculitis.
Pediatr Nephrol 25(6):1037–1048
53. Walter MA, Melzer RA, Schindler C, Muller-Brand J,
Tyndall A, Nitzsche EU (2005) The value of [18F]
FDG-PET in the diagnosis of large-vessel vasculitis
and the assessment of activity and extent of disease.
Eur J Nucl Med Mol Imaging 32(6):674–681
54. Gonzalez-Gay MA, Vazquez-Rodriguez TR, Lopez-
Diaz MJ et al (2009) Epidemiology of giant cell
33 Miscellaneous Vascular Diseases
arteritis and polymyalgia rheumatica. Arthritis Rheum
61(10):1454–1461
55. Gonzalez-Gay MA, Garcia-Porrua C, Gonzalez-
Juanatey C, Miranda-Filloy JA, Blanco R, Llorca J
(2013) Biopsy-proven giant cell arteritis patients with
coronary artery disease have increased risk of aortic
aneurysmal disease and arterial thrombosis. Clin Exp
Rheumatol 31(1 Suppl 75):S94
56. Ly KH, Regent A, Tamby MC, Mouthon L (2010)
Pathogenesis of giant cell arteritis: more than just an
inflammatory condition? Autoimmun Rev 9(10):
635–645
57. Gonzalez-Gay MA, Garcia-Porrua C, Llorca J
et al (2000) Visual manifestations of giant cell arteritis.
Trends and clinical spectrum in 161 patients. Medicine
(Baltimore) 79(5):283–292
58. Schmidt WA (2013) Imaging in vasculitis. Best Pract
Res Clin Rheumatol 27(1):107–118
59. Schmidt WA, Kraft HE, Vorpahl K, Volker L,
Gromnica-Ihle EJ (1997) Color duplex ultrasonogra-
phy in the diagnosis of temporal arteritis. N Engl J Med
337(19):1336–1342
60. Prieto-Gonzalez S, Arguis P, Garcia-Martinez A
et al (2012) Large vessel involvement in biopsy-proven
giant cell arteritis: prospective study in 40 newly diag-
nosed patients using CT angiography. Ann Rheum Dis
71(7):1170–1176
61. Bley TA, Uhl M, Carew J et al (2007) Diagnostic value
of high-resolution MR imaging in giant cell arteritis.
AJNR Am J Neuroradiol 28(9):1722–1727
62. Segelmark M, Selga D (2007) The challenge of man-
aging patients with polyarteritis nodosa. Curr Opin
Rheumatol 19(1):33–38
63. Provenzale JM, Allen NB (1996) Neuroradiologic
findings in polyarteritis nodosa. AJNR Am J
Neuroradiol 17(6):1119–1126
64. Korematsu S, Uchiyama S, Miyahara H et al (2007)
The characterization of cerebrospinal fluid and serum
cytokines in patients with Kawasaki disease. Pediatr
Infect Dis J 26(8):750–753
65. Gibelin A, Maldini C, Mahr A (2011) Epidemiology
and etiology of Wegener granulomatosis, micro-
scopic polyangiitis, churg-strauss syndrome and
goodpasture syndrome: vasculitides with frequent
lung involvement. Semin Respir Crit Care Med 32(3):
264–273
66. Wieczorek S, Holle JU, Epplen JT (2010) Recent pro-
gress in the genetics of wegener’s granulomatosis and
churg-strauss syndrome. Curr Opin Rheumatol 22(1):
8–14
67. Comarmond C, Pagnoux C, Khellaf M et al (2013)
Eosinophilic granulomatosis with polyangiitis (churg-
strauss): clinical characteristics and long-term
followup of the 383 patients enrolled in the french
vasculitis study group cohort. Arthritis Rheum
65(1):270–281
68. Mahr A, Moosig F, Neumann T et al (2014) Eosino-
philic granulomatosis with polyangiitis (churg-
strauss): evolutions in classification, etiopathogenesis,
779
assessment and management. Curr Opin Rheumatol
26(1):16–23
69. Almouhawis HA, Leao JC, Fedele S, Porter SR (2013)
Wegener’s granulomatosis: a review of clinical features
and an update in diagnosis and treatment. J Oral Pathol
Med 42(7):507–516
70. Mujagic S, Sarihodzic S, Huseinagic H, Karasalihovic
Z (2011) Wegener’s granulomatosis of the paranasal
sinuses with orbital and central nervous system
involvement-diagnostic imaging. Acta Neurol Belg
111(3):241–244
71. Seror R, Mahr A, Ramanoelina J, Pagnoux C, Cohen P,
Guillevin L (2006) Central nervous system involve-
ment in Wegener granulomatosis. Medicine (Balti-
more) 85(1):54–65
72. Jagiello P, Aries P, Arning L et al (2005) The PTPN22
620W allele is a risk factor for Wegener’s
granulomatosis. Arthritis Rheum 52(12):4039–4043
73. Manna R, Cadoni G, Ferri E et al (2008) Wegener’s
granulomatosis: an update on diagnosis and therapy.
Expert Rev Clin Immunol 4(4):481–495
74. Yong TY, Li JY, Amato L et al (2008) Pituitary involve-
ment in Wegener’s granulomatosis. Pituitary 11(1):
77–84
75. Lohrmann C, Uhl M, Warnatz K, Kotter E, Ghanem N,
Langer M (2006) Sinonasal computed tomography in
patients with Wegener’s granulomatosis. J Comput
Assist Tomogr 30(1):122–125
76. Borowoy AM, Pope JE, Silverman E et al (2012) Neu-
ropsychiatric lupus: the prevalence and autoantibody
associations depend on the definition: results from the
1000 faces of lupus cohort. Semin Arthritis Rheum
42(2):179–185
77. Unterman A, Nolte JE, Boaz M, Abady M,
Shoenfeld Y, Zandman-Goddard G (2011) Neuropsy-
chiatric syndromes in systemic lupus erythematosus:
a meta-analysis. Semin Arthritis Rheum 41(1):1–11
78. Steup-Beekman GM, Zirkzee EJ, Cohen D et al (2013)
Neuropsychiatric manifestations in patients with sys-
temic lupus erythematosus: epidemiology and radiol-
ogy pointing to an immune-mediated cause. Ann
Rheum Dis 72(Suppl 2):ii76–ii79
79. Hanly JG, Walsh NM, Sangalang V (1992) Brain
pathology in systemic lupus erythematosus.
J Rheumatol 19(5):732–741
80. Luyendijk J, Steens SC, Ouwendijk WJ et al (2011)
Neuropsychiatric systemic lupus erythematosus: les-
sons learned from magnetic resonance imaging. Arthri-
tis Rheum 63(3):722–732
81. Okamoto K, Tokiguchi S, Furusawa T et al (2003) MR
features of diseases involving bilateral middle cerebel-
lar peduncles. AJNR Am J Neuroradiol 24(10):
1946–1954
82. Bosma GP, Steens SC, Petropoulos H et al (2004)
Multisequence magnetic resonance imaging study of
neuropsychiatric systemic lupus erythematosus.
Arthritis Rheum 50(10):3195–3202
83. Axford JS, Howe FA, Heron C, Griffiths JR (2001)
Sensitivity of quantitative (1)H magnetic resonance
780
spectroscopy of the brain in detecting early neuronal
damage in systemic lupus erythematosus. Ann Rheum
Dis 60(2):106–111
84. Appenzeller S, Li LM, Costallat LT, Cendes F (2007)
Neurometabolic changes in normal white matter may
predict appearance of hyperintense lesions in systemic
lupus erythematosus. Lupus 16(12):963–971
85. Popescu A, Kao AH (2011) Neuropsychiatric systemic
lupus erythematosus. Curr Neuropharmacol 9(3):
449–457
86. D’Cruz D (1998) Vasculitis in systemic lupus
erythematosus. Lupus 7(4):270–274
87. Askalan R, Laughlin S, Mayank S et al (2001)
Chickenpox and stroke in childhood: a study of fre-
quency and causation. Stroke 32(6):1257–1262
88. Nagel MA, Cohrs RJ, Mahalingam R et al (2008) The
varicella zoster virus vasculopathies: clinical, CSF,
imaging, and virologic features. Neurology 70(11):
853–860
89. Saito K, Moskowitz MA (1989) Contributions from the
upper cervical dorsal roots and trigeminal ganglia to
the feline circle of Willis. Stroke 20(4):524–526
M.H. Machnowska and R.I. Aviv
90. Gilden D, Cohrs RJ, Mahalingam R, Nagel MA (2009)
Varicella zoster virus vasculopathies: diverse clinical
manifestations, laboratory features, pathogenesis, and
treatment. Lancet Neurol 8(8):731–740
91. Gutierrez J, Ortiz G (2011) HIV/AIDS patients with
HIV vasculopathy and VZV vasculitis: a case series.
Clin Neuroradiol 21(3):145–151
92. Kastenbauer S, Pfister HW (2003) Pneumococcal men-
ingitis in adults: spectrum of complications and prog-
nostic factors in a series of 87 cases. Brain 126(Pt 5):
1015–1025
93. Chow FC, Marra CM, Cho TA (2011) Cerebrovascular
disease in central nervous system infections. Semin
Neurol 31(3):286–306
94. Mat C, Yurdakul S, Sevim A, Ozyazgan Y, Tuzun Y
(2013) Behcet’s syndrome: facts and controversies.
Clin Dermatol 31(4):352–361
95. Kocer N, Islak C, Siva A et al (1999) CNS involvement
in neuro-behcet syndrome: an MR study. AJNR Am J
Neuroradiol 20(6):1015–1024
96. Koc Y, Gullu I, Akpek G et al (1992) Vascular involve-
ment in Behcet’s disease. J Rheumatol 19(3):402–410
 Vascular Loop Syndromes
34
Divyata R. Hingwala and Kesavadas Chandrasekharan

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 781 Vascular loop syndromes are disorders usually
associated with vascular compression of the
Trigeminal Neuralgia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 782
Clinical Features, Classification, and Definition . . . . 782
specific cranial nerves. Neuroimaging may
Etiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 782 play an important role in the preoperative diag-
Pathogenesis and Pathophysiology . . . . . . . . . . . . . . . . . . 782 nosis of these syndromes. MRI can detect the
Necessity for Neuroimaging in TN . . . . . . . . . . . . . . . . . . 782 vascular loops compressing the cranial nerves
Neuroimaging Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . 782
Neuroimaging Findings in TN . . . . . . . . . . . . . . . . . . . . . . . 783
(fifth, seventh, ninth, eighth, and fourth) and
can also rule out secondary causes. The aim of
Hemifacial Spasm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 784 this chapter is to outline the clinical features
Clinical Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 784
Pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 785 and pathogenesis of neurovascular syndromes
Role of Neuroimaging in HFS . . . . . . . . . . . . . . . . . . . . . . . 785 and give an overview of neuroimaging tech-
Neuroimaging Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 785 niques and findings.
Other Rare Manifestations of Vascular
Loop Syndrome and Their Imaging Features . . . . 786 Keywords
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 787 Trigeminal neuralgia • Hemifacial spasm •
Glossopharyngeal neuralgia • Superior oblique
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 787
myokymia

Introduction

Vascular loop syndromes are disorders usually

associated with vascular compression of the spe-
cific cranial nerves. This compression can cause
paroxysmal attacks of pain or abnormal move-
D.R. Hingwala (*)
ments in the distribution of the affected nerves
NM Medical Centre, Mumbai, Maharashtra, India
e-mail: divyatahingwala@gmail.com [1]. Neuroimaging may play an important role in
the preoperative diagnosis of these syndromes.
K. Chandrasekharan
NM Medical Centre, Mumbai, Maharashtra, India The aim of this chapter is to outline the clinical
features and pathogenesis of neurovascular syn-
Sree Chitra Tirunal Institute for Medical Sciences and
Technology, Trivandrum, Kerela, India dromes and give an overview of neuroimaging
e-mail: chandkesav@yahoo.com techniques and findings.
# Springer Science+Business Media New York 2016 781
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_9
782
Trigeminal Neuralgia
Clinical Features, Classification,
and Definition
Trigeminal neuralgia (TN) has been defined by the
International Association for the Study of Pain as
sudden, usually unilateral, severe, brief, stabbing,
recurrent episodes of pain in the distribution of
one or more branches of the trigeminal nerve
[2]. Specific clinical criteria have been described
like characteristic, stereotyped patterns of pain
attacks, no objective neurological deficit, and no
other identified causes of pain [3].
According to the latest classification of the
International Headache Society, TN is classified
as classic and symptomatic. Classic TN (CTN) is
distinguished from symptomatic TN (STN) by the
absence of an established etiology and clinically
evident neurologic deficit [3].
This pain is localized to the trigeminal nerve
distribution. It often affects the maxillary (V2) or
mandibular (V3) divisions of the trigeminal nerve
or a combination of adjacent divisions. The topo-
graphic distribution of pain is determined by the
location of the vascular loop. Clinical stimuli
within specific trigger zones or specific activities
like chewing, shaving, or brushing teeth may pre-
cipitate pain attacks. Occasionally, in the period
following a pain attack, the patient may be refrac-
tory to further pain [4].
Neuroepidemiological studies have projected
the incidence of TN to be between 4 and 5 per
100,000 per year. It is more frequent with advanc-
ing age and in females as compared with
males [5].
Etiology
CTN is caused by neurovascular compression (see
section “Pathophysiology”). Potential etiologies
of STN include multiple sclerosis plaques, tumors
(schwannomas, meningiomas, epidermoids, etc.),
and abnormalities of the skull base [6]. Stretching
of the trigeminal nerve in Chiari I malformation is
also a rare cause of trigeminal neuralgia [7].
D.R. Hingwala and K. Chandrasekharan
Pathogenesis and Pathophysiology
Trigeminal neuralgia is caused by demyelination
of the trigeminal sensory fibers within the
trigeminal nerve root [8]. Walter Dandy was the
first to observe that the nerve is compressed by
an overlying blood vessel [9], and it has been
confirmed by various authors including
Hamlyn [10] that vascular compression at the
root entry zone (REZ) is the cause of TN in
80–90 % of the cases. The REZ of the trigeminal
nerve is the transition zone between the central
myelin (formed by oligodendrocytes) and periph-
eral myelin (formed by Schwann cells) and is
located approximately 1–2.5 mm from the
pons [11].
Pathological studies have demonstrated
demyelinated axons which are in direct apposition
with few intervening glial processes [8]. This
allows the generation of ectopic impulses and
non-synaptic ephaptic transmission of impulses
[12]. The pulsatility of overlying arteries may
contribute to the demyelination and abnormal
impulse generation [8]. Nerve fibers subserving
light touch and pain are the nearest in proximity at
the REZ. This theory further explains how light
touch provokes pain [4].
Necessity for Neuroimaging in TN
Studies have found that STN and CTN cannot be
reliably differentiated on the basis of clinical fea-
tures, trigeminal reflex, and trigeminal nerve-
evoked potentials [13, 14], though younger age
and abnormal trigeminal nerve-evoked potentials
may suggest the presence of a secondary cause
[15]. The role of MRI is not only to exclude
secondary causes but also to depict vascular
loops causing TN.
Neuroimaging Techniques
CT is inadequate for evaluating a patient with
vascular loop syndromes [16]. Conventional
MRI images may be useful for excluding
34 Vascular Loop Syndromes
nonvascular causes of HFS. However, these may
fail to satisfactorily delineate the neurovascular
relationships [17].
Useful techniques [18] include a heavily
T2-WI sequence like Constructive Interference
in Steady State (CISS, Siemens) or 3D Fast Imag-
ing with Steady State (FIESTA, GE) sequence in
which the CSF is bright and other cisternal struc-
tures like nerves, arteries, and veins appear dark.
MR angiography is also very useful to delineate
the culprit artery due to superb contrast between
the artery and CSF, but the nerves are not well
defined as compared with the CSF [19]. These
sequences have high spatial resolution due to
thin contiguous slices. Contrast-enhanced high-
resolution 3D T1-WI sequences provide addi-
tional delineation of the arterial and venous struc-
tures causing neurovascular conflict and other
enhancing lesions like AVM, tumors, and devel-
opmental venous anomalies.
Neuroimaging Findings in TN
An intraoperative finding of arterial or venous
compression of the trigeminal nerve root is seen
in 90–95 % of the patients with CTN as reported
in a series of patients who undergo microvascular
decompression (MVD) [20, 21].
The most commonly implicated vessels are the
SCA, AICA, and the petrosal vein [22]. The supe-
rior cerebellar artery is by far the most common
artery seen compressing the trigeminal nerve,
seen in about 75 % of the cases. Other arteries
associated with compression include the anterior
inferior cerebellar artery (9.6 %), posterior infe-
rior cerebellar artery (0.7 %), vertebral artery
(1.6 %), basilar artery (0.7 %), and labyrinthine
artery (0.2 %). Multivessel conflict may also be
seen [23].
Venous causes of compression include the
transverse pontine vein (identification of which
may change the surgical approach) [24].
The site of neurovascular conflict between the
superior cerebellar artery and trigeminal nerve is
superior, superomedial, or medial. If the culprit
783
vessel is the anterior inferior cerebellar artery, the
site of conflict is inferior or medial (Fig. 1). Con-
flict at the lateral aspect of the nerve is extremely
uncommon. The pattern of venous conflict is
variable [22].
Other vascular causes of TN include aneu-
rysms and vascular malformations [dural arterio-
venous fistulas and arteriovenous malformations
(including microAVMs)] (Fig. 2) [24].
According to increasing severity, MRI findings
are graded as Grade I (simple contact
between nerve and vessel), Grade II (displace-
ment/distortion of nerve roots by vessel), and
Grade III (indentation and thinning of nerve
root) [25].
The sensitivity and specificity of high-
resolution MRI for identification of neurovascular
conflict varied from 52–100 % to 29–93 %,
respectively. Because of the inconsistency of the
results, according to the practice parameters pre-
scribed by the AAN, there is insufficient evidence
to support or refute the usefulness of MRI to
identify vascular contact in CTN or to indicate
the most reliable MRI technique.
Other imaging indicators of TN include an
atrophy of the trigeminal nerve, a small ipsilateral
cerebellopontine angle cistern, a short length of
the cisternal segment of the trigeminal nerve, and
a narrower trigeminopontine angle [25, 26].
Apart from neurovascular conflict, neuroimag-
ing detected other structural causes in 10–18 % of
cases [27].
The pitfalls of MRI include partial volume
effects due to the small size of structures being
imaged and a false-positive rate of 14–21 % of
vascular compression in asymptomatic
individuals [28].
In addition to anatomical imaging, diffusion
tensor imaging (DTI) can objectively quantify
the microstructural changes in the affected nerve.
Coregistering anatomic and DTI images can
enhance the localization of small structures. The
fractional anisotropy in the ipsilateral nerve has
been described to be reduced as compared with
the contralateral side. ADC values remain
unchanged [29].
784
Fig. 1 A sixty-eight-year-old female patient with left tri-
geminal neuralgia. Axial CISS images (a, b), axial time-of-
flight MRA source images (c, d), and coronal
Fig. 2 Other vascular causes of trigeminal neuralgia: (a)
posterior fossa arteriovenous malformation, (b) brainstem
cavernoma, (c) microAVM, and (d) developmental venous
Hemifacial Spasm
Clinical Features
Hemifacial spasm (HFS) is a peripheral move-
ment disorder clinically characterized by parox-
ysms of tonic or clonic contractions involving
D.R. Hingwala and K. Chandrasekharan
reconstruction (e, f) show AICA loop (yellow arrow) cross-
ing inferomedial to the left trigeminal nerve and causing
lateral displacement and thinning of the REZ
anomaly in the vicinity of the trigeminal nucleus (Drawing
adapted from Harsha et al. [24])
muscles innervated by the ipsilateral facial
nerve, predominantly the periorbital or perioral
facial musculature [30]. Clinically, it needs to be
differentiated from other causes of facial dyskine-
sias like blepharospasm, tics, myokymia, focal
seizures, or psychogenic conditions. Usually uni-
lateral, it may be bilateral in about 0.6–5 % cases
[31]. While this condition is not life threatening,
34 Vascular Loop Syndromes
the clinical significance of diagnosing and treating
this disorder lies in its disabling nature [32].
Pathophysiology
Dandy in 1932 [9], Campbell et al. in 1947 [33],
and Gardner in 1962 [34] had initially reported
that vascular compressions of the fifth and seventh
cranial nerves were possible causes of trigeminal
neuralgia and hemifacial spasm.
In 1977, Jannetta et al. [35] were the first to
describe the site of compression of these nerves.
They reported that vascular compression of the
seventh cranial nerve at the REZ was a cause of
HFS. They further elaborated that the
compression of the nerve must be perpendicular
to its course. Peripheral vessels and vessels run-
ning parallel to the nerve were previously not
thought to be associated with HFS. The REZ of
the facial nerve is the point where the cells
forming the myelin sheath change from the central
oligodendroglia to the central Schwann cells,
making it most susceptible to irritation by
mechanical stimulation such as vascular
compression [36].
Other authors have described a far more distal
location of compression midway between the
REZ and the internal acoustic meatus (IAM) or
at the IAM associated with HFS [37].
This compression is seen generally on the ante-
rior surface of the facial nerve and is associated
with typical HFS [38]. Rarely, it can occur at the
rostral and posterior surface of the nerve and is
then associated with atypical HFS [39].
Configuration variation of the facial-
vestibulocochlear nerve complex may also be
associated with the AICA compressing the distal
portion of the facial nerve. In the case report
described by Kawashima et al., the facial nerve
arose 5 mm away from the vestibulocochlear
nerve without forming a complex with it and
was strongly bent posteriorly [40].
Patients with young-onset HFS (<30 years)
account for up to 6 % of the patients overall.
This subgroup may present with similar clinical
features and imaging findings as the older sub-
group. However, the older subgroup has more
785
incidence of hypertension and its related changes
in the vasculature.
HFS has a higher incidence in Asians, proba-
bly due to the smaller size of the posterior fossa.
HFS can also rarely be caused by other vascu-
lar abnormalities like vertebrobasilar ectasia
(Fig. 3) and fusiform aneurysms at the vertebral
artery (VA)–posterior inferior cerebellar artery
(PICA) junction [41] and of the VA [42]. The
VA fusiform aneurysm may also compress the
contralateral facial nerve, causing HFS [43]. How-
ever, in the presence of aneurysms, the possibility
of coexistence of additional causative vessels like
PICA or AICA adjacent to the aneurysm should
be ruled out.
Role of Neuroimaging in HFS
HFS is mostly related to neurovascular compres-
sion of the facial nerve [32, 44]. Microvascular
decompression is the surgical treatment of choice
for providing symptomatic relief to the patients
[45]. The decision to perform this surgery is based
on clinical history and examination. However,
rarely, another cause like a tumor or a demyelin-
ating disorder may be detected [46]. Preoperative
imaging is thus important to rule out a
nonvascular cause. In addition, imaging of the
neurovascular relationships at the REZ and delin-
eating the responsible vessel may be important in
the surgical planning to prevent/minimize
intraoperative complications like hearing loss,
facial paresis, and meningitis [47]. Imaging can
also aid in patient counseling as the detection of a
small compressing artery or vein may be associ-
ated with a higher postoperative recurrence rate of
symptoms [48].
Neuroimaging Findings
On imaging, we have to identify the course of the
facial nerve and follow it from its origin, along the
cisternal segment and up to the intracanalicular
segment for any compression, contact, or inden-
tation. The exact location of these findings on the
circumference of the nerve has to be documented.
786 D.R. Hingwala and K. Chandrasekharan

Fig. 3 A patient with left

hemifacial spasm. Axial
CISS (a) and high-
resolution T2 (b) images
and axial time-of-flight MR
angiography source images
show a tortuous left
vertebral artery (yellow
arrow) compressing the left
seventh to eighth nerve
complex

In addition, the dominance of the vertebrobasilar
system and rotation of the brainstem, if any, are
noted. Finally, as described above, other findings
like aneurysms, arteriovenous malformations, and
nonvascular causes like tumors have to be
excluded.
Neurovascular contact has been found in
86–90 % [49] of the patients on MRA and defor-
mity of the nerves in 56 % of the cases. The MRA
findings correspond with the surgical findings in
87 % of the cases. A false-negative rate of about
15 % has also been observed [50]. CISS sequence
has a sensitivity of 100 % for demonstrating
neurovascular compression [18].
Neurovascular compression of the facial nerve
may also be seen in asymptomatic controls and in
the contralateral asymptomatic side in patients
with unilateral HFS. Thus, this imaging finding
may not always lead to symptoms. According to
Tan et al. [51], the MRI finding of contralateral
NVC has a poor predictive value for the presence
of bilateral symptoms. However, patients with
bilateral symptoms have a significantly higher
incidence of presence of contralateral NVC in
addition to the ipsilateral NVC. Also, in patients
with bilateral symptoms, there is a correlation
between the severity of HFS symptoms and the
degree of neurovascular compression at the REZ.
Other Rare Manifestations of Vascular
Loop Syndrome and Their Imaging
Features
Glossopharyngeal neuralgia is a rare condition
characterized by severe, unilateral paroxysmal
pain in the oropharynx, ear, or both locations,
triggered by specific activities like swallowing,
34 Vascular Loop Syndromes
chewing, or coughing. It may also be manifested
by other sensory phenomena like palatal myoclo-
nus, syncope, or rarely even cardiac arrest. MRI
can depict vascular compression of the
glossopharyngeal nerve at the supraolivary
fossette (especially by the posterior inferior cere-
bellar artery) [52].
Superior oblique myokymia is an unusual
movement disorder with tonic or phasic abnormal
eye movements caused by vascular compression
of the trochlear nerve. High-resolution MRI can
demonstrate arterial contact at the root entry zone
of the affected trochlear nerve [53].
Vascular compression of the vestibulocochlear
nerve has been suggested as a cause of tinnitus
and/or vertigo by many authors [54]. However,
other authors [55] have refuted these findings.
Conclusion
Vascular loop syndromes (TN and HFS) are
caused by vascular compression at the REZ and
focal demyelination. These are often refractory
conditions that may require surgical management
of the neurovascular conflict. Neuroimaging
(MRI) plays an important role in the evaluation
of these patients. The entire cisternal course of the
affected nerve should be imaged in multiple
planes to verify the presence of neurovascular
conflict.
References
1. Langner S, Schroeder HW, Hosten N, Kirsch M (2012)
Diagnosing neurovascular compression syndromes.
Rofo 184(3):220–228
2. Merskey H, Bogduk N (1994) Classification of chronic
pain: descriptions of chronic pain syndromes and def-
initions of pain terms. IASP Press, Seattle, pp 59–71
3. Headache Classification Subcommittee of the Interna-
tional Headache Society (2004) The international clas-
sification of headache disorders. 2nd edition.
Cephalalgia 24(1):126–136
4. Prasad S, Galetta S (2009) Trigeminal neuralgia histor-
ical notes and current concepts. Neurologist 15
(2):87–94
5. Katusic S, Williams DB, Beard CM, Bergstralh EJ,
Kurland LT (1991) Epidemiology and clinical features
of idiopathic trigeminal neuralgia and
787
glossopharyngeal neuralgia: similarities and differ-
ences, Rochester, Minnesota, 1945–1984. Neuroepi-
demiology 10:276–281
6. Gronseth G, Cruccu G, Alksne J, Argoff C, Brainin M,
Burchiel K, Nurmikko T, Zakrzewska JM (2008) Prac-
tice parameter: the diagnostic evaluation and treatment
of trigeminal neuralgia (an evidence-based review)
report of the Quality Standards Subcommittee of the
American Academy of Neurology and the European
Federation of Neurological Societies. Neurology
71:1183–1190
7. Papanastassiou AM, Schwartz RB, Friedlander RM
(2008) Chiari I malformation as a cause of
trigeminal neuralgia: case report. Neurosurgery 63(3):
E614–E615
8. Love S, Coakham HB (2001) Trigeminal neuralgia:
pathology and pathogenesis. Brain 124:2347–2360
9. Dandy WE (1934) Concerning the cause of trigeminal
neuralgia. Am J Surg 24:447–455
10. Hamlyn PJ, King TT (1992) Neurovascular compres-
sion in trigeminal neuralgia: a clinical and anatomic
study. J Neurosurg 76:948–954
11. Peker S, Kurtkaya O, Uzun I, Pamir MN (2006) Micro-
anatomy of the central-peripheral myelin transition
zone of the trigeminal nerve. Neurosurgery 59
(2):354–359
12. Smith KJ, McDonald WI (1980) Spontaneous and
evoked activity due to central demyelinating lesion.
Nature 286(5769):154–155
13. Cruccu G, Biasiotta A, Galeotti F, Iannetti GD,
Truini A, Gronseth G (2006) Diagnostic accuracy of
trigeminal reflex testing in trigeminal neuralgia.
Neurology 66(1):139–141
14. Mursch K, Schafer M, Steinhoff BJ, Behnke – Mursch
J (2002) Trigeminal evoked potentials and sensory
deficits in atypical facial pain – a comparison with
results in trigeminal neuralgia. Funct Neurol
17(3):133–136
15. Bradley WG Jr, Waluch V, Yadley RA, Wycoff RR
(1984) Comparison of CT and MR in 400 patients
with suspected disease of the brain and cervical spinal
cord. Radiology 152(3):695–702
16. Kent DL, Larson EB (1988) Magnetic resonance imag-
ing of the brain and spine. Is clinical efficacy
established after the first decade? Ann Intern Med
108(3):402–424
17. Papanagiotou P, Grunwald IQ, Politi M, Struffert T,
Ahlhelm F, Reith W (2006) Vascular anomalies of the
cerebellopontine angle. Radiologe 46(3):216–222
18. Tarnaris A, Rendowen S, Coakham HB (2007) A com-
parison of magnetic resonance angiography and con-
structive interference in steady state-three-dimensional
Fourier transformation magnetic resonance imaging in
patients with hemifacial spasm. Br J Neurosurg
21(4):375–381
19. Sindou M, Howeidy T, Acevedo G (2002) Anatomic
observations during microvascular decompression for
idiopathic trigeminal neuralgia with correlations
between topography of pain and site of the
788
neurovascular conflict: prospective study in a series of
579 patients. Acta Neurochir (Wien) 144(1):1–12
20. Sindou M, Leston J, Decullier E, Chapuis F (2007)
Microvascular decompression for trigeminal neuralgia:
long-term effectiveness and prognostic factors in a
series of 362 consecutive patients with clear-cut
neurovascular conflicts who underwent pure decom-
pression. J Neurosurg 107(6):1144–1153
21. Meaney JMF, Miles JB, Nixon TE, Whitehouse GH,
Ballantyne ES, Eldridge PR (1994) Vascular contact
with the fifth cranial nerve at the pons in patients with
trigeminal neuralgia: detection with 3D FISP imaging.
AJR 163(6):1447–1452
22. Jannetta PJ (1967) Arterial compression of the trigem-
inal nerve at the pons in patients with trigeminal neu-
ralgia. J Neurosurg 17:159–180
23. Seoane ER, Rhoton AL Jr (1999) Suprameatal exten-
sion of the retrosigmoid approach: microsurgical anat-
omy. Neurosurgery 44(3):553–560
24. Harsha KJ, Kesavadas C, Chinchure S, Thomas B,
Jagtap S (2012) Imaging of vascular causes of trigem-
inal neuralgia. J Neuroradiol 39(5):281–289
25. Kress B, Rasche D, Fiebach J, Tronnier V, Sartor K,
Stippich C (2004) MR volumetry of the trigeminal
nerve in patients with unilateral facial pain. RoFo
176:719–723
26. Ha SN, Kim SH, Yoo EH, Han IB, Shin DA, Cho KG,
Chung SS, Park YS (2012) Patients with idiopathic
trigeminal neuralgia have a sharper-than-normal
trigeminal-pontine angle and trigeminal nerve atrophy.
Acta Neurochir 154(9):1627–1633
27. Goh BT, Poon CY, Peck RH (2001) The importance of
routine magnetic resonance imaging in trigeminal neu-
ralgia diagnosis. Oral Surg Oral Med Oral Pathol Oral
Radiol Endod 92:424–429
28. Tash AR, Sze G, Leslie DR (1989) Trigeminal neural-
gia: MR imaging features. Radiology 172(3):767–770
29. Liu Y, Li J, Butzkueven H, Duan Y, Zhang M, Shu N,
Li Y, Zhang Y, Li K (2013) Microstructural abnormal-
ities in the trigeminal nerves of patients with trigeminal
neuralgia revealed by multiple diffusion metrics. Eur J
Radiol 82(5):783–786
30. Wang A, Jankovic J (1998) Hemifacial spasm: clinical
findings and treatment. Muscle Nerve 21
(12):1740–1747
31. Tan E-K, Jancovik J (1999) Bilateral hemifacial spasm:
a report of five cases and a literature review. Mov
Disord 14:345–349
32. Tan EK, Fook-Chong S, Lum SY, Lim E (2004) Botu-
linum toxin improves quality of life in hemifacial
spasm: validation of a questionnaire (HFS-30).
J Neurol Sci 219(12):151–155
33. Campbell E, Keedy C (1947) Hemifacial spasm: a note
on the etiology in two cases. J Neurosurg 4:342–347
34. Gardner WJ (1962) Concerning the mechanism of tri-
geminal neuralgia and hemifacial spasm. J Neurosurg
19:947–958
35. Jannetta PJ, Abbasy M, Maroon JC, Ramos FM, Albin
MS (1977) Etiology and definitive microsurgical
D.R. Hingwala and K. Chandrasekharan
treatment of hemifacial spasm. Operative techniques
and results in 47 patients. J Neurosurg 47(3):321–328
36. Tomii M, Onoue H, Yasue M, Tokudome S, Abe T
(2003) Microscopic measurement of the facial nerve
root exit zone from central glial myelin to peripheral
Schwann cell myelin. J Neurosurg 99(1):121–124
37. Ryu H, Yamamoto S, Sugiyama K, Uemura K,
Miyamoto T (1998) Hemifacial spasm caused by vas-
cular compression of the distal portion of the facial
nerve. Report of seven cases. J Neurosurg 88
(3):605–609
38. Ho SL, Cheng PW, Wong WC, Chan FL, Lo SK, Li LS,
Tsang KL, Leong LL (1999) A case-controlled
MRI/MRA study of neurovascular contact in
hemifacial spasm. Neurology 53(9):2132–2139
39. Barker FG II, Jannetta PJ, Bissonette DJ, Shields PT,
Larkins MV, Jho HD (1995) Microvascular decom-
pression for hemifacial spasm. J Neurosurg 82
(2):201–210
40. Kawashima M, Yamada M, Sato S, Oka H, Fujii K,
Matsushima T (2009) Hemifacial spasm caused by
vascular compression of the distal portion of the facial
nerve associated with configuration variation of the
facial and vestibulocochlear nerve complex. Turk
Neurosurg 19(3):269–275
41. Choi SK, Rhee BA, Park BJ, Lim YJ (2008)
Hemifacial spasm caused by fusiform aneurysm at
vertebral artery posterior inferior cerebellar artery junc-
tion. J Korean Neurosurg Soc 44(6):399–400
42. Uchino M, Nomoto J, Ohtsuka T, Kuramitsu T (2005)
Fusiform aneurysm of the vertebral artery presenting
with hemifacial spasm treated by microvascular
decompression. Acta Neurochir (Wien) 147
(8):901–903
43. Nakagawa I, Takayama K, Kurokawa S, Wada T,
Nakagawa H, Kichikawa K, Nakase H (2011)
Hemifacial spasm due to contralateral aneurysmal
compression of the facial nerve successfully treated
with endovascular coil embolization: case report.
Neurosurgery 69(3):E768–E772
44. Jannetta P (1970) Microsurgical exploration and
decompression of the facial nerve in hemifacial
spasm. Curr Top Surg Res 2:217–220
45. Chung SS, Chang JH, Choi JY, Chang JW, Park YG
(2001) Microvascular decompression for hemifacial
spasm: a long-term follow-up of 1,169 consecutive
cases. Stereotact Funct Neurosurg 77(1–4):190–193
46. Galvez-Jimenez N, Hanson MR, Desai M (2001)
Unusual causes of hemifacial spasm. Semin Neurol
21(1):75–83
47. Coakham H (2000) The microneurosurgical treatment
of trigeminal neuralgia, hemifacial spasm and
glossopharyngeal neuralgia. In: Robertson HC, Rob-
ertson JH (eds) Cranial base surgery. Churchill Living-
stone, London
48. Engh JA, Horowitz M, Burkhart L, Chang YF, Kassam
A (2005) Repeat microvascular decompression for
hemifacial spasm. J Neurol Neurosurg Psychiatry 76
(11):1574–1580
34 Vascular Loop Syndromes
49. Fukuda H, Ishikawa M, Okumura R (2003) Demon-
stration of neurovascular compression in trigeminal
neuralgia and hemifacial spasm with magnetic reso-
nance imaging comparison with surgical findings in
60 consecutive cases. Surg Neurol 59(2):93–100
50. Tan EK, Chan LL (2006) Young onset hemifacial
spasm. Acta Neurol Scand 114(1):59–62
51. Tan E-K, Chan LL (2004) Clinico-radiologic correla-
tion in unilateral and bilateral hemifacial spasm.
J Neurol Sci 222(1–2):59–64
52. Hiwatashi A, Matsushima T, Yoshiura T, Tanaka A,
Noguchi T, Togao O, Yamashita K, Honda H (2008)
MRI of glossopharyngeal neuralgia caused by
neurovascular compression. AJR 191(2):578–581
789
53. Yousry I, Dieterich M, Naidich TP, Schmid UD,
Yousry TA (2002) Superior oblique myokymia:
magnetic resonance imaging support for the
neurovascular compression hypothesis. Ann Neurol
51(3):361–368
54. McDermott AL, Dutt SN, Irving RM, Pahor AL,
Chavda SV (2003) Anterior inferior cerebellar artery
syndrome: fact or fiction. Clin Otolaryngol 28
(2):75–80
55. Gultekin S, Celik H, Akpek S, Oner Y, Gumus T,
Tokgoz N (2008) Vascular loops at the
cerebellopontine angle: is there a correlation with tin-
nitus? Am J Neuroradiol 29(9):1746–1749
 Migraine
35
Mark Kruit

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 792 This chapter focuses on diagnostic clinical neu-
Pathophysiology of Migraine . . . . . . . . . . . . . . . . . . . . . . . . 793 roimaging in migraine. In most migraine cases,
Diagnostic Neuroimaging Indications
patient history, details of symptoms, and careful
in Migraine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 794 clinical neurological examination are together
AAN Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 794 the most important tools in diagnosing and
EFNS Guidelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 795 treating migraine, and consequently, there is
Summarized Recommendations . . . . . . . . . . . . . . . . . . . . . 795
mostly no need for further laboratory tests or
Migraine and Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 796 neuroimaging. In selected non-acute headache
Migrainous Infarction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 796 cases, neuroimaging is warranted, and recom-
Migraine as a Risk Factor for Clinical
Ischemic Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 803 mendations are provided. Imaging in acute
Migraine as a Risk Factor for Clinical Hemorrhagic headaches is not subject of this chapter.
Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 804 Good understanding of the migraine patho-
Migraine as a Risk Factor for Subclinical Stroke . . . 805 physiology allows better interpretation of neu-
Migraine and T2 Hyperintensities on MRI . . . . . . . 807 roimaging findings notably when patients
White Matter Lesions in Migraine (WMLs) . . . . . . . . 807 present acutely, in or outside an attack. The
“Subclinical” MRI Findings in the Individual neuroradiological findings relevant in under-
Migraine Patient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 811 standing the complex relationship between
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 811 migraine and stroke will therefore be discussed,
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 812
including details on findings during migraine
aura. On MRI scans in migraine patients, fre-
quently, white matter hyperintensities and (cer-
ebellar) infarcts can be present. Knowledge of
the epidemiological findings that have set
migraine today also as a risk factor for (progres-
sive) subclinical brain lesions is relevant in the
everyday neuroradiological practice.

Keywords
Migraine disorders • Headache • Pathophysi-
M. Kruit (*)
ology • MRI • Stroke • Aura • White matter
Department of Radiology, Leiden University Medical
Center, Leiden, The Netherlands lesions • Infarct • Cerebellar infarcts
e-mail: m.c.kruit@lumc.nl

# Springer Science+Business Media New York 2016 791

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_24
792
Introduction
Migraine is a common, multifactorial,
multiphasic, disabling primary headache disorder.
Effects on the brain (structure, neurochemistry,
function) and neurovascular system have been
widely documented during the different phases
of migraine.
The disorder is typically characterized by
recurrent attacks of disabling headache that last
4–72 h and are accompanied by nausea or
vomiting and photophobia and phonophobia
(migraine without aura). The pain is typically
located unilaterally, of moderate to severe
intensity, and has mostly a pulsating character.
The headache phase may be preceded by a
premonitory phase (occurring hours or days
before the headache) and followed by a
resolution phase. In the premonitory and resolu-
tion phases, patients often experience various
symptoms, including hyper- or hypoactivity,
food craving, yawning, fluid retention, and
depression [1].
Up to one third of patients also have neurolog-
ical aura symptoms (migraine with aura): these
focal neurological symptoms usually precede or
sometimes accompany the headache, and
most often affect the visual field, but may also
involve the sensory or motor system, or can affect
speech. Aura symptoms usually develop gradu-
ally over 5–20 min, last for less than 60 min, and
are fully reversible. Headache with the features of
migraine without aura usually follows the aura
symptoms. Less commonly, headache lacks
migrainous features or is completely absent (aura
without migraine) [2]. Migraine is diagnosed
according to the International Classification of
Headache Disorders (ICHD, 2nd edition, 2005;
Table 1) [1].
Epidemiologic studies described that at least
12 % of the general population suffers from reg-
ular migraine attacks [4]. Women are roughly
three times more often affected than men.
Migraine patients suffer a median of 12 migraine
attacks per year; 25 % have at least two attacks per
month. Migraine is rated by the World Health
Organization among the most disabling chronic
M. Kruit
Table 1 Diagnostic criteria for migraine without and with
aura [3]
Migraine without aura
A. At least 5 attacks fulfilling criteria B–D
B. Headache attacks lasting 4–72 h (untreated or
unsuccessfully treated)
C. Headache has at least two of the following
characteristics
a. Unilateral location
b. Pulsating quality
c. Moderate or severe pain intensity
d. Aggravation by or causing avoidance of routine
physical activity (e.g., walking or climbing stairs)
D. During headache, at least one of the following
a. Nausea and/or vomiting
b. Photophobia and phonophobia
E. Not attributed to another disorder
Previously used terms: common migraine, hemicrania
simplex
Migraine with aura
A. At least 2 attacks fulfilling criteria B–D
B. Aura consisting of at least one of the following, but
no motor weakness
a. Fully reversible visual symptoms including positive
features (e.g., flickering lights, spots, or lines) and/or
negative features (i.e., loss of vision)
b. Fully reversible sensory symptoms including
positive features (i.e., pins and needles) and/or negative
features (i.e., numbness)
c. Fully reversible dysphasic speech disturbance
C. At least two of the following
a. Homonymous visual symptoms and/or unilateral
sensory symptoms
b. At least one aura symptom develops gradually over
5 min and/or different aura symptoms occur in
succession over 5 min
c. Each symptom lasts 5 and 60 min
D. Headache fulfilling criteria B–D for migraine
without aura begins during the aura or follows aura
within 60 min
E. Not attributed to another disorder
Previously used terms: Classic or classical migraine,
ophthalmic, hemiparesthetic, hemiplegic or
aphasic migraine, migraine accompagnée, complicated
migraine
disorders [5] and has been estimated to be the
most costly neurological disorder in the
European countries, largely explained by days
missed from work and reduced productivity at
35 Migraine
work [6]. Acute attack treatments are in less than
half of the patients fully satisfactory, and more
effective and better-tolerated prophylactic agents
to prevent attacks are needed. Mainstream
migraine prophylactic treatments have been
largely based on serendipitous observations and
presumed class effects and have disappointing
efficacy and tolerability.
Pathophysiology of Migraine
Understanding of the pathophysiology of
migraine starts in acknowledging that migraine
is not simply a disease of intermittently occurring
pain, but that it relates to processes that over time
affect the brain, or act as triggers on a predisposed
brain. These processes seem to lead to increased
sensitivity or hyperexcitability of brain areas,
allowing headache and aura to develop
paroxysmally [7]. The mechanisms behind such
triggering processes are poorly understood, but
are likely multifactorial, and involve several
brain structures. Glutamatergic-mediated neuro-
nal hyperexcitability seems an important trigger-
ing mechanism [8], but environmental factors
(e.g., physiologic factors, stress, weather), comor-
bid depression, and obesity are also of
influence [9].
Mechanisms Behind Migraine Headache
The migraine headache results from activation of
the trigeminovascular system (TGVS), which
consists of neural connections between dural
blood vessels and trigeminal brainstem nuclei
[10]. Functional imaging illustrated dynamic
involvement of brainstem areas in migraine
[11–15]. In the ictal phase, nociceptive (pain)
signals are projected from the TGVS to the hypo-
thalamus, the thalamus, and higher CNS centers.
These signals are normally controlled by a net-
work involving the hypothalamus and brainstem
areas (periaqueductal grey matter and ventral teg-
mental area) [16, 17]. Paroxysmal dysfunction of
(parts of) this network leads to ineffective
antinociceptive modulation and vascular control,
resulting in the ictal symptoms (pain and auto-
nomic symptoms) [18].
793
Migraine Aura: Cortical Spreading
Depression
The pathophysiological mechanism behind the
migraine aura symptoms is cortical spreading
depression (CSD). This is a transient depression
of activity in neural tissue and is often preceded by
a brief burst of action potential firing, which
slowly propagates in brain tissue [19–21]. CSD
most often involves the occipital lobe, leading to
visual symptoms in about 30 % of patients. Dur-
ing the CSD, regional brain hyper- and
hypoperfusion (supposed to remain above ische-
mic thresholds), reductions of blood–brain barrier
integrity, and plasma extravasation have been
described (see below in section “Neuroimaging
Findings in (Prolonged) Aura”). It is unknown
why CSD occurs episodically, although tempo-
rary changes in excitability seem to be crucial,
and likely already develop gradually in hours to
days before the symptoms present. The potential
role of CSD in activating the migraine headache is
compelling in animal experiments but still contro-
versial in humans [22].
Role of Neuroimaging in Understanding
the Pathophysiology of Migraine
Effects on the brain (structure, neurochemistry,
function) and neurovascular system have been
widely documented during the different phases
of migraine, and neuroimaging has played a sig-
nificant role in the current understanding of path-
ophysiological processes behind migraine.
However, these processes are still only partially
understood, are likely multifactorial, and involve
several brain structures.
Neuroimaging has contributed significantly to
the current neuroscientific knowledge on struc-
tural and functional brain changes during the
ictal phase (aura and headache) of migraine. It is
however largely unknown which parts of the brain
structurally, functionally, or biochemically
change earlier on, during the premonitory phase.
The thalamus, hypothalamus, and probably other
deep brain and brainstem structures seem to play a
role. Further research focusing on such early
changes in the premonitory phase may provide
insight in when and why brainstem nuclei and
pain networks become paroxysmally
794
dysfunctional, how the trigeminovascular system
becomes activated, and what pathophysiological
changes precede and characterize the aura symp-
toms. Given the diagnostic focus of this chapter,
the neuroscientific neuroimaging findings (based
on, e.g., voxel-based morphometry, DTI, MRS,
etc.) behind parts of migraine pathophysiology are
considered out of scope here and will therefore not
be further discussed.
Diagnostic Neuroimaging Indications
in Migraine
The diagnosis of primary headaches is exclusively
a clinical task based on history taking and careful
neurologic examination. In cases of acute head-
ache (like primary thunderclap headache or after
trauma, etc.) or suspected symptomatic headache,
the need for neuroimaging is mostly evident and
will not further be discussed in this chapter. For
non-acute headache, as applies to most migraine
patients, neuroimaging is overused and is only in
selected cases considered to be appropriate.
CT scans and notably MRI scans are frequently
requested and performed in migraine patients who
seek medical help. Often this is driven by patient’s
anxiety about having an underlying pathologic
condition, or to improve patient overall satisfac-
tion and medical care. The fact that radiological
examinations are not particularly invasive or
uncomfortable reduces thresholds further.
However, notably, in patients presenting with
typical primary headaches, the very low likeli-
hood of detecting explanatory underlying diseases
that change treatment or diagnosis must be con-
sidered. Combined results of imaging studies
(CT and MRI) in over 3,700 headache patients
(not exclusively “typical migraine”) together
show a low yield of about 0.4 % in migraine
patients [23]. In “typical migraine” patients, the
yield is likely to be even lower.
Potential risk of “unnecessary imaging” is the
discovery of an incidental finding, which eventu-
ally needs further diagnostic workup of follow-up.
Further, potential problems include false-positive
studies, false reassurance from an inadequate
study, allergic reaction to contrast agent, etc.
M. Kruit
Furthermore, the current resource-restricted med-
ical environment more and more requires
evidence-based justification of diagnostic
imaging.
AAN Recommendations
In 2000, the quality standards subcommittee of the
American Academy of Neurology (AAN)
published an evidence-based guideline on the role
of neuroimaging in patients with headache [24], to
assist the physicians in making appropriate choices
in diagnostic workup. A total of 28 studies (from
1966 to 1998) were reviewed. Based on reported
rates of “abnormalities related to headache that
may require further action” (like acute cerebral
infarct, neoplastic disease, hydrocephalus, aneu-
rysm, or arteriovenous malformation) combined
with data from patient history and neurological
examination, the following symptoms were identi-
fied to significantly increase the odds of finding a
significant abnormality on neuroimaging in
patients with non-acute headache [24]:
• Rapidly increasing headache frequency
• History of lack of coordination
• History of localized neurologic signs or a his-
tory such as subjective numbness or tingling
• History of headache causing awakening from
sleep (although this can occur with migraine
and cluster headache)
Based on these findings, the following AAN
recommendations for non-acute headache were
formulated:
• Consider neuroimaging in patients with an
unexplained abnormal finding on the neuro-
logic examination (Grade B).
• Consider neuroimaging in patients with atypi-
cal headache features or headaches that do not
fulfill the strict definition of migraine or other
primary headache disorders (or have some
additional risk factor, such as immune defi-
ciency), when a lower threshold for neuroim-
aging may be applied (Grade C).
35 Migraine
• Neuroimaging is not usually warranted in
patients with migraine and a normal neurologic
examination (Grade B).
EFNS Guidelines
The European Federation of Neurological Socie-
ties (EFNS) published a similar guideline for the
management of non-acute headache (revision in
2010) [25], also mainly based on review of
published evidence. This EFNS guideline
includes the following summarized statements
relevant for migraine (grade B recommendations):
• In adult and pediatric patients with migraine,
with no recent change in pattern, no history of
seizures, and no other focal neurological signs
or symptoms, the routine use of neuroimaging
is not warranted.
• Exception of these rules should be made in the
diagnosis of trigeminal autonomic headaches
and headaches that are aggravated by exertion
or a Valsalva-like maneuver [23].
• In patients with atypical headache patterns, a
history of seizures, or neurological signs or
symptoms or symptomatic illness such as
tumors, acquired immunodeficiency syndrome
(AIDS), and neurofibromatosis, MRI may be
indicated.
• When neuroimaging is warranted, the most
sensitive method should be used, and MRI
(and not CT) is recommended in these cases.
• With a normal unenhanced MRI, in the
absence of other diseases and suspicion on
metastasis/vasculitis/etc., there is no need for
additional scanning with gadolinium.
• There is no role for conventional Röentgen
techniques.
• Digital subtraction angiography is not appro-
priate in the screening of patients with head-
ache for intracranial disease.
Summarized Recommendations
In the recent meta-analysis by Detsky et al., data
from over 3,700 patients were included [23]. The
795
Table 2 Recommendation on the use of diagnostic neu-
roimaging in non-acute headache
A. Consider neuroimaging in non-acute headache
patients with “red flags”
Unexplained abnormal findings on neurologic
examination
(Atypical) headaches that do not fulfill ICHD-II
criteria for primary headaches
Additional risk factors (e.g., immune deficiency,
tumors, etc.)
History of (associated) seizures
Recent changes in headache pattern
Nonvisual (e.g., sensory or motor) or atypical aura
pattern
B. Neuroimaging is not usually warranted in patients
with typical migraine with or without aura and
normal neurological examination
C. When neuroimaging is warranted, MRI is method
of primary choice
recommendations from this meta-analysis are
consistent with the guidelines by AAN and
EFNS, although additional recommendations
were included to perform imaging in
cases (1) with nonvisual of aura (sensory or
motor), (2) an aura that has changed in character,
or (3) an aura that cannot be clearly described as
typical of migraine aura.
Since a 2007 case series demonstrated that
even cluster headache with a typical time
pattern and an excellent response to typical
treatment can still be caused by underlying
structural pathology such as a pituitary tumor,
patients with trigeminal autonomic headaches
should be considered to undergo
neuroimaging [26].
In summary, patient history, details of symp-
toms, and careful clinical neurological examina-
tion are together the most important tools in
diagnosing and treating migraine. In most patients
with non-acute headache, this will lead to a reli-
able diagnosis (applying the ICHD criteria) and
do not require any further laboratory tests or
neuroimaging.
Table 2 summarizes the combined recommen-
dations on the use of neuroimaging in non-acute
headache patients.
796
Migraine and Stroke
Accumulating evidence from the last 3–4 decades
has expanded the spectrum of neurovascular
pathology linked to migraine. Initial case reports
of “migrainous stroke” were followed by retro-
spective and prospective mostly hospital-based
case-control studies assessing the prevalence of
clinical ischemic and hemorrhagic stroke in
migraine patients, showing a consistent associa-
tion between migraine with aura and stroke; the
association with migraine without aura is less
evident [27, 28]. MRI studies have further identi-
fied that migraine is also associated with markers
of small vessel disease, including (progressive)
white matter lesions (WMLs), brainstem T2
hyperintensities, posterior circulation subclinical
infarcts, and microbleeds [29–31]. And finally,
reports from epidemiological studies on associa-
tions between migraine and coronary events [32,
33] and all-cause mortality [34] further illustrate
the broad spectrum of lesions associated with
migraine, likely to be explained via complex
relationships [35].
The relationship between migraine and stroke
is complex, and the following paragraphs will
expand on different aspects of this relationship;
neuroimaging examples will illustrate how
migraine patients with (suspicion of) hemorrhagic
or ischemic stroke may present.
First, migrainous infarction and its imaging
appearances will be described, to be followed by
a paragraph on “aura-related” neuroimaging find-
ings, because the clinical symptoms of either
“infarction” or “aura” are often very similar in
the acute and subacute moments of presentation.
Migrainous Infarction
Kurth et al. suggested that probably the first report
on migrainous infarction came from Féré, who
described a patient with migraine who died after
2 months of headache, visual disturbances, and
hemiplegia [35, 36]. Various case reports of
migrainous infarction have been published since
then and have made clear that migraine can act as
a direct cause of ischemic stroke. In such cases,
M. Kruit
Table 3 Diagnostic criteria for migrainous infarction [3]
A. The present attack in a patient with migraine with aura
is typical of previous attacks except that one or more aura
symptoms persist for >60 min
B. Neuroimaging demonstrates ischemic infarction in a
relevant area
C. Not attributed to another disorder
stroke is assumed to be directly and causally
related to an acute migraine attack.
Because it is often impossible by clinical
examination alone to differentiate between TIA,
prolonged aura, and migrainous infarction, MRI
(notably with diffusion weighting) today plays a
key role in the diagnosis of migrainous infarction.
The current ICHD-II criteria strictly define
migrainous infarction as ischemic stroke that
occurs when during a typical migraine with aura
attack, one or more migrainous aura symptoms
persist longer than 60 min, with neuroimaging
proof of an associated ischemic brain lesion in
an appropriate region and absence of other under-
lying causes (Table 3).
From this definition, it is indirectly evident that
migraine patients who present with “prolonged
aura symptoms” require an appropriate neuroim-
aging workup with MRI and, when an acute ische-
mic lesion is present, additional diagnostic
workup to exclude other underlying disorders. In
cases of coexisting other causes (like cardiac
arrhythmia, coagulation disorders, embolism
through a patent foramen ovale, cervical artery
dissection), the diagnosis then has to be (changed
to) ischemic stroke coexisting with migraine. The
same applies when in a patient with a history of
migraine without aura, an ischemic lesion
develops during or after a migraine attack. In
other cases, when criteria are not completely ful-
filled, ischemic stroke in a migraine patient may
be categorized as cerebral infarction of other
causes presenting with symptoms resembling
migraine with aura [3].
In the past decades, the diagnostic criteria for
migrainous infarction have been changed (ICHD-
I vs. ICHD-II), and studies inconsistently applied
the criteria. This explains probably the relatively
wide range (0.8–3.4 per 100,000) of reported
35 Migraine 797

annual incidences and points at a probable amount was relatively favorable. Although no studies
of overdiagnosis [37]. Although this implies that have systematically examined the appearance of
migrainous infarction is a rare condition, which is migrainous infarcts, notably when supraten-
further illustrated by Wolf et al. who estimated torially, from a number of case reports, it is
that it accounts for approximately two among suggested that the ischemic insults predominantly
1,000 “overall” strokes per year [38], it needs to may affect the cortex. Similarly, cortical ischemia
be considered that migrainous infarction predom- that crosses different vascular territories may also
inantly affects younger patients. In that age cate- point at a migrainous infarct mechanism, but this
gory, migrainous infarction was estimated to is probably an infrequent finding, since in the
account for 13 % of first-ever ischemic study by Wolf et al., no such “crossing” lesions
strokes [39]. were identified, neither on DWI nor on PWI
[38]. However, aura-related hypoperfusion (see
Neuroimaging Findings in Migrainous below) quite typically seems to cross territories.
Infarction The underlying mechanisms of migrainous
The largest series of migrainous infarction cases infarction are unknown, but are probably related
to date are reported by Wolf et al. (17 cases) [38] to CSD-related changes, including
and Laurell et al. (33 cases) [40]. In both reports, hypoperfusion, and changes in blood–brain bar-
patients underwent an appropriate stroke workup, rier permeability (which might lead to an exacer-
and diagnoses according to ICHD-II criteria were bation of local cellular injury caused by ischemia).
reported. Table 4 shows the main study findings. Together with factors predisposing to
Both studies reported a clear predominance of coagulopathy and release of vasoactive neuropep-
infarcts in the posterior circulation, supporting tides, further changes in cerebral hemodynamics,
previous observations. The low age at stroke arterial thrombosis, and infarctions may be
onset is a further key finding in both studies; explained [41].
therefore, when treating a young patient In the Figs. 1, 2, 3, and 4, case descriptions
presenting with stroke, migrainous infarction illustrate various presentations and appearances of
should be kept in mind. In both studies, outcome migrainous infarcts. In Fig. 5, a case with “cere-
bral infarction presenting with symptoms resem-
Table 4 Recent case series of migrainous infarction bling migraine with aura” is described and
Laurell illustrates that it can be difficult to apply a correct
et al. [40] Wolf et al. [38] and meaningful diagnosis, given the strict ICHD-
Cases n = 33 (all n = 17 (n = II criteria.
ICHD-II) 11 ICHD-II)a
M:F 39 %:61 % 23 %:77 % Neuroimaging Findings
Age at stroke 19–76, median 20–71, mean
onset 39 years 45 years
in (Prolonged) Aura
Posterior 82 % 71 %
According to the ICHD-II criteria, a diagnosis of
circulation persistent aura without infarct can be applied
Cerebellum 21 % 6% when aura symptoms remain present longer than
Multiple lesions 41 % 1 week and when there is no neuroradiological
Family history of 75 % 24 % evidence of ischemia. This is a rare condition that
migraine seems to affect genetic forms of migraine (like
Patent foramen 40 % 65 % familial hemiplegic migraine) somewhat more
ovale
often.
ICHD-II = International Classification of Headache Dis-
In “regular” migraine with aura patients,
orders, Second Edition [3]
a
n = 6 had a history of MO (thus not fulfilling the ICHD-II aura symptoms incidentally also persist longer
criteria in the strict sense) and presented with first-ever than 60 min. When more than one aura
neurologic symptoms compatible with MA and concomi- symptom is present (e.g., visual and sensory
tant migraine headaches
symptoms together or in succession) for each
798
Fig. 1 Bilateral occipital migrainous infarction. A
33-year-old female patient with migraine without aura
from childhood and visual aura attacks since age 21. She
presented with usual visual aura symptoms but with now a
persisting visual field defect, and persisting positive scin-
tillating scotoma, accompanied by migraine headache. No
other neurological signs and symptoms. Family history for
migraine with aura was positive. An MRI scan performed
after 1 day of symptoms revealed bilaterally in the occipital
lobe small cortical areas of diffusion restriction. There
type, 60 min may be accepted. When aura persists
longer, this might point at migrainous
infarction, although most often symptoms still
spontaneously normalize, and probably, only inci-
dentally, patients will be scanned. Imaging studies
in such “non-infarct” cases by definition do not
show ischemic changes, but various case reports
and series have described other CSD- or aura-
M. Kruit
were no other abnormalities. CTA of the cervical and
intracranial arteries was negative (not shown). There
were no other underlying causes. Follow-up MRI after
3 weeks showed only minimal residual hyperintensity on
the FLAIR images, consistent with nearly normalization of
the ischemic foci. Images: a1 and a2 FLAIR images; a3 and
a4 corresponding B1000 diffusion-weighted images in the
acute setting; b1 and b2 FLAIR images after 3 weeks
follow-up
related effects on the brain tissue and
neurovascular system that will be discussed in
the next paragraphs.
Cutrer et al. [42] and Sanchez del Rio et al. [43]
studied spontaneous migraine episodes with
perfusion-weighted MRI, including six patients
studied during regular (not prolonged) visual
aura, within 31  6 min after the onset of visual
35 Migraine
Fig. 2 Bilateral occipital and thalamic migrainous
infarction. A 61-year-old female patient, with long history
of migraine with aura, presented with persisting left upper
quadrant visual field defect that had developed during a
regular migraine with visual aura attack. Initial noncontrast
CT (a) dubiously showed some reduced gray-white matter
differentiation in the right occipital lobe. CTA (not shown)
showed normal caliber of the carotids, the vertebrobasilar
system and the posterior cerebral arteries; the posterior
symptoms. In all studies, perfusion deficits were
observed in the occipital visual cortex from which
the hemifield defect was originating. Maximum
measured changes were 37 % decrease in CBF,
33 % decrease in rCBV, and 82 % increase in
799
communicating arteries were not identified. Whole brain
perfusion CT (b) demonstrated reduced CBV and CBF and
prolonged MTT and TTP values in the right occipital lobe,
but also to a less degree in the left occipital lobe. MRI after
one day (c) confirmed recent bilateral infarction, with signs
of hemorrhagic transformation on the left side (arrow
head), but also identified right-sided thalamic infarction.
In the following diagnostic workup, no other underlying
causes were identified
MTT. Several small series and case reports have
been published since then, mostly with consistent
findings of mild regional hypoperfusion, uni- or
bilaterally affecting overlapping vascular
territories.
800
Fig. 3 Bilateral cerebellar migrainous infarction. A
45-year-old male patient with migraine with aura and
visual aura attacks since age 35, with an average attack
frequency of 2 years, presented with usual visual aura
symptoms that lasted longer than normal, and were accom-
panied by sensory symptoms over his whole body, and
diplopia and dysarthria. Non-enhanced CT (upper row)
Förster et al. reported a prospective study in
which patients with suspected acute ischemic
stroke were evaluated. In this study, 33 patients
with a final diagnosis of migraine with aura were
compared with age-matched patients with a final
diagnosis of acute ischemic stroke [44]. As a
consequence of the study methodology, in this
cohort, the number of patients with “rare” aura
symptoms (like hemihypesthesia, hemiparesis,
aphasia) was overrepresented, as well as “acute
onset” of symptoms. In 54 % (n = 18) of migraine
with aura patients, PWI showed hypoperfusion
that was involving more than the PCA territory
in all but one of the patients, although the PCA
territory was predominantly involved in 61 % of
cases. In seven patients (39 %), the hypoperfusion
extended to the parietal, temporal, or frontal lobe.
M. Kruit
was performed after few hours and showed bilateral cere-
bellar hypodensities consistent with bilateral cerebellar
infarction. A subsequent MRI scan (T2 images, lower
row) confirmed the presence of three cerebellar and one
vermian infarct. MRA of the cervical and intracranial
arteries was negative (not shown). No other underlying
causes were identified
There was no clear association between clinical
symptoms and location of perfusion changes.
There were no diffusion-weighted imaging
(DWI) abnormalities related to the aura symp-
toms, and there were no vessel occlusions or ste-
noses on MRA. In comparison with acute
ischemic stroke patients, the aura patients more
often had hypoperfusion involving more than one
territory and less increased TTP and MTT ratios.
A few reports pointed at the occurrence of
“crossed-cerebellar diaschisis” in cases with
aura-related perfusion changes. Dodick
et al. reported a migraine patient with typical
attacks of sensory aura during 30–60 min
followed by headache, who showed reversible
reduction in CBF in the left cerebral hemisphere
and associated crossed-cerebellar diaschisis
35 Migraine
Fig. 4 Subacute right occipital migrainous infarct. A
58-year-old male patient with migraine with aura and
visual and sometimes sensory aura attacks since childhood
presented with a history of recent persisting visual aura
symptoms for >1 week, followed by partial spontaneous
recovery. MRI was performed to exclude ischemia or other
underlying causes. The scan was performed 12 days after
(hypoperfusion in the right cerebellar hemisphere)
during a typical attack on a brain SPECT scan
[45]. Iizuka et al. observed a case with prolonged
aura crossed-cerebellar hyperperfusion on brain
SPECT scan on day 2 after symptom onset,
which was explained by the authors as a conse-
quence of uncoupled hyperperfusion with low
function in the left cerebral hemisphere
(corresponding to the neurological deficits). Both
the crossed hyperperfusion and hypoperfusion
illustrate the possibility of associated flow (and
metabolic) alterations distant to and opposite of
the primary site of disturbances, which could be
relevant in understanding the occurrence of silent
801
the start of symptoms and showed a subacute occipital
corticosubcortical infarct on the right: hyperintensity on
FLAIR and T2 slices (open arrows); cortical diffusion
restriction and signs of cortical necrosis (arrows). Note a
small lipoma in the vermis, initially diagnosed as suspi-
cious for an AVM (arrowhead)
ischemic lesions in the cerebellum in migraine
patients (see later).
Besides flow alterations, a variety of other
imaging findings have been described on brain
imaging during (prolonged) migraine aura that
all together may be explained by temporary
changes in blood–brain barrier function that are
probably secondary to aura-related cortical
spreading depression and/or spreading
hypoperfusion. A number of reports described
“vasogenic leakage,” that may present as regional
sulcal hyperintensity on native FLAIR MRI
images [46], as (delayed) subarachnoid (sulcal),
leptomeningeal, or cortical gadolinium
802 M. Kruit

Fig. 5 (continued)
35 Migraine
Fig. 5 Cerebral infarction presenting with symptoms
resembling migraine with aura. A 46-year-old female
patient with migraine with aura and regular attacks of
(left-sided) visual aura since age 22, woke up in the morn-
ing with right-sided hemiparesis and aphasia, with associ-
ated migraine-like headache, followed in the hospital by
unilateral pulsating headache with nausea and vomiting,
resembling her prior migraine attacks. The noncontrast CT
in the acute setting showed slight hypodensity (not shown).
An MRI scan 1 day later confirmed a cortically restricted
zone of cytotoxic edema in the left MCA territory involv-
ing parts of the frontoparietal and insular cortex, consistent
with her symptoms (a, T2-weighted images; b, B1000
enhancement on FLAIR or T1-weighted MRI
scans [47–50] or as increased vascular permeabil-
ity on perfusion-weighted MRI [51]. In few cases
with (prolonged) aura, symptoms may be associ-
ated with reversible cortical swelling and
hyperintense signal changes on FLAIR images,
distributed along the regional cortical ribbon
corresponding to the symptoms, without clear
diffusion restriction (vasogenic edema) [52].
Migraine as a Risk Factor for Clinical
Ischemic Stroke
Besides the (rare) presentation of “migrainous
infarction” (i.e., directly related to a migraine
attack), migraine patients have also a higher
chance to present with an ischemic or hemor-
rhagic stroke unrelated to a migraine attack.
803
diffusion-weighted images; c, sagittal and coronal FLAIR
images). Additional diagnostic workup remained negative
for other underlying causes. The patient recovered
completely, although a follow-up MRI scan after 1 year
(d, FLAIR images) showed postischemic cortical paren-
chymal loss. The cortically restricted infarct was consid-
ered somewhat unusual for ordinary MCA stroke and
might have been due to aura-related CSD. However,
because the infarction was not clearly temporally related
to a migraine with aura attack that was similar to previous
attacks, the diagnosis was cerebral infarction presenting
with symptoms resembling migraine with aura and not
migrainous infarction
Over the past four decades, many observa-
tional studies, hospital-based stroke case-control
studies, and population-based studies have evalu-
ated the association between migraine and clinical
ischemic stroke. Meta-analyses [27, 53] of these
studies summarized that migraine patients are
about at doubled increased risk, which notably
applies to patients with migraine with aura.
Sch€urks calculated a pooled relative risk of 1.7
(95 % CI 1.3–2.3) for migraineurs compared to
controls [27]. From eight studies, data specified
by migraine subtype was available, resulting in a
pooled relative risk of 2.2 (95 % CI: 1.5–3.0) for
patients with migraine with aura and 1.2 (95 % CI:
0.9–1.7) for patients with migraine without aura,
compared to controls. A higher migraine fre-
quency seems also to further increase the risk of
ischemic stroke. Due to the lower prevalence of
migraine in men, the association between
804
migraine and ischemic stroke is less certain
in men.
The risk is highest in women, notably at youn-
ger age (<45 years), and in this group, the risk
further increases (up to 10x increased risk) with
concurrent use of oral contraceptives. In individ-
ual studies, effects of concurrent hypertension and
heavy smoking showed a greater than multiplica-
tive effect on the risk [54, 55]. Associations with
markers of endothelial dysfunction and factors
linked to prothrombotic or proinflammatory con-
ditions have also been suggested as explanatory
for the increased stroke risk in migraine
[56]. Genetic factors, including polymorphisms
in the MTHFR, ACE, MEPE, and IRX4 genes,
have been linked with both migraine and ischemic
stroke, but further studies need to assess the path-
ophysiological relevance of such findings.
Topographic and Pathophysiological
Considerations
Few structured data exist on the topography of
non-migrainous infarcts in migraine patients. In a
large series of 3,500 patients with acute stroke,
130 (3.7 %) had active migraine, and about 50 %
of these were <45 years of age. In the younger
patients, posterior circulation involvement (55 %)
was characteristic [57]. From case reports and
small series, it was also suggested that the occipital
lobe and/or the posterior cerebral artery territory
seem to be overrepresented in migraine patients
with clinical ischemic stroke. Recently, Oygarden
et al. retrospectively investigated whether the
lesion pattern on diffusion-weighted MRI scans
was different between migraineurs (with and with-
out aura; n = 196) and non-migraineurs (n = 720)
with clinical ischemic stroke [58]. In the migraine
group, younger patients and women were overrep-
resented, and infarcts were more often due to
cardioembolism and less often due to small vessel
disease than in the control group. Migraine patients
presented more often with symptoms from the
posterior circulation and had more cortical
(OR 1.8 CI: 1.3–2.5), small (OR 1.9 CI:
1.04–3.5), cerebellar (P = 0.026), and occipital
infarcts (13.3 % vs. 8.6 %; P = 0.05). Migraine
patients with infarcts had a three times higher
chance to have a patent foramen ovale (PFO).
M. Kruit
Oygarden et al. suggested that the higher fre-
quency of smaller lesions in migraineurs with aura
may point at a higher vulnerability to damage
following minor ischemic insults (similarly as
was described in experiments in mice with famil-
ial hemiplegic migraine) or that ischemia in
migraineurs may simultaneously lead to a small
“clinically silent” ischemic lesion (that normally
would not be noticed) and ischemia-triggered
CSD resulting in “clinical” stroke like neurologi-
cal deficits leading to hospital admission [58].
Although the earlier study results were criti-
cized (e.g., because of potential misclassification,
referral bias, lack of neuroimaging proof, etc.), the
consistent findings in several studies over the
years have migraine set today as an acknowledged
ischemic stroke risk factor. It has to be noted that
in absolute terms, stroke in young women with
migraine (estimated at 5.5/100,000 annually) [59]
remains rare, although at the same time, in this age
group, migraine should be considered as an
important risk factor.
Migraine as a Risk Factor for Clinical
Hemorrhagic Stroke
Data on a possible association between migraine
and hemorrhagic stroke have remained inconsis-
tent for years. However, in 2013, Sacco
et al. meta-analyzed a total of four case-control
and four cohort studies that together included
320,539 individuals in whom 1,600 intracerebral
or subarachnoid hemorrhages occurred
[28]. Migraine patients were found to be at signif-
icantly increased risk for hemorrhagic stroke
(OR 1.48; 95 % CI 1.2–1.9; P = 0.001). There
was no clear difference between migraine with
aura (OR 1.6; 95 % CI 0.9–3.0; P = 0.13) or
migraine without aura (OR 1.4; 95 % CI
0.7–2.62 P = 0.3), male and female patients, or
younger and older subjects with migraine,
although Kuo et al. reported higher risks of
migraine with aura in male patients and patients
<45 years of age [60]. A limitation in the review
by Sacco et al. was that subarachnoid and intrapar-
enchymatous hemorrhages could not be sepa-
rated. Next to ischemic stroke, migraine patients
35 Migraine
are thus likely also at increased risk for hemor-
rhagic stroke, although mechanisms likely differ
and remain to be elucidated, and subgroups most
at risk still need to be identified.
Migraine as a Risk Factor for Subclinical
Stroke
Posterior Circulation Infarcts
The population-based cross-sectional CAMERA
MRI study (n = 435) assessed whether migraine
cases were at increased risk of several types of
“silent” (or subclinical) brain lesions and whether
certain areas of the brain were particularly vulner-
able [61]. None of the participants reported a
history of stroke or transient ischemic attack or
showed relevant abnormalities at standard neuro-
logical examination.
In the migraine with aura group, 8 % of sub-
jects had one or more posterior circulation
infarcts. This was significantly higher compared
to controls (0.7 %; P = 0.005) and migraineurs
without aura (2 %). The highest risk was found in
MA with 1 attack per month (OR, 15.8; 95 % CI,
1.8–140). Infarct size ranged from 2 to 21 mm.
Most lesions were located in the cerebellum
(Fig. 6), typically in a borderzone location. The
average number of lesions per subject was 1.8.
Although migraine-related clinical strokes seem
to have a predilection for the occipital lobes (see
above), in the CAMERA study, none of the pos-
terior circulation infarcts were in the occipital
lobes.
Migraine patients with posterior circulation
infarcts were significantly older, but cardiovascu-
lar risk factors were not more prevalent, and the
presence of these lesions was not significantly
associated with supratentorial brain changes,
such as WMLs. These two observations suggest
that the lesions are not atherosclerotic in origin.
The combination of vascular distribution, deep
borderzone location, shape, size, and imaging
characteristics on MR imaging makes it likely
that the lesions have an infarctious origin. The
most likely etiologic mechanism seems to be
hypoperfusion and/or embolism, rather than ath-
erosclerosis or small vessel disease.
805
The higher risk for subjects with higher attack
frequency may point at migraine attack-related
mechanisms. During and after migraine attacks,
sluggish cerebral flow below an ischemic thresh-
old has been described [42, 43, 62–64]. A
decrease in brain perfusion pressure (e.g., during
migraine) theoretically affects the clearance and
destination of embolic particles; narrowing of the
arterial lumen and endothelial abnormalities stim-
ulate formation of thrombi; occlusive thrombi
further reduce blood flow and brain perfusion.
Because the deep cerebellar territories have a pat-
tern of progressively tapering arteries with only
few anastomoses present, they are likely to be
particularly vulnerable to hypoperfusion-related
borderzone infarct mechanisms. This
hypoperfusion-related concept matches the find-
ings of previous studies in which the small cere-
bellar borderzone infarcts, in particular when
multiple, were strongly associated with severe
occlusive and/or (artery-to-artery) embolic dis-
ease based on vertebrobasilar atherosclerosis,
likely to result in hypoperfusion and infarction.
The population-based AGES-Reykjavik study
(n = 4,689) confirmed these findings and also
reported a significantly higher prevalence of cer-
ebellar infarcts on MRI scans in female migraine
with aura patients at late life (23 % vs. 15 %; P <
0.001) [65]. In that study, there was no increased
risk for cortical or subcortical infarcts on MRI for
migraine patients.
In the 9-year follow-up CAMERA 2 study,
66 % of the original sample was rescanned with
the same MRI scanners and protocols. None of the
infarcts present at baseline had disappeared. Only
in the migraine group new posterior circulation
infarcts had occurred (5 % vs. 0 %; P = 0.07) [31].
Other Population-Based Evidence
for Silent Infarcts in Migraine
In the EVA study (Epidemiology of Vascular Age-
ing; n = 780), migraine with aura patients (n =
17) had over a threefold increased risk (OR 3.4;
95 % CI 1.2–9.3) compared to controls (n = 617)
for any infarct on MRI, and there was a suggestion
that migraine with aura patients were at increased
risk for multiple infarcts (OR 3.7; 95 % CI
0.8–17) [66]. In this study, most infarcts were
806 M. Kruit

Fig. 6 Cerebellar infarcts in migraine with aura patients. Corresponding T2-weighted (left) and FLAIR (right) MRI
images showing (multiple) cerebellar infarcts (arrowheads) in three migraine with aura patients from the CAMERA study

located outside the cerebellum or brainstem, and
the number of patients was unfortunately too
small for further statistical testing of specific
infarct locations like the cerebellum (5.9 %
vs. 2.8 %) and thalamus (11.8 % vs. 2.1 %) in
migraine with aura versus control participants.
The MRI substudy of the NOMAS study
(Northern Manhattan Study) included n =
546 racial/ethnically diverse population-based
participants; 65 % were Hispanic. Migraine
patients had a significantly higher risk of subclin-
ical brain infarcts (OR 2.1; 95 % CI 1.0–4.2),
which was even higher in migraineurs
without aura (OR 2.6; 95 % CI 1.3–5.5).
In the groups of participants >75 years of
age, 30 % of migraineurs had at least
1 infarct on MRI, compared to 15 % of controls.
In this study, infarcts were found most commonly
in the white matter (13 %) and cerebellum
(10 %) [67].
35 Migraine
In summary, an association between migraine
and subclinical infarcts is well established by now.
Similarly as for clinical ischemic stroke, notably,
migraine with aura patients are at highest risk. The
exact mechanisms and potential consequences
need further evaluation and research. The poste-
rior circulation territory is probably most vulner-
able in migraine patients.
Migraine and T2 Hyperintensities
on MRI
White Matter Lesions in Migraine
(WMLs)
Already from the early days of MRI, reports exist
on the presence of WMLs on T2-weighted images
in migraine patients. Initial clinic-based studies
presented data on WMLs in migraine patients, but
results were often uncontrolled, inconsistent, or
conflicting. Some studies evaluated migraine sub-
types, also with inconsistent results. One study
found that migraine attack frequency was related
to an increased prevalence of WMLs. Despite the
various limitations and discrepancies in those MRI
studies, a meta-analysis [29] summarized the
results of seven case-control studies and reported
that migraine patients were at 4 increased risk
(OR: 3.9; 95 % CI 2.3–6.7) for WMLs, regardless
of comorbidities like cardiovascular risk factors,
demyelinating disease, inflammatory conditions,
and valvular heart disease.
Fig. 7 Focal small- to medium-sized deep WMLs in a 47-year-old female migraine patient
807
WMLs in Migraine: The CAMERA-Study
1 and 2
Because the clinic-based studies mentioned above
suffered from methodological difficulties, and
might have evaluated a more than average severe
subgroup of migraine patients (due to their
“clinic-based” origin), the cross-sectional popula-
tion-based CAMERA MRI study was carried out
(see above in paragraph “Migraine as a Risk Fac-
tor for Clinical Ischemic Stroke”). Among women
of the CAMERA cohort, the risk of having a high
deep WML load (top 20th percentile of the distri-
bution of deep WML load, Fig. 7) was increased
in migraineurs compared to controls (OR 2.1;
95 % CI 1.0–4.1). This risk was higher in those
with higher attack frequency (1 attack/per
month; OR 2.6; 95 % CI 1.2–5.7) and was similar
among women with migraine with and without
aura. Among men, prevalence of deep WMLs
did not differ between controls and migraineurs.
No association was found between severity of
periventricular WMLs and migraine, irrespective
of gender or migraine frequency or subtype.
The finding of higher risks in those with a
higher attack frequency was suggestive of a
potential causal relationship between migraine
attacks or attack severity and the development of
WMLs. Therefore, the 9-year follow-up CAM-
ERA 2 study (n = 286, mean age 57  8 years)
was carried out to test for accumulative effects of
recurrent attacks and to study the underlying
mechanisms and possible cognitive
consequences [31].
808
Fig. 8 FLAIR images showing preexisting (arrowhead) and newly developed (arrows) deep WMLs after 9 years of
follow-up in a 37-year-old (at baseline) female patient with migraine without aura
In the follow-up study, there were again no
differences in WMLs between male participants
with migraine versus controls. However, deep
WML volume was higher in female migraineurs
compared to controls (P = .04), and their deep
WML progression was more severe (77 %
vs. 60 %; P = .02), which was highest in migraine
without aura (83 %). Multivariate logistic regres-
sion showed that migraine was independently
associated with deep WML progression (OR 2.1;
95 % CI 1.0–4.1; P = .04). The increase in total
deep WML volume was explained by an
increased number of new lesions, rather than by
increase in size of preexisting lesions. Figure 8
shows an example of incident lesions in a
female migraine patient after 9 years of
follow-up. The mean size of individual
hyperintensities at follow-up did not differ
between participants with migraine and controls.
Among females with migraine, deep WMLs
appeared to be more diffusely distributed com-
pared to controls (Fig. 9). Hypertension and dia-
betes were not associated with a higher incidence
of DWMH progression. Exploratory analyses
showed no association of number of migraine
attacks, migraine attack duration, migraine fre-
quency, type of attack, migraine therapy, with
deep WML progression.
M. Kruit
WMLs in Migraine: Other Population-
Based Evidence
In the longitudinal EVA study (n = 780; mean age
69  3), participants with a history of any severe
headache (including migraine) were at increased
risk of higher WML volumes (OR 2.0, 95 % CI
1.3–3.1) compared to controls, with similar find-
ings for deep and periventricular WMLs [66]. In
participants with migraine with aura, the associa-
tions with WMLs were strongest.
A subset of participants of the ARIC study
(Atherosclerosis Risk in Communities cohort
study; n = 1,028) received two MRI examina-
tions, between 8 and 12 years apart. The authors
showed also in this study that migraine without
aura (OR 1.9; 95 % CI 1.0–3.4) is associated with
WMLs in cross-sectional analysis. However, they
failed to demonstrate that migraine was associated
with WML volume progression. Differences in
lesion–quantification methodology, in the defini-
tion of progression, in the older age, size, and
other characteristics of the cohort, etc., might
explain why the findings seem to contrast with
the findings of the CAMERA 2 study.
Unexpectedly, in the NOMAS study (n = 546;
two thirds of participants was >60 years of age),
no association between WML volume and
migraine or its subgroups was found. The authors
35 Migraine
Fig. 9 Topography and progression of deep WMLs in
female migraine patients and controls. Baseline and
follow-up deep WMLs are projected on transparent 3D
suggested that no difference was detected because
of the high burden of other cardiovascular risk
factors in the racially diverse older cohort.
Pathophysiological Mechanisms
Although the pathologic substrate of WMLs in
migraine remains unknown, the most likely histo-
logical substrate of these abnormalities is incom-
plete infarction with changes such as gliosis and
demyelination [68]. The causative mechanisms,
however, remain largely unknown. Different
explanatory options have been described.
Attack-related mechanisms may play a role, as
was suggested by the higher risk of WMLs in the
CAMERA 1 study and as seems likely given the
occurrences of migrainous infarcts. During
attacks, reduced blood flow in large and/or small
arteries [43], possibly in combination with vaso-
constriction or activation of the clotting system/
platelets, might lead to formation of local thrombi.
Alternatively, local tissue changes during
migraine attacks, such as excessive neuronal acti-
vation, neurogenic inflammation, neuropeptide
and cytokine release [69], or excitotoxity [70],
may occur, and such changes may directly lead
809
maps (normalized for differences in group size; controls,
N = 52; migraine with aura, N = 75; migraine without
aura, N = 57)
to tissue damage. Reversible MRI abnormalities
during migraine aura, including areas of increased
vasogenic leakage [52] and evidence of
blood–brain barrier dysfunction in prolonged
aura [47, 48, 50], seem to illustrate the possibility
of direct effects to the brain during attacks, most
likely first acting on the level of the microvascu-
lature [71], and possibly being enhanced by other
factors like matrix metalloproteinase nine-
dependent cascade mechanism, that may increase
the risk of local tissue damage [72].
On the other hand, because the progression of
WMLs and occurrence of infarcts in the CAM-
ERA cohort were not dependent on persisting
migraine activity, and also because of the attack-
unrelated increased risk of clinical ischemic
strokes in migraineurs, attack-unrelated factors
also seem to play a role. With increasing age,
when attacks generally diminish, other systemic
migraine “disease-related” conditions leading to
WMLs are possibly increasing and likely compli-
cate the detection of attack-related mechanisms.
Attack-unrelated factors could, e.g., include
chronic procoagulatory or proinflammatory
changes due to endothelial dysfunction [73, 74],
810
Fig. 10 Infratentorial hyperintense lesions (IHLs) in
migraine. T2-weighted images of baseline and 9-year
follow-up of three migraine patients from the CAMERA
study, showing increasing loads of IHLs: either newly
elevated homocysteine levels [75], or recurrent
paradoxical (micro-) emboli due to right-to-left
shunts [76]. In addition, population-based evi-
dence that migraine (with aura) is associated
with a higher cardiovascular risk profile might
contribute to the risk of (ischemic) brain
lesions [77].
Brainstem Lesions
In the CAMERA 1 study, also infratentorial
hyperintense lesions (IHLs) were significantly
more prevalent in migraine patients (4.4.%) com-
pared with controls (0.7 %; P = 0.04) [61]. The
majority of the IHLs were located in the pons. The
CAMERA 2 study showed that after 9 years
follow-up, the prevalence of IHLs remained
higher in women with versus without migraine
(21 % vs. 4 %; adjusted OR, 6.5; 95 % CI,
1.5–28.3; P = .01) and that they also more often
showed progression of IHLs (15 % vs. 2 %;
adjusted OR, 7.7; 95 % CI, 1.0–59.5; P = .05)
M. Kruit
developed lesions (a, arrowheads), additional lesions (b,
open arrows), or increases in size (c, arrows), compared to
baseline
[31]. Figure 10 shows typical examples of T2
hyperintensities in the brainstem and their pro-
gression over time.
Typically, brainstem lesions were located in the
dorsal basis pontis, adjacent to the tegmentum, at
the level of, and slightly cranial to, the entry zone
of the trigeminal nerve. In nearly all cases, lesions
were located bilaterally, sometimes extending to
the midline; none reached the surface of the pons.
Hyperintensities seem to involve the
pontocerebellar fibers, the pontine nuclei, or the
nucleus reticularis tegmenti pontis, and, in some
cases, parts of the corticopontine (pyramidal tract)
or medial lemniscus fibers. These anatomical
locations are supplied by the anteromedial and
anterolateral groups arising from the basilar
artery.
Earlier reports in non-migraineurs linked
pontine IHLs to patients with cardiovascular
risk factors, leukoaraiosis, lacunar infarcts,
and poor clinical outcome after stroke
35 Migraine
[78]. Histopathologically, pontine IHLs corre-
spond to myelin pallor and reactive astrocytosis.
The pathophysiology of these lesions is assumed
to be comparable as for lesions in subcortical
arteriosclerotic encephalopathy, i.e., ischemia sec-
ondary to small artery sclerosis. There is a good
correlation between MRI findings and histopa-
thology [79]. Similar hyperintense brainstem
lesions are also frequent in CADASIL, a disorder
in which migraine with aura is often the
presenting symptom [80]. In CADASIL,
decreased cerebral perfusion secondary to
changes in the wall of cerebral arteries leads to
early damage of the white matter. Regions that are
irrigated by the longest perforating arteries are
most vulnerable to hypoperfusion. This is partic-
ularly the case for the central part of the pons.
Similarly, as discussed above for DWMLs,
migraine attacks might be causally related to the
brainstem lesions, because repeated or prolonged
reduced perfusion has been described in migraine
attacks, although not specifically in the pons
[64]. Alternatively, attack-unrelated “systemic”
factors could also explain the association (see
above), may be in combination with suggested
impaired adaptive cerebral hemodynamic mecha-
nisms in the posterior circulation of migraine
patients.
“Subclinical” MRI Findings
in the Individual Migraine Patient
The knowledge from hospital-based and
population-based studies that migraine is an inde-
pendent risk factor for DWMLs, IHLs, and sub-
clinical infarcts does not imply that when such
lesions are identified on a brain MRI scan in a
migraine patient, migraine is “the cause.” In rou-
tine neuroradiological practice, T2 hyperintense
lesions in the white matter are frequently encoun-
tered in subjects above age 50 but also appear in
younger individuals. No studies found clear pref-
erential locations or distribution patterns for
WMLs migraine patients. Therefore, in most
patients with migraine and WMLs, the lesions
will remain “aspecific.” The identification of a
limited number of small- to medium-sized lesions
811
in adults with migraine does therefore generally
not need further medical attention.
Incidentally, WMLs are observed on MRI
scans in relatively young patients, e.g., below
age 40. In such cases, number, location, aspect,
and distribution of the lesions have to be carefully
evaluated, and based on the images and clinical
history together, the likelihood of other diseases
that are associated with white matter lesions, such
as multiple sclerosis, vasculitis, CADASIL,
MELAS, coagulation disorders, cardiac abnor-
malities, etc., has to be considered. “Migraine” is
part of this differential diagnostic list, but other
options should never be disregarded.
Similarly, the identification of one or more
silent cerebellar infarcts on an MRI scan can’t
directly be attributed to “migraine,” although
these findings – with the data from the CAMERA
study, showing 8 % of patients with migraine with
aura from the general population affected – are not
so unusual in migraine patients with aura. In such
cases, if the observation is incidental, no direct
clinical consequence seems to be necessary. If
future research indicates that there is a risk of
progression of number or size of lesions or if
there are associated functional consequences,
additional study of causes and evaluation of use-
fulness of preventive therapy has to be initiated.
However, a patient who presents with an acute
cerebellar infarct, irrespective of the size of the
lesion or presence of migraine, has to be screened
for (embolic) sources of the infarction and treated
accordingly.
Summary
Migraine is a very prevalent disorder. Knowledge
of the migraine pathophysiology allows better
interpretation of neuroimaging findings in
migraine patients, who may present both in acute
situations and in regular outpatient settings.
A migraine patient may present during a
migraine attack with neurologic deficits. Normal
neurologic aura symptoms need to be differenti-
ated from “prolonged aura” and ischemic stroke.
When a clinical diagnosis is not possible, urgent
neuroimaging is mandatory. Neuroimaging may
812
show heterogeneous changes during “prolonged
aura” that mostly can be differentiated from find-
ings in true ischemia. If ischemia is identified
during a migraine attack, this might be either due
to true migrainous infarction, due to ischemic
stroke coexisting with migraine, or due to cerebral
infarction of other causes presenting with symp-
toms resembling migraine with aura.
Migraine patients, notably those with aura and
in the younger age category, are also at increased
risk of clinical ischemic of hemorrhagic stroke
unrelated to a migraine attack. The exact
mechanisms behind these associations are still
unknown, although attack-unrelated “systemic”
proinflammatory, prothrombotic, endothelial, car-
diac, and genetic factors probably play a role and
may be part of “migraine as a disease.” Migraine
is associated with subclinical infarcts,
supratentorial, and infratentorial T2 hyperintense
lesions on MRI. The higher prevalence of such
lesions is probably explained by similar patho-
physiological mechanisms. Migraine is part of
the differential diagnostic list of WMLs and cere-
bellar infarcts, but other options should never be
disregarded.
Neuroimaging in migraine patients is overused
in non-acute situations; recommendations have
been formulated to identify patient in whom neu-
roimaging is warranted.
References
1. Olesen J (2004) The international classification of
headache disorders, 2nd edn (ICDH-II). J Neurol
Neurosurg Psychiatry 75:808–811. doi:10.1136/
jnnp.2003.031286
2. Goadsby PJ (2001) Migraine, aura, and cortical spread-
ing depression: why are we still talking about it?
AnnNeurol 49:4–6
3. Olesen J (2006) International classification of headache
disorders, second edition (ICHD-2): current status and
future revisions. Cephalalgia 26:1409–1410.
doi:10.1111/j.1468-2982.2006.01214.x
4. Launer LJ, Terwindt GM, Ferrari MD (1999) The prev-
alence and characteristics of migraine in a population-
based cohort: the GEM study. Neurology 53:537–542
5. Menken M, Munsat TL, Toole JF (2000) The global
burden of disease study: implications for neurology.
Arch Neurol 57:418–420
M. Kruit
6. Linde M, Gustavsson A, Stovner LJ, Steiner TJ,
Barré J, Katsarava Z, Lainez JM, Lampl C, Lantéri-
Minet M, Rastenyte D, Ruiz de la Torre E, Tassorelli C,
Andrée C (2012) The cost of headache disorders in
Europe: the Eurolight project. Eur J Neurol
19:703–711. doi:10.1111/j.1468-1331.2011.03612.x
7. Sprenger T, Borsook D (2012) Migraine changes the
brain: neuroimaging makes its mark. Curr Opin Neurol
25:252–262. doi:10.1097/WCO.0b013e3283532ca3
8. Ligthart L, De VB, Smith AV, Ikram MA, Amin N,
Hottenga JJ, Koelewijn SC, Kattenberg VM, de Moor
MH, Janssens AC, Aulchenko YS, Oostra BA, de Geus
EJ, Smit JH, Zitman FG, Uitterlinden AG, Hofman A,
Willemsen G, Nyholt DR, Montgomery GW, Terwindt
GM, Gudnason V, Penninx BW, Breteler M, Ferrari
MD, Launer LJ, van Duijn CM, van den Maagdenberg
AM, Boomsma DI (2011) Meta-analysis of genome-
wide association for migraine in six population-based
European cohorts. Eur J Hum Genet 19:901–907
9. Breslau N, Lipton RB, Stewart WF, Schultz LR, Welch
KM (2003) Comorbidity of migraine and depression:
investigating potential etiology and prognosis. Neurol-
ogy 60:1308–1312
10. Goadsby PJ, Lipton RB, Ferrari MD (2002) Migraine –
current understanding and treatment. N Engl J Med
346:257–270
11. Weiller C, May A, Limmroth V, Juptner M, Kaube H,
Vonschayck R, Coenen HH, Diener HC (1995) Brain-
stem activation in spontaneous human migraine
attacks. Nat Med 1:658–660
12. Cao Y, Aurora SK, Nagesh V, Patel SC, Welch KMA
(2002) Functional MRI-BOLD of brainstem structures
during visually triggered migraine. Neurology
59:72–78
13. Sánchez del Rio M, Linera JA (2004) Functional neu-
roimaging of headaches. Lancet Neurol 3:645–651
14. May A (2009) New insights into headache: an
update on functional and structural imaging
findings. Nat Rev Neurol 5:199–209. doi:10.1038/
nrneurol.2009.28
15. Stankewitz A, May A (2011) Increased limbic and
brainstem activity during migraine attacks following
olfactory stimulation. Neurology 77(5):476–482.
doi:10.1212/WNL.0b013e318227e4a8
16. Miranda-Cardenas Y, Rojas-Piloni G, Martinez-
Lorenzana G, Rodriguez-Jimenez J, Lopez-Hidalgo M,
Freund-Mercier MJ, Condes-Lara M (2006) Oxytocin
and electrical stimulation of the paraventricular hypo-
thalamic nucleus produce antinociceptive effects that
are reversed by an oxytocin antagonist. Pain
122:182–189
17. Holden JE, Farah EN, Jeong Y (2005) Stimulation of
the lateral hypothalamus produces antinociception
mediated by 5-HT1A, 5-HT1B and 5-HT3 receptors
in the rat spinal cord dorsal horn. Neuroscience
135:1255–1268
18. Akerman S, Holland PR, Goadsby PJ (2011) Dience-
phalic and brainstem mechanisms in migraine. Nat Rev
Neurosci 12:570–584. doi:10.1038/nrn3057
35 Migraine
19. Ayata C (2009) Spreading depression: from serendipity
to targeted therapy in migraine prophylaxis.
Cephalalgia 29:1095–1114
20. Welch KM, Cao Y, Aurora S, Wiggins G, Vikingstad
EM (1998) MRI of the occipital cortex, red nucleus,
and substantia nigra during visual aura of migraine.
Neurology 51:1465–1469
21. Lauritzen M (1994) Pathophysiology of the migraine
aura. The spreading depression theory. Brain 117
(Pt 1):199–210
22. Zhang X, Levy D, Kainz V, Noseda R, Jakubowski M,
Burstein R (2011) Activation of central trigemino-
vascular neurons by cortical spreading depression.
Ann Neurol 69:855–865
23. Detsky ME, McDonald DR, Baerlocher MO,
Tomlinson GA, McCrory DC, Booth CM (2006)
Does this patient with headache have a migraine or
need neuroimaging? JAMA 296:1274–1283.
doi:10.1001/jama.296.10.1274
24. Silberstein SD (2000) Practice parameter: evidence-
based guidelines for migraine headache (an evidence-
based review): report of the Quality Standards Sub-
committee of the American Academy of Neurology.
Neurology 55:754–762
25. Sandrini G, Friberg L, Coppola G, Jänig W, Jensen R,
Kruit M, Rossi P, Russell D, Sanchez Del Rìo M,
Sand T, Schoenen J (2011) Neurophysiological tests
and neuroimaging procedures in non-acute headache
(2nd edition). Eur J Neurol 18:373–381. doi: 10.1111/
j.1468-1331.2010.03212.x
26. Wilbrink LA, Ferrari MD, Kruit MC, Haan J (2009)
Neuroimaging in trigeminal autonomic cephalgias:
when, how, and of what? Curr Opin Neurol
22:247–253. doi:10.1097/WCO.0b013e32832b4bb3
27. Schurks M, Rist PM, Bigal ME, Buring JE, Lipton RB,
Kurth T (2009) Migraine and cardiovascular disease:
systematic review and meta-analysis. BMJ 339:b3914.
doi:10.1136/bmj.b3914
28. Sacco S, Ornello R, Ripa P, Pistoia F, Carolei A (2013)
Migraine and hemorrhagic stroke: a meta-
analysis. Stroke 44(11):3032–3038. doi:10.1161/
STROKEAHA.113.002465
29. Swartz RH, Kern RZ (2004) Migraine is associated
with magnetic resonance imaging white matter abnor-
malities: a meta-analysis. Arch Neurol 61:1366–1368.
doi:10.1001/archneur.61.9.1366
30. Kruit M, Launer L, Van Buchem M, Terwindt G,
Ferrari M (2005) MRI findings in migraine. Rev
Neurol (Paris) 161:661–665
31. Palm-Meinders IH, Koppen H, Terwindt GM, Launer
LJ, Konishi J, Moonen JME, Bakkers JTN, Hofman P
a M, van Lew B, Middelkoop H a M, van Buchem M a,
Ferrari MD, Kruit MC (2012) Structural brain changes
in migraine. JAMA 308:1889–1897. doi:10.1001/
jama.2012.14276
32. Kurth T, Gaziano JM, Cook NR, Logroscino G, Diener
H-C, Buring JE (2006) Migraine and risk of cardiovas-
cular disease in women. JAMA 296:283–291.
doi:10.1001/jama.296.3.283
813
33. Bigal ME, Kurth T, Santanello N, Buse D, Golden W,
Robbins M, Lipton RB (2010) Migraine and cardio-
vascular disease: a population-based study. Neurology
74:628–635. doi:10.1212/WNL.0b013e3181d0cc8b
34. Gudmundsson LS, Scher AI, Aspelund T, Eliasson JH,
Johannsson M, Thorgeirsson G, Launer L, Gudnason
V (2010) Migraine with aura and risk of cardiovascular
and all cause mortality in men and women: prospective
cohort study. BMJ 341:c3966. doi:10.1136/bmj.c3966
35. Kurth T, Chabriat H, Bousser M-G (2012) Migraine
and stroke: a complex association with clinical impli-
cations. Lancet Neurol 11:92–100. doi:10.1016/
S1474-4422(11)70266-6
36. Féré C (1883) Note sur un cas de migraine ophtalmique
à accès répétés suivis de mort. Rev Med 3:194–201
37. Bousser M-G (2004) Estrogens, migraine, and
stroke. Stroke 35:2652–2656. doi:10.1161/01.
STR.0000143223.25843.36
38. Wolf ME, Szabo K, Griebe M, Förster A, Gass A,
Hennerici MG, Kern R (2011) Clinical and MRI
characteristics of acute migrainous infarction.
Neurology 76:1911–1917. doi:10.1212/
WNL.0b013e31821d74d5
39. Arboix a, Massons J, García-Eroles L, Oliveres M,
Balcells M, Targa C (2003) Migrainous cerebral infarc-
tion in the Sagrat Cor Hospital of Barcelona stroke
registry. Cephalalgia 23:389–394
40. Laurell K, Artto V, Bendtsen L, Hagen K, Kallela M,
Meyer EL, Putaala J, Tronvik E, Zwart JA, Linde M
(2011) Migrainous infarction: A Nordic multicenter
study. Eur J Neurol 18:1220–1226. doi: 10.1111/
j.1468-1331.2011.03364.x
41. Kruit MC, Launer LJ, Ferrari MD, van Buchem M
(2005) Infarcts in the posterior circulation territory
in migraine. The population-based MRI CAMERA
study. Brain 128:2068–2077. doi:10.1093/brain/
awh542
42. Cutrer FM, Sorensen a G, Weisskoff RM,
Ostergaard L, Sanchez del Rio M, Lee EJ, Rosen BR,
Moskowitz M a (1998) Perfusion-weighted imaging
defects during spontaneous migrainous aura. Ann
Neurol 43:25–31. doi:10.1002/ana.410430108
43. Sanchez del Rio M, Bakker D, Wu O, Agosti R,
Mitsikostas D, Ostergaard L, Wells W, Rosen B,
Sorensen G, Moskowitz M, Cutrer F (1999) Perfusion
weighted imaging during migraine: spontaneous visual
aura and headache. Cephalalgia 19:701–707.
doi:10.1046/j.1468-2982.1999.019008701.x
44. Förster A, Wenz H, Kerl HU, Brockmann MA, Groden
C (2014) Perfusion patterns in migraine with aura.
Cephalalgia 34:11 870–876
45. Dodick DW, Roarke MC (2008) Crossed cerebellar
diaschisis during migraine with prolonged aura: a pos-
sible mechanism for cerebellar infarctions. Cephalalgia
28:83–86. doi:10.1111/j.1468-2982.2007.01458.x
46. Gómez-Choco M, Capurro S, Obach V (2008)
Migraine with aura associated with reversible sulcal
hyperintensity in FLAIR. Neurology 70:2416–2418.
doi:10.1212/01.wnl.0000314693.57386.f0
814
47. Iizuka T, Sakai F, Suzuki K, Igarashi H, Suzuki N
(2006) Implication of augmented vasogenic leakage
in the mechanism of persistent aura in sporadic hemi-
plegic migraine. Cephalalgia 26:332–335.
doi:10.1111/j.1468-2982.2005.01025.x
48. Iizuka T, Sakai F, Yamakawa K, Suzuki K, Suzuki N
(2004) Vasogenic leakage and the mechanism of
migraine with prolonged aura in Sturge-Weber syn-
drome. Cephalalgia 24:767–770. doi:10.1111/j.1468-
2982.2004.00769.x
49. Lanfranconi S, Corti S, Bersano A, Costa A, Prelle A,
Sciacco M, Bresolin N, Ghione I (2009) Aphasic
and visual aura with increased vasogenic leakage: an
atypical migrainosus status. J Neurol Sci 285:227–229.
doi:10.1016/j.jns.2009.06.009
50. Smith M, Cros D, Sheen V (2002) Hyperperfusion with
vasogenic leakage by fMRI in migraine with prolonged
aura. Neurology 58:1308–1310
51. Rotstein DL, Aviv RI, Murray BJ (2012) Migraine with
aura associated with unilateral cortical increase in vas-
cular permeability. Cephalalgia 32:1216–1219.
doi:10.1177/0333102412462286
52. Resnick S, Reyes-Iglesias Y, Carreras R, Villalobos E
(2006) Migraine with aura associated with reversible
MRI abnormalities. Neurology 66:946–947.
doi:10.1212/01.wnl.0000203646.05540.ec
53. Etminan M, Takkouche B, Isorna FC, Samii A (2005)
Risk of ischaemic stroke in people with migraine: sys-
tematic review and meta-analysis of observational stud-
ies. BMJ 330:63. doi:10.1136/bmj.38302.504063.8F
54. Chang CL, Donaghy M, Poulter N (1999) Migraine
and stroke in young women: case–control study. The
World Health Organisation Collaborative Study of Car-
diovascular Disease and Steroid Hormone Contracep-
tion. BMJ 318:13–18
55. Tzourio C, Kittner SJ, Bousser MG, Alperovitch A
(2000) Migraine and stroke in young women.
Cephalalgia 20:190–199
56. Tietjen G (2009) Migraine as a systemic vasculopathy.
Cephalalgia 29:987–996
57. Milhaud D, Bogousslavsky J, Van Melle G, Liot P
(2001) Ischemic stroke and active migraine. Neurology
57:1805–1811
58. Oygarden H, Kvistad CE, Bjørk M, Thomassen L,
Waje-Andreassen U, Naess H (2014) Diffusion-
weighted lesions in acute ischaemic stroke patients
with migraine. Acta Neurol Scand Suppl 129:41–46.
doi:10.1111/ane.12236
59. WHO Collaborative Study of Cardiovascular Disease
and Steroid Hormone Contraception (1996) Ischaemic
stroke and combined oral contraceptives: results of an
international, multicentre, case–control study. WHO
Collaborative Study of Cardiovascular Disease and
Ste. Lancet 348:498–505
60. Kuo C-Y, Yen M-F, Chen L-S, Fann C-Y, Chiu Y-H,
Chen H-H, Pan S-L (2013) Increased risk of hemor-
rhagic stroke in patients with migraine: a population-
based cohort study. PLoS One 8:e55253. doi:10.1371/
journal.pone.0055253
M. Kruit
61. Kruit MC, van Buchem M, Hofman PM, Bakkers JTN,
Terwindt GM, Ferrari MD, Launer LJ, Zaidat OO
(2004) Migraine as a risk factor for subclinical brain
lesions. JAMA 291:427–434. doi:10.1001/
jama.291.4.427
62. Woods RP, Iacoboni M, Mazziotta JC (1994) Brief
report: bilateral spreading cerebral hypoperfusion dur-
ing spontaneous migraine headache. N Engl J Med
331:1689–1692
63. Olesen J, Friberg L, Olsen TS, Iversen HK, Lassen NA,
Andersen AR, Karle A (1990) Timing and topography
of cerebral blood flow, aura, and headache during
migraine attacks. Ann Neurol 28:791–798
64. Bednarczyk EM, Remler B, Weikart C, Nelson AD,
Reed RC (1998) Global cerebral blood flow, blood
volume, and oxygen metabolism in patients with
migraine headache. Neurology 50:1736–1740
65. Scher AI, Gudmundsson LS, Sigurdsson S,
Ghambaryan A, Aspelund T, Eiriksdottir G, van
Buchem M a, Gudnason V, Launer LJ (2009) Migraine
headache in middle age and late-life brain infarcts.
JAMA 301:2563–2570. doi:10.1001/jama.2009.932
66. Kurth T, Mohamed S, Maillard P, Zhu Y-C, Chabriat H,
Mazoyer B, Bousser M-G, Dufouil C, Tzourio C
(2011) Headache, migraine, and structural brain lesions
and function: population based epidemiology of vas-
cular ageing-MRI study. BMJ 342:c7357. doi:10.1136/
bmj.c7357
67. Monteith T, Gardener H, Rundek T, Dong C,
Yoshita M, Elkind MSV, DeCarli C, Sacco RL, Wright
CB (2014) Migraine, white matter hyperintensities,
and subclinical brain infarction in a diverse
community: the northern Manhattan study.
Stroke 45:1830–1832. doi:10.1161/
STROKEAHA.114.005447
68. Rocca MA, Colombo B, Pratesi A, Comi G, Filippi M
(2000) A magnetization transfer imaging study of the
brain in patients with migraine. Neurology 54:507–509
69. Goadsby PJ (1997) Pathophysiology of migraine: a dis-
ease of the brain. In: Goadsby PJ, Silberstein SD (eds)
Headache. Butterworth-Heinemann, Boston, pp 5–25
70. Eggers AE (2001) New neural theory of migraine. Med
Hypotheses 56:360–363
71. Ghabriel MN, Lu MX, Leigh C, Cheung WC, Allt G
(1999) Substance P-induced enhanced permeability of
dura mater microvessels is accompanied by pro-
nounced ultrastructural changes, but is not dependent
on the density of endothelial cell anionic sites. Acta
Neuropathol 97:297–305
72. Gursoy-Ozdemir Y, Qiu J, Matsuoka N, Bolay H,
Bermpohl D, Jin H, Wang X, Rosenberg GA, Lo EH,
Moskowitz MA (2004) Cortical spreading depression
activates and upregulates MMP-9. J Clin Invest
113:1447–1455. doi:10.1172/JCI200421227.The
73. Lee ST, Chu K, Jung KH, Kim DH, Kim EH, Choe VN,
Kim JH, Im WS, Kang L, Park JE, Park HJ, Park HK,
Song EC, Lee SK, Kim M, Roh JK (2008) Decreased
number and function of endothelial progenitor cells in
patients with migraine. Neurology 70:1510–1517
35 Migraine
74. Tietjen GE, Herial N a, White L, Utley C, Kosmyna JM,
Khuder S a (2009) Migraine and biomarkers of endothe-
lial activation in young women. Stroke 40:2977–2982.
doi:10.1161/STROKEAHA.109.547901
75. Hamed S a (2009) The vascular risk associations with
migraine: relation to migraine susceptibility and pro-
gression. Atherosclerosis 205:15–22. doi:10.1016/j.
atherosclerosis.2008.10.016
76. Schwedt TJ, Demaerschalk BM, Dodick DW (2008)
Patent foramen ovale and migraine: a quantitative sys-
tematic review. Cephalalgia 28:531–540
77. Scher AI, Terwindt GM, Picavet HSJ, Verschuren
WMM, Ferrari MD, Launer LJ (2005) Cardiovascular
risk factors and migraine: the GEM population-based
815
study. Neurology 64:614–620. doi:10.1212/01.
WNL.0000151857.43225.49
78. Mantyla R, Pohjasvaara T, Vataja R, Salonen O,
Aronen HJ, Standertskjold-Nordenstam CG, Kaste M,
Erkinjuntti T (2000) MRI pontine hyperintensity after
supratentorial ischemic stroke relates to poor clinical
outcome. Stroke 31:695–700
79. Pullicino P, Ostrow P, Miller L, Snyder W, Munschauer
F (1995) Pontine ischemic rarefaction. Ann Neurol
37:460–466
80. Chabriat H, Mrissa R, Levy C, Vahedi K, Taillia H,
Iba-Zizen MT, Joutel A, Tournier-Lasserve E, Bousser
MG (1999) Brain stem MRI signal abnormalities in
CADASIL. Stroke 30:457–459
 Moyamoya Disease (Spontaneous
Occlusion of the Circle of Willis) 36
Akira Yamamoto, Tomohisa Okada, and Jun C. Takahashi

Contents Clinical Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 827

An Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 827
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 819 Ischemic Symptoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 827
Epidemiological Features . . . . . . . . . . . . . . . . . . . . . . . . . . 819 Hemorrhage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 828
Geographical Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . 819 Headache . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 829
Gender and Age Distribution . . . . . . . . . . . . . . . . . . . . . . . . 819 Involuntary Movement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 831

Familial Moyamoya Disease . . . . . . . . . . . . . . . . . . . . . . . 821 Diagnostic Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 831

Current Diagnostic Criteria
Pathogenesis of Moyamoya Disease . . . . . . . . . . . . . . . 821 of Moyamoya Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 831
Cytokines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 821 Definitive Moyamoya Disease and Probable
Genetic Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 822 Moyamoya Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 832
CT and CT Angiography in Moyamoya Disease . . . . 832
Asymptomatic Moyamoya Disease . . . . . . . . . . . . . . . . 822
MRI/MRA Diagnosis of Moyamoya Disease . . . . . . . 834
Unilateral Moyamoya Disease . . . . . . . . . . . . . . . . . . . . . 824 FLAIR Ivy Sign . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 834
Feasibility of 3-T MRI/MRA
Quasi-moyamoya Disease . . . . . . . . . . . . . . . . . . . . . . . . . . 825 in Moyamoya Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 836
Systemic Vessel Involvement Susceptibility Weighted Imaging in Moyamoya
of Moyamoya Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 825 Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 837

Pathological Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 827 Staging by Conventional Angiography

in Moyamoya Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 838
Angiographical Staging Progression
of Moyamoya Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 840
Aneurysm in Moyamoya Disease . . . . . . . . . . . . . . . . . . . 841
Roles of Morphological and Hemodynamic
Functional Imaging for MMD . . . . . . . . . . . . . . . . . . . . . 842
Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 842
Staging of Moyamoya Disease by MRI/MRA . . . . . . 844
Altered Perfusion Status in MMD . . . . . . . . . . . . . . . . 845
A. Yamamoto (*) • T. Okada Evaluation of Perfusion Status . . . . . . . . . . . . . . . . . . . . 846
Department of Diagnostic Imaging and Nuclear Medicine, Oxygen-15 PET . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 847
Graduate School of Medicine, Kyoto University, SPECT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 847
Sakyo-ku, Kyoto, Japan CT Perfusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 848
e-mail: yakira@kuhp.kyoto-u.ac.jp; MR Perfusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 848
tomokada@kuhp.kyoto-u.ac.jp Arterial Spin Labeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 848
J.C. Takahashi
Department of Neurosurgery, National Cerebral and
Cardiovascular Center, Suita City, Osaka, Japan
e-mail: juntak@ncvc.go.jp

# Springer Science+Business Media New York 2016 817

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_25
818 A. Yamamoto et al.

CBF Changes After Bypass Surgery . . . . . . . . . . . . . . 849 Abstract

Visualization of Lenticulostriate Moyamoya is a disease affecting the cerebral
Artery (LSA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 850 arteries, which predisposes individuals to
Treatment of Moyamoya Disease . . . . . . . . . . . . . . . . . . 851 recurrent ischemic attack in association with
Treatment for Ischemic Moyamoya Disease . . . . . . . . 851 progressive steno-occlusive change of the
Treatment for Hemorrhagic Moyamoya Disease . . . . 855 intracranial internal carotid arteries and their
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 857 proximal branches. The disorder is typically
characterized by a reduction in blood flow in
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 858
the major vessels of the anterior circulation of
the brain. This results in the development of a
compensatory collateral vasculature around
the stenosed vessels, near the apex of the inter-
nal carotid, on the cortical surface,
leptomeninges, and branches of the external
carotid artery and the base of the skull. In
advanced cases, the posterior circulation is
also involved, including the basilar and poste-
rior cerebral arteries. While it was first
described as a “hypoplasia of the bilateral
internal carotid arteries,” the characteristic
appearance of the network of abnormally
dilated collateral vessels on conventional angi-
ography was compared to “a puff of cigarette
smoke” (Arch Neurol 20(3):288–899, 1969) or
“moyamoya” in Japanese. The designation
“spontaneous occlusion of the circle of Willis”
was recently suggested as an alternative by the
International Classification of Diseases (ICD).
Moyamoya disease patients are predominantly
found in Eastern Asian countries such as Japan
and the Republic of Korea. Moyamoya disease
patients show bimodal distribution, adult type
and pediatric type. Symptom of adult
moyamoya patients can be either ischemic or
hemorrhagic; nevertheless, pediatric patients
usually present with ischemia.
In ischemic moyamoya disease, the thera-
peutic effect of surgical revascularization
(extracranial-intracranial bypass) has been
well established. Bypass surgery can improve
the impaired cerebral hemodynamic state and
decrease the recurrent ischemic events. Man-
agement of the hemorrhagic moyamoya dis-
ease, on the other hand, presents a serious
challenge. Despite the extremely high rate of
rebleeding attacks, no therapeutic method has
been established until recently. In 2013, the
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis)
Japan Adult Moyamoya (JAM) Trial, a multi-
centered prospective randomized controlled
trial, was completed. It has been revealed that
bypass surgery significantly decreases the rate
of rebleeding attacks and improve the patients’
prognosis during the following 5 years. This
epoch-making study is expected to establish a
guiding principle for the treatment of hemor-
rhagic moyamoya disease.
Keywords
Moyamoya disease • MRI • MRA • SWI •
Perfusion • ASL • Direct bypass • STA-MCA
anastomosis • Indirect bypass • SPECT • PET •
Intracranial hemorrhage • JAM trial • Preg-
nancy • Delivery
Introduction
Moyamoya disease represents a unique cerebro-
vascular disorder characterized by steno-
occlusive change at the terminal portion of the
internal carotid artery (ICA) and a fine vascular
network, the so-called “moyamoya” vessels. This
fine vascular network forms a collateral pathway
and compensates for the reduced cerebral blood
flow due to the steno-occlusion occurring in the
major arteries. The disease shows a differential
incidence and predominantly affects Eastern
Asia. Its clinical features include ischemia and
hemorrhage, epilepsy, headache, etc. While
young patients usually present with ischemia,
adult patients present either ischemia or hemor-
rhage. Evidence suggests that revascularization
surgery is highly efficient at preventing an ische-
mic event; in addition, the latest report [2] sug-
gests its effect on prevention of hemorrhagic
events. Due to a familial occurrence of 10–15 %,
intense investigations were made towards the
identification of gene candidates, which led to
the identification of several genetic loci associated
with the disease. Nevertheless, its pathogenesis
remains largely unknown.
Recently, for the diagnosis and treatment of
moyamoya disease (spontaneous occlusion of
the circle of Willis) have been updated [3]. In
819
this chapter, state-of-the-art description, symp-
tom, epidemiology, diagnosis, and treatment of
moyamoya disease based on this latest guideline
will be discussed (Fig. 1).
Epidemiological Features
Geographical Distribution
East Asian countries feature the highest incidence
of moyamoya disease, more particularly Japan
and the Republic of Korea. A study conducted in
Japan reported the total annual number of
moyamoya disease patients at 3,900, with preva-
lence and annual incidence rates of 3.16 and 0.35
per 100,000, respectively, in 1994. The female-to-
male ratio was 1.8 and 10 % of the patients had a
family history of the disease. The distribution of
the age at onset has two peaks: a dominant peak at
5 years of age and a more modest peak around the
late 20s to 30 years of age [4]. A recent review of
regional differences shows that incidences per
100,000 patient-years ranged in Japan from 0.35
to 0.94 (95 % CI 0.69–1.19) and in the USA from
0.05 ( 0.04–0.12) in Iowa to 0.17 ( 0.06–0.40)
in Hawaii and were 0.41 (0.28–0.54) in Nanjing,
China, and 0.02 (0.003–0.04) in Taiwan. The
female-to-male ratio ranged from 1.1 (0.9–1.5)
in Nanjing to 2.8 (1.2–6.1) in Iowa. Proportions
with cerebral hemorrhage as the initial presenta-
tion were 56 % in China, 52 % in Taiwan, 29 % in
Hawaii, 21 % in Japan, and 10 % in Iowa. Patients
with childhood onset presented most often with
ischemia (more than 75 %) in all regions [5].
Gender and Age Distribution
MMD has two peak ages of onset, initially at
5 years of age (pediatric or juvenile type) and
subsequently at 30–50 years of age (adult type).
Peak age in male is 10–14 years and 35–49 years,
whereas peak in female is 20–24 years and 50–54
years [6]. In 2003, the total number of patients
treated in Japan was estimated at 7,700. The
female-to-male ratio was 1.8, and a family history
was found in 12.1 % of patients. The prevalence
820
Fig. 1 Conventional digital subtraction angiography of
the left internal carotid arteries of a pediatric moyamoya
disease patient. Anterior-posterior view (a–c) and lateral
rate was calculated at 6.03 per 100,000, and the
annual rate of newly diagnosed cases was 0.54 per
100,000. In 2009, an analysis of the regional
all-inclusive epidemiological data obtained in
Hokkaido, a major island of Japan counting a
population of 5.63 million [7], reported a preva-
lence of 10.5 patients per 100,000 and an annual
incidence of 0.94 per 100,000, both of which
greatly exceeded the results of the previous
A. Yamamoto et al.
view (d–f) show steno-occlusive change of the terminal
portion of the left ICA and moyamoya vessels in the
perforator artery of the basal ganglia
surveys. The female-to-male ratio was 2.18 and
the age of onset occurred in two peaks with a
different trend: the highest peak arose between
45 and 49 years, and the second occurred between
5 and 9 years. A familial history was observed in
15.4 % of patients. These epidemiological fea-
tures differ significantly from the data obtained
in previous studies. However, the higher detection
and prevalence reported may not reflect an actual
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis)
increase in the incidence of moyamoya disease,
but may rather be due to the increased availability
of noninvasive diagnostic tools, such as MRI/MR
angiography.
Moyamoya disease patients’ age distribution
shows characteristic bimodal pattern, pediatric
and adult patients. The precise mechanism for
this bimodal distribution has not been determined.
In the pediatric patients, the rapidly progressing
maturation of the brain may precipitate deficiency
of a balance with cerebral blood flow maturation,
and as a consequence, patients progress to
develop ischemic attack. However, still many
things are unknown about pediatric normal devel-
opment of cerebral blood flow. This lack of bal-
ance between maturation of the brain and
development of cerebral blood flow might not be
able to explain the full mechanism of the devel-
opment of ischemic attack in pediatric moyamoya
disease patients. In adult moyamoya disease
patients, hemorrhagic symptom can occur more
commonly than in pediatric patients. In adult
moyamoya patients, prevalence of ischemic and
hemorrhagic symptom is almost the same.
Another unsolved issue of moyamoya disease
is why the number of patients between these pedi-
atric and adult moyamoya patients is small. About
this issue, ischemic symptom which commonly
occurs in pediatric period may become unlikely
to occur because a balance between brain matura-
tion and cerebral blood flow will be achieved.
After getting over this dangerous pediatric period,
moyamoya disease patient may become symp-
tomatic due to a decrease of cerebral blood flow
by arteriosclerosis in the adult period.
Familial Moyamoya Disease
An epidemiological feature of moyamoya disease
is the high incidence of familial occurrence, which
accounts for about 15 % of patients [8]. In patients
with familial incidence, the ratio of women to men
was reported to equal 5.0, as opposed to only 1.6
in sporadic cases; the mean (SD) age of onset was
11.8 (11.7) years in familial cases vs. 30.0 (20.9)
years in sporadic cases. Out of eight parent-
offspring pairs, all were paired with the mother.
821
While parents presented the onset of symptoms at
the age of 22–36 (mean 30.7 [7.5] years), their
offspring presented symptoms much earlier,
within 5–11 years of age (mean 7.2 [2.7] years).
This phenomenon is called clinical anticipation.
Clinical anticipation was reported in triplet repeat
disease; however, triplet repeat has not been
reported in moyamoya disease patients. This indi-
cates a strong genetic association with a female
predominance in cases of familial moyamoya
disease [9].
Pathogenesis of Moyamoya Disease
Cytokines
Most studies about cytokines in moyamoya dis-
ease patients have been performed by the exami-
nation of cerebrospinal fluid (CSF) sampled from
the subarachnoid space during surgical therapy.
Angiogenetic cytokines such as basic fibroblast
growth factor (b-FGF), hepatocyte growth factor,
and transforming growth factor (TGF) showed
increased concentrations in the CSF as well as in
other surgical specimens such as the arterial wall
specimens and the meningeal specimens. Among
these cytokines, angiogenetic cytokine b-FGF has
been thought to be related with several findings in
moyamoya disease: steno-occlusive lesion in the
circle of Willis, the development of moyamoya
vessels and dilatation of the cortical small arteri-
oles, and angiogenesis after surgical indirect
synangiosis therapy. Reportedly, b-FGF induces
the proliferation of vascular endothelial cells,
which may precipitate the stenosis of the major
arteries. On the contrary, b-FGF has also an
angiogenetic and dilatation effect on the small
arteries, which may explain the development of
moyamoya vessels and the dilated pial arterioles
on the cortex in moyamoya disease [10].
The role of cytokine abnormality in moyamoya
disease has not yet been totally determined. Ele-
vation of cytokines in the CSF may simply reflect
the response of cytokines to hypo-oxygenation
status in moyamoya disease. In this scheme, ele-
vated concentrations of cytokines may simply be a
consequence of hypo-oxygenation but not a direct
822
pathogenesis of moyamoya disease. Validation
study for the elevation mechanism of cytokine
levels will be necessary [10].
Genetic Factors
In cases of familial moyamoya disease, the gene
loci 3p24-p26 and 8q23 have been identified
using genome-wide analysis, while loci 6q25
and 17q25 were determined using chromosomal
search.
Cases of unilateral moyamoya disease
progressing to bilateral moyamoya disease have
been well known, as well as progression of major
artery stenosis on the contralateral side to the
initial disease. It has thus been suspected that
major artery stenosis, unilateral moyamoya dis-
ease, and bilateral moyamoya disease were among
a spectrum of relational phenomena, based on
genetic susceptibility. In addition, familial
moyamoya disease has been known to be an auto-
somal dominantly inherited disease with incom-
plete penetrance, and various stages of the disease
are thus observed in familial moyamoya disease
patients. From these points, a combination of
internal genetic factors and external environmen-
tal factors has been believed to be relevant for
disease occurrence or progression. These mecha-
nisms of moyamoya disease have been reported to
be relevant with three heterogeneities, disease
heterogeneity, genetic heterogeneity, and locus
heterogeneity [11].
Recently, a genome-wide association study
identified ring finger protein (RNF)213 (http://
omim.org/entry/613768) as the first moyamoya
disease gene [12]. Also, another genome-wide
linkage analysis by assuming the inheritance pat-
tern of moyamoya disease as autosomal dominant
mode with incomplete penetrance and whole
genome-exome analysis provided evidence
suggesting the involvement of RNF213 in genetic
susceptibility to moyamoya disease [13]. Further
studies are ongoing to clarify the biochemical
function and pathological role of RNF213 in
moyamoya disease [14].
A. Yamamoto et al.
Asymptomatic Moyamoya Disease
In recent years, asymptomatic cases of moyamoya
disease and moyamoya disease manifesting with
nonspecific symptoms only, such as headache,
have been reported. “Asymptomatic” patients
with moyamoya disease have been defined as
those who have experienced neither ischemic
nor hemorrhagic episodes since birth. According
to previous reports, clues to the diagnosis of
asymptomatic moyamoya disease included
nonspecific symptoms such as tension-type head-
ache, dizziness, head trauma, etc. Some patients
were incidentally diagnosed using MRI/MRA
which was performed for brain checkup or scru-
tiny for the disease because of family history of
moyamoya disease [15]. The increase in the
number of asymptomatic patients could be attrib-
uted, at least partially, to the current increasing
availability of MRI/MRA examination in
Japan. In an all-inclusive survey of moyamoya
disease in Hokkaido, the authors reported that
asymptomatic patients comprised 17.8 % of the
267 newly registered patients with moyamoya
disease between 2002 and 2006 [7]. From this
observation, the prevalence of asymptomatic
moyamoya disease may be much higher than pre-
viously thought [16].
The reported imaging finding in asymptomatic
moyamoya disease patients has been limited in
number. Kuroda et al. reported cerebral infarction
in 16 out of 77 (20.8 %) involved hemispheres in
asymptomatic moyamoya disease patients
[15]. About 40 % of the involved hemispheres
had a moderate or severe reduction of cerebral
perfusion reserve in asymptomatic moyamoya
disease [15]. The annual risk for ischemic or hem-
orrhagic stroke was estimated at 3.2 %. Disease
progression was associated with ischemic events
or silent infarction in four of five patients. Also,
asymptomatic abnormal findings were identified
in another three patients [15].
Asymptomatic moyamoya disease may thus be
closely related to the development of asymptom-
atic cerebral infarction. This observation may be
consistent with the findings that steno-occlusive
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis)
Fig. 2 A case of asymptomatic adult moyamoya patient.
Conventional digital subtraction angiography lateral view
image shows stenosis of ICA and moyamoya vessels at the
frontal area and basal area. (a) Source image of TOF-MRA
arterial change in adult moyamoya disease
patients may progress in both the anterior and
posterior circulation, in both bilateral and unilat-
eral disease types, and in both symptomatic and
asymptomatic patients [17]. The natural course of
asymptomatic moyamoya disease is still not fully
understood [3]. Therefore, it is important to know
that moyamoya disease progression may occur
asymptomatically and can suddenly cause ische-
mic or hemorrhagic stroke even in asymptomatic
patients. It is currently unknown how and when
823
(b) and T2-WI image (c) shows both stenosis of ICA.
Previous T2-WI image (d) shows apparent flow void of
bilateral MCA compared to (c). MIP image of TOF-MRA
shows bilateral ICA stenosis apparently (e)
moyamoya disease deteriorates from an asymp-
tomatic to symptomatic. It might be feasible to
monitor asymptomatic moyamoya disease
patients by less invasive imaging methods such
as MRI/MRA for disease progression and ische-
mic and/or hemorrhagic stroke findings [16, 18].
Recently, in Japan, asymptomatic moyamoya
registry project (AMORE) has been started.
By achievements of this study, progression mech-
anism and the natural course of asymptomatic
moyamoya disease might be elucidated (Fig. 2).
824
Unilateral Moyamoya Disease
Based on the current diagnostic guideline [3],
unilateral moyamoya disease in adult patient is
diagnosed as probable moyamoya disease and
refers to the presence of unilateral steno-occlusive
change of the terminal portion of the internal
carotid arteries accompanied by the formation of
moyamoya vessels around that region. These uni-
lateral changes may occur concurrently with other
underlying diseases, such as hyperthyroidism,
intracranial arteriovenous malformation,
Down’s syndrome, Apert syndrome, von
Recklinghausen’s disease (neurofibromatosis
type 1), postirradiation therapy of the brain, sys-
temic lupus erythematosus, and Sjögren’s syn-
drome; in these situations, the patient’s condition
is classified as quasi-moyamoya disease and not
as unilateral moyamoya disease [3]. In children,
unilateral moyamoya disease associated with ste-
nosis of the terminal portion of the internal carotid
arteries should be considered as definitive
moyamoya disease and not as probable
moyamoya disease [19].
The frequency of unilateral moyamoya disease
was reported as 10.6 % within 2,635 patients with
moyamoya disease, including initially diagnosed
and re-diagnosed patients [3]. A family history is
occasionally present in patients with unilateral
moyamoya disease. An analysis of 15 families
having a family history of moyamoya disease in
3 or more generations revealed 5 patients with
Fig. 3 A case of unilateral moyamoya disease patient
shows right ICA stenosis on MRA. MIP image of
TOF-MRA shows right ICA stenosis (a). Moyamoya
A. Yamamoto et al.
concurrent unilateral moyamoya disease in addi-
tion to 43 patients with definitive moyamoya dis-
ease and suggested the possibility that the disease
was inherited by the same autosomal dominant
inheritance pattern. From this observation, unilat-
eral moyamoya disease with a family history can
also be considered as a subtype of moyamoya
disease. In addition, unilateral moyamoya disease
without a positive family history should be distin-
guished from definitive bilateral moyamoya
disease [20].
The symptoms of unilateral moyamoya disease
are the same as those of definitive bilateral
moyamoya disease. Ischemic symptoms, hemor-
rhagic symptoms, concurrent cerebral aneurysm,
involuntary movement, and headache may be
noted.
The reported frequency of progression from
unilateral to bilateral moyamoya disease varies
from 10 % to 39 % [21]. Each study cohort con-
sists of a limited number of patients; therefore,
progression rate from unilateral to bilateral dis-
ease may be influenced by the number of patients.
Progression to bilateral moyamoya disease may
not only affect pediatric patients but also adult
patients. The statistically significant risk factors
for progression to bilateral disease have been
reported to be the presence of equivocal or mild
stenotic changes in the terminal portion of the
ICA, the middle cerebral artery (MCA), or the
anterior cerebral artery (ACA) of the contralateral
side [22] (Fig. 3).
vessels are not prominent in the cistern both on T2-WI
image and TOF-MRA source image (b, c)
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis)
Quasi-moyamoya Disease
Quasi-moyamoya disease refers to the presence of
steno-occlusive change of the terminal portion of
the ICA accompanied by an abnormal vascular
network in association with an underlying dis-
ease. Even in cases with unilateral lesions, if an
underlying disease is present, the condition can be
considered as quasi-moyamoya disease. Unilat-
eral moyamoya disease in adult patient without
underlying disease should be considered as prob-
able moyamoya disease and should be differenti-
ated from quasi-moyamoya disease. Quasi-
moyamoya disease has been reported to affect
people of all races. Concurrent occurrence with
underlying congenital disease is more frequent in
children; however, concurrent occurrence with
acquired underlying disease is more frequent in
adults [3].
The following illnesses have been reported as
underlying diseases: atherosclerosis, autoimmune
disease (systemic lupus erythematosus,
antiphospholipid antibody syndrome, periarteritis
nodosa, and Sjogren’s syndrome), meningitis, von
Recklinghausen’s disease (neurofibromatosis type I),
brain tumors, Down’s syndrome, head injury, irra-
diation, hyperthyroidism, stenocephaly, Turner’s
syndrome, Alagille syndrome, Williams syn-
drome, Noonan’s syndrome, Marfan syndrome,
tuberous sclerosis, Hirschsprung’s disease, glyco-
gen storage disease type I, Prader-Willi syndrome,
Wilms tumor, primary oxalosis, sickle cell dis-
ease, Fanconi’s anemia, spherocytosis, eosino-
philic granuloma, type II plasminogen
deficiency, leptospirosis, pyruvate kinase defi-
ciency, protein S deficiency, protein C deficiency,
fibromuscular hyperplasia, osteogenesis
imperfecta, polycystic kidney, oral contracep-
tives, and poisoning (cocaine poisoning) [3].
Epidemiology of quasi-moyamoya disease was
reported in the annual report of the Research
Committee on Spontaneous Occlusion of the Cir-
cle of Willis (Moyamoya Disease) of the Ministry
of Health, Labour and Welfare, Japan 2011 (http://
mhlw-grants.niph.go.jp/niph/search/NIDD02.do?
resrchNum=201128175A). Prevalence of quasi-
moyamoya disease is reported as 5.4 % in a total
825
of 7,941 moyamoya disease patients and 0.34
patients per 100,000 populations. Annual inci-
dence of quasi-moyamoya disease is 0.11 patients
per 100,000 populations.
Symptom of quasi-moyamoya disease may be
epilepsy, headache, or asymptomatic. Concurrent
presence of symptoms associated with mental
retardation due to underlying disease and symp-
toms associated with cerebrovascular disorder
may cause a relatively complicated clinical con-
dition for the precise diagnosis. Cerebral conven-
tional digital subtraction angiography can show
various findings from those very similar to defin-
itive moyamoya disease to be rather different,
such as atherosclerotic lesions. Pathological find-
ings also vary according to the underlying disease.
Treatment of quasi-moyamoya disease is like that
of definitive moyamoya disease. The effect of
revascularization on the prevention of rebleeding
in patients with quasi-moyamoya disease has not
yet been clarified. In quasi-moyamoya disease,
unilateral disease may progress to bilateral dis-
ease. The nature of the underlying diseases influ-
ences the prognosis of patients with quasi-
moyamoya disease (Fig. 4).
Systemic Vessel Involvement
of Moyamoya Disease
In addition to intracranial arterial steno-occlusive
lesions associated with moyamoya disease, steno-
occlusive lesions of extracranial arteries including
renal, coronary, pulmonary, mesenteric, and
peripheral arteries have also been reported
[23]. Among these extracranial arteries,
moyamoya disease most commonly affects the
renal artery. The prevalence of renal artery
involvement in moyamoya disease patients ranges
from 5 % to 8 %. In most cases of moyamoya
disease, in which the renal artery was involved,
renal artery stenosis mainly involves the proximal
one-third of the main branch. Severe stenosis case
can precipitate renovascular hypertension.
Although the definitive cause and pathogenesis
of moyamoya disease remain unclear, pathologic
studies of the steno-occlusive lesions of the
826
Fig. 4 Images of a quasi-moyamoya disease (neurofibro-
matosis type I, von Recklinghausen disease) patient are
shown. MIP image of TOF-MRA (a) shows steno-
occlusive change of bilateral ICA. Source image of
TOF-MRA shows cisternal moyamoya vessels as high-
signal spots, which are also apparent on T2-WI image
A. Yamamoto et al.
flow voids in the cistern. (b, c) Conventional digital sub-
traction angiography images (d, e) show steno-occlusive
change and moyamoya vessels more apparently. On T2-WI
image, right optic nerve mass lesion and high-signal area in
the right cerebellar peduncle are apparent; those are char-
acteristic lesion of neurofibromatosis type I (f)
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis)
cerebral vessels typically reveal fibrous thicken-
ing of the intima with a small amount of lipid
deposition. Inflammatory cell infiltration is not
noted in the vascular walls, and the internal elastic
lamina is well preserved. Similar histological
findings of intimal fibrous thickening in the extra-
cranial vessels have been reported in autopsy
cases of moyamoya disease [23].
Pathological Features
Pathological evaluation of involved vessels in
moyamoya disease has shown that the outer diam-
eters of the terminal portions of the relevant ICA
are markedly diminished. In addition to outer
diameter change, fibrocellular thickening of the
intima, an irregular undulation (waving) of the
internal elastic lamina, and attenuation of the
media have been reported in a histological finding
in the vascular wall of the ICA [24, 25]. Data from
recent studies have suggested that caspase-3-
dependent apoptosis might be associated with
these histological changes in the vascular wall of
moyamoya disease patients [26].
Moyamoya vessels consist of dilated perforat-
ing arteries, which can show a wide range of
histological findings, including fibrin deposits in
the wall, fragmented elastic lamina, attenuated
media, and the formation of microaneurysms. In
the early stage of the disease (stage I according to
Suzuki’s angiographic classification), moyamoya
vessels are rarely observed. Steno-occlusive
change of the arterial lumen and subsequent
thrombosis can also be seen in the moyamoya
vessels [27]. Therefore, these histological changes
might have close association with the onset or
progression of ischemic or hemorrhagic stroke in
moyamoya disease.
Clinical Features
An Overview
The clinical features of moyamoya disease show
substantial difference between pediatric and adult
patients. Most pediatric patients with moyamoya
827
disease develop transient ischemic attack (TIA) or
cerebral infarction, whereas about half of adult
patients develop intracranial bleeding, and half
develop TIA or cerebral infarction or both
[28]. The symptoms and clinical course may
change according to the age and the disease type
manifested at the initial attack, and a wide range of
severity of symptoms have been noted, such as
TIAs and hemorrhagic stroke which resulted in
permanent neurological deficits.
According to the recent increase in the avail-
ability of MRI/MRA, increasing numbers of
patients have been reported, who were inciden-
tally diagnosed as having moyamoya disease dur-
ing the asymptomatic stage [15] or with only
minor complaint such as headache [3]. In pediatric
population, acute infantile hemiplegia is one of
the most important symptoms of moyamoya dis-
ease. Benign familial nocturnal alternating hemi-
plegia of childhood has been also reported to be a
possible symptom of moyamoya disease.
The frequency of each initial symptom in the
1,127 definitive moyamoya disease patients reg-
istered until 2000 is presented in Table 1 for the
patients with the hemorrhagic type and ischemic
type (infarction type, TIA type, and frequent TIA
type) of initial attack. For both types, muscle
weakness, consciousness disturbance, headache,
speech disorder, and sensory disturbance were the
most frequent; however, the incidence of con-
sciousness disturbance and headache was higher.
Incidence of muscle weakness was lower for
patients with the hemorrhagic-type than for the
ischemic-type initial attacks ( p < 0.01) [3].
Ischemic Symptoms
Moyamoya disease patients usually show cerebral
ischemic change in the anterior territory of the
ICA, particularly in the frontal lobe. Therefore,
most patients with symptomatic moyamoya dis-
ease may present with focal neurological deficits,
such as dysarthria, aphasia, or hemiparesis. How-
ever, moyamoya disease patients might also occur
with other relatively atypical symptoms such as
consciousness disturbance, sensory disturbance,
visual symptoms, or involuntary movements,
828 A. Yamamoto et al.

Table 1 Initial symptom of moyamoya disease (n = ventilation. Pediatric moyamoya disease patients
1,127) sometimes develop TIA when slow waves appear
Hemorrhagic Ischemic again during EEG scans. This reappearance of
Initial symptom type (%) type (%) slow waves is pathognomonic for pediatric
Muscle weakness 58.6 79.8a moyamoya disease patients and is known as
Consciousness 70.4a 14.1 rebuild-up phenomenon [32]. Recently, this
disturbance
rebuild-up phenomenon has been reported to be
Headache 64.6a 18.8
Seizure 8.5 8.0
induced by post-hyperventilation hypoxia com-
Psychiatric 8.7 2.5 bined with a hyperventilation-induced reduction
symptom in cerebral blood flow and originates in the deep
Speech disorder 24.5 20.1 cortical sulci, where the cerebral perfusion reserve
Sensory 18.4 19.3 is diminished [33]. Rebuild-up phenomenon can
disturbance completely disappear after effective surgical
Involuntary 3.3 3.0 revascularization [33] (Fig. 5).
movement
Intellectual 5.3 6.2
disturbance
Visual impairment 2.0 3.2 Hemorrhage
Visual field defect 3.9 5.0
Reprinted with permission from Guidelines for Diagnosis About a half of adult moyamoya disease patients
and Treatment of Moyamoya Disease (Spontaneous Occlu- develop intracranial hemorrhage, though hemor-
sion of the Circle of Willis). #The Japan Neurosurgical rhagic symptom is rare in pediatric moyamoya
Society [3] patients [24]. Also, hemorrhagic symptom has
a
Significantly more frequent as compared with the others
( p < 0.05) been known to be more common in female
patients than male patients. Hemorrhagic symp-
tom patients show more severe symptom and
particularly in children [29, 30]. Some pediatric deteriorated outcome than ischemic symptom
patients develop intellectual impairment owing to patients.
frontal lobe ischemia, infarction, or both [31]. In Two main and one additional causes of intra-
adult moyamoya disease patients, cognitive dys- cranial hemorrhage in moyamoya disease have
function, such as short-term memory disturbance, been assumed: (1) rupture of dilated, fragile
irritability, or agitation, has been reported. moyamoya vessels or (2) rupture of saccular aneu-
Patients who present with these atypical symp- rysms in the circle of Willis especially in the
toms may be misdiagnosed with psychiatric dis- posterior part. For the first cause, the rupture
orders, such as schizophrenia, depression, or may be due to persistent hemodynamic stress on
personality disorder. the moyamoya vessels and occurs mainly in the
In pediatric patients, ischemic attacks are typ- basal ganglia, thalamus, or periventricular region,
ically induced by hyperventilation, for example, which is common with basal moyamoya vessels.
when crying or playing with a pinwheel. Electro- Intraventricular hemorrhage is more frequent in
encephalogram (EEG) findings in pediatric moyamoya disease patients than other origin hem-
moyamoya patients are characteristic. In healthy orrhagic patients [34]. Peripheral aneurysms in
children, high-amplitude slow waves are induced the collateral vessels or moyamoya vessels might
during hyperventilation, which will disappear be identified on conventional digital subtraction
after hyperventilation stops (an event known as cerebral angiography [35].
build-up phenomenon). However, in pediatric For the second cause, rupture of saccular aneu-
moyamoya disease patients, the slow waves rysms located around the circle of Willis occurs
appear again after a recovery of normal most commonly at the basilar artery bifurcation or
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis)
Fig. 5 3-T MR images of a pediatric moyamoya patient
are shown. On T2-WI image (a), moyamoya vessels are
apparent in the bilateral basal ganglia. On the source image
of time-of-flight MRA (b), moyamoya vessels are apparent
at the junction of the basilar artery and the supe-
rior cerebellar artery. The vertebrobasilar system
has an important role in providing collateral cir-
culation for steno-occlusive change of bilateral
internal carotid arteries in moyamoya disease
patients. Consequently, hemodynamic stress
probably may precipitate the formation of a
saccular aneurysm predominantly in the posterior
part of the circle of Willis or vertebrobasilar
system. Rupture of a saccular aneurysm in the
vertebrobasilar system can cause subarachnoid
hemorrhage [36]. Therefore, in the imaging
diagnosis of moyamoya disease patients,
scrutiny for aneurysm especially in the
vertebrobasilar system is important. 3-T
MRI/MRA examination has superior signal-to-
noise ratio, higher spatial resolution, and less
invasiveness and, therefore, is suitable for clinical
routine follow-up examination for moyamoya dis-
ease patients [18].
There is increasing reports that adult
moyamoya disease patients might present sub-
arachnoid hemorrhage over the cerebral cortex
despite the absence of an intracranial aneurysm
[37]. A third cause of intracranial hemorrhage in
adult moyamoya disease patients is rupture of the
dilated collateral pial superficial arteries on the
829
in the bilateral basal ganglia and thalamus. FLAIR image
(c) shows tiny high-signal area in the bilateral occipital
lobe (FLAIR ivy sign) and chronic infarction
brain cortical surface, although this is rare [37]
(Figs. 6 and 7).
Headache
Recently, headache is also considered to be one of
the common clinical presentations of pediatric
moyamoya disease patients. In many clinical
cases, pediatric patients complaining headache
may not be considered as having moyamoya dis-
ease, and other origins of headache such as men-
ingitis or brain tumor will be looked for. Headache
can be seen in about 20–30 % of moyamoya
disease patients. Specific type of headache for
moyamoya disease patients has not been reported;
however, patients can show migraine in the more
severely affected side. Also, patients with alter-
nating hemiplegia of childhood showing migraine
and hemiplegia have been reported
[38]. Moyamoya disease patients may complain
of headache before and after revascularization
surgery. Preoperative headache in moyamoya dis-
ease patients was considered to be related to
hypoperfusion or hypo-oxygenation because it
has been reported that headaches disappeared
after revascularization surgery [39].
830
Fig. 6 In a case of patient presenting with headache,
unenhanced CT scan image shows subtle high-density
area in the left frontal lobe surface with obliteration of the
left frontal lobe sulci. (a) FLAIR (b) and SWI (c) images
show superficial high signal and low signal in the left
frontal lobe, respectively. TOF-MRA MIP image shows
A. Yamamoto et al.
stenosis of the left ICA and MCA (d). Arterial spin labeling
perfusion-weighted image shows decreased flow in the left
hemisphere. (e) Diagnosis of probable moyamoya disease
was confirmed by a conventional digital subtraction angi-
ography (f, g)
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis)
Fig. 7 Difference of appearance of hemorrhage among
different MRI sequences. T2-WI image (a), T2-star-
weighted image, and MiniP image of SWI of hemorrhage
in the right basal ganglia are shown. T2-star-weighted
image shows (b) hemorrhage as more pronounced (c)
low-signal area compared to T2-WI image. This difference
Involuntary Movement
Involuntary movements are relatively rare symp-
toms in moyamoya disease patients, and their
frequency is estimated about 5 % [40]. Association
of various involuntary movement disorders with
moyamoya disease has been reported in the liter-
ature. The spectrum of these involuntary move-
ments includes chorea, choreoathetosis,
dyskinesia, dystonia, limb shaking, and epilepsia
partialis continua. Some of them may occur in a
paroxysmal manner, while others can be triggered
by the initiation of some movement
(“kinesigenic” manner) [41], by some particular
form of exercise, or by various types of hyperven-
tilation maneuvers. The involuntary movements
are usually transient, ranging in duration from
several seconds to a few months, and rarely
constant [40].
The main mechanism of onset of these invol-
untary movements is thought to be cerebral
hypoperfusion induced by steno-occlusive change
of the internal carotid arteries in the circle of
Willis. Limb shaking is typically triggered by
diffuse hemispheric hypoperfusion caused by
severe stenosis of the ICA. Any activities induc-
ing hyperventilation, such as singing and crying,
may precipitate the involuntary movement,
because of induced cerebral hypoperfusion. This
831
comes from effect of 180 refocus pulse used in T2-WI
image. In addition, MiniP image of SWI shows hemor-
rhage as more pronounced low-signal area compared to
T2-star-weighted image. This difference comes from the
fact that SWI utilizes phase information to enhance the
local susceptibility difference
hypothesis is supported by the fact that revascu-
larization surgery is associated with improvement
of the involuntary movements or at least a signif-
icant decrease in their frequency [40].
Diagnostic Evaluation
Current Diagnostic Criteria
of Moyamoya Disease
1. Cerebral angiography is considered essential
for the diagnosis and must show at least the
following findings [3, 42]:
i. Stenosis or occlusion of the terminal por-
tion of the intracranial ICA or proximal
portions of the anterior and/or the MCA
ii. Abnormal vascular networks in the vicinity
of the occlusive or stenotic lesions in the
arterial phase
iii. Bilaterality of findings (i) and (ii)
2. However, when magnetic resonance imaging
(MRI) and magnetic resonance angiography
(MRA) findings meet all of the following
criteria, cerebral angiography can be omitted:
i. MRA shows stenosis or occlusion of the
terminal portion of the intracranial ICA or
proximal portions of the anterior and/or
the MCA.
832
ii. MRA shows abnormal vascular networks
in the basal ganglia.
Note: When two or more visible flow
voids are present in the basal ganglia on
MRI, at least unilaterally, they can be
deemed as representing an abnormal vas-
cular network.
iii. Bilaterality of findings (i) and (ii).
3. Moyamoya disease is an illness of unknown
etiology. The differential diagnosis of this
disease includes similar cerebrovascular
lesions associated with the following underly-
ing diseases, which should, therefore, be
excluded: (i) atherosclerosis, (ii) autoimmune
disease, (iii) meningitis, (iv) brain tumors,
(v) Down’s syndrome, (vi) von
Recklinghausen’s disease, (vii) head injury,
(viii) cerebrovascular lesions after head irradi-
ation, and (ix) others.
4. Pathological findings that can be used as refer-
ences for the diagnosis:
i. Thickening of the arterial intima, mainly
in the terminal portion of the internal carotid
arteries, and narrowing or blockage of the
lumen caused by this change, usually bilat-
eral. Occasionally, lipid deposits are also
present in the thickened intima.
ii. Arteries such as the anterior, middle, and
posterior cerebral arteries forming the cir-
cle of Willis occasionally show varying
degrees of stenosis or occlusion associated
with fibrocellular thickening of the intima,
waviness of the internal elastic lamina, and
thinning of the media.
iii. Numerous small vascular channels (perfo-
rating and anastomotic branches) can be
seen around the circle of Willis.
iv. Pia mater may also show reticular con-
glomerates of small vessels (Fig. 8).
Reprinted with permission from Guidelines
for Diagnosis and Treatment of Moyamoya
Disease (Spontaneous Occlusion of the
Circle of Willis). #The Japan Neurosurgical
Society [3].
A. Yamamoto et al.
Definitive Moyamoya Disease
and Probable Moyamoya Disease
Moyamoya disease should be classified as defin-
itive or probable based on the abovementioned
items (1) to (4) [3]. For definitive moyamoya
disease, all criteria listed in (1) or (2) and in
(3) should be fulfilled. In pediatric patients, how-
ever, the criteria in item (1) or (2) (i) and (ii), on
one side, and visible stenosis around the terminal
portion of the internal carotid arteries, on the other
side, are sufficient for a definitive moyamoya
disease diagnosis. For probable moyamoya dis-
ease, all criteria are fulfilled except item (1) (iii)
and/or item (2) (iii) among the criteria of (1) or
(2) and (3).
CT and CT Angiography in Moyamoya
Disease
CT scan is one of the primary survey modalities
for patients suspected of stroke; however, any
definitive diagnostic information of moyamoya
disease may not be obtained by this modality. In
advanced stages of moyamoya disease, chronic
cerebral infarction and unusual cerebral atrophy
can be seen on CT scan; however, definitive diag-
nosis cannot be done [43]. On unenhanced CT
scan, the steno-occlusive change and develop-
ment of moyamoya vessels are not apparent. For
the assessment of cerebral vessels, contrast-
enhanced CT scan and CT angiography is neces-
sary, and they can precisely show the steno-
occlusive change and development of moyamoya
vessels. During the diagnosis process of
moyamoya disease, findings of unenhanced CT
scan of stroke is not enough to exclude moyamoya
disease, and further examination such as
MRI/MRA or conventional angiography is
necessary [43].
Unenhanced CT scan is a reliable imaging
modality when the diagnosis of an ischemic or
hemorrhagic stroke is suspected. In adult
moyamoya disease patients, hemorrhage is
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis) 833

Fig. 8 In a pediatric moyamoya disease patient, right vertebral artery digital subtraction angiography images are shown.
Anterior-posterior view (a–c) and lateral view (d–f) show collateral moyamoya vessels from vertebrobasilar systems
834
common such as intraventricular hemorrhage,
cerebral hemorrhage, and subarachnoid hemor-
rhage due to rupture of moyamoya vessels or
aneurysms. In any cases of supratentorial cerebral
hemorrhage seen on unenhanced CT scan, hemor-
rhage due to moyamoya disease must be consid-
ered as a differential diagnosis, and further
examination such as contrast-enhanced CTA is
recommended [43].
The recent advancement of multi-detector row
computed tomography (MDCT) scanners has
enabled high-resolution three-dimensional recon-
struction. CTA and MRA were compared in terms
of the steno-occlusive changes exhibited in each
vessel. CTA and MRA scores were assigned on
the basis of the severity of occlusive changes in
the ICA, MCA, ACA, and posterior cerebral
artery (PCA). CTA scores were significantly cor-
related with MRA scores, and their scores were in
complete agreement in 39.6 %. The mean CTA
score was significantly lower than the mean MRA
score. CTA using MDCT is having a higher spatial
resolution, therefore, can be a more reliable
method than MRA for diagnosing moyamoya dis-
ease, particularly in emergency cases [44] (Figs. 9
and 10).
MRI/MRA Diagnosis of Moyamoya
Disease
In moyamoya disease patients, anterior circula-
tion (steno-occlusive change of the ICA and col-
lateral moyamoya vessels), posterior circulation
(steno-occlusive change of the PCA and
leptomeningeal anastomosis collateral vessels
from the posterior circulation), and transdural
anastomosis collateral vessels such as ethmoidal
moyamoya vessels and vault moyamoya vessels
should be evaluated to monitor the hemodynamic
state of moyamoya disease patients. Time-of-
flight (TOF) MRA method is more suitable than
phase contrast MRA or other MRA methods
because TOF-MRA can show the steno-occlusive
change of the major artery and development of
moyamoya vessels simultaneously and more pre-
cisely. MRI/MRA examination can provide the
significant diagnostic information less invasively
A. Yamamoto et al.
(without ionizing radiation exposure) compared
to conventional angiography.
Cerebral angiography is essential for a defini-
tive diagnosis of moyamoya disease [3]. MRI/
MRA can also be used for a definitive diagnosis
when the following findings are fulfilled on
TOF-MRA imaging conducted using a scanner
with a static magnetic field strength of more than
1.5-T (especially 3-T MR scanner is preferable):
1. On MRA, stenosis or occlusion of the terminal
portion of the intracranial ICA or proximal
portion of the anterior and/or middle cerebral
arteries
2. On MRA, abnormal vascular networks in the
basal ganglia
Note: When two or more visible flow voids in
the basal ganglia are present at least unilaterally
on MRI, they can be deemed as representing an
abnormal vascular network.
3. Bilaterality of findings (1) and (2)
Moyamoya disease stage classification has
also been reported based on the MR findings
(Table 2). In this classification system, the
stage is determined by simply assigning
scores to the MRA findings and then totaling the
scores. The stage classification using this
method corresponds well to the conventional clas-
sification based on angiography and has been
reported to show high sensitivity and
specificity [45].
FLAIR Ivy Sign
Fluid-attenuated inversion recovery (FLAIR)
image has been widely used for the diagnosis of
stroke and other cerebral abnormalities with a
high capability of lesion detection by nulling the
CSF signal with an employment of sophisticated
inversion pulse [46]. Characteristic imaging find-
ing on FLAIR has been reported in moyamoya
disease patients [46]. On FLAIR image,
moyamoya disease patients may show linear or
serpentine high signal in the sulci along the corti-
cal surface, which is believed to represent dilated
leptomeningeal collateral vessels [46], and this
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis) 835

Fig. 9 CT scan image of an adult moyamoya patient spots. (a, b) In the same patient, MRA (d) and T2-WI
shows chronic infarction as low-density area (c). images (e) show steno-occlusive change of bilateral ICA
Contrast-enhanced CT angiography images show and moyamoya vessels. Final diagnosis was confirmed by
moyamoya vessels in the cisternal area as tiny high-density conventional digital subtraction angiography (f, g)

finding is called “FLAIR ivy sign” because of its including moyamoya disease. This mechanism of
appearance resembling an ivy on the wall. This high signal on FLAIR image may at least partly
sign is frequently observed in distal areas with share with the mechanism of findings intra-arterial
decreased perfusion caused by vascular stenosis high signal on FLAIR ivy sign may partly share
836
Fig. 10 In a case of unilateral moyamoya disease, contrast-enhanced CT angiography image shows steno-occlusive
change of the right ICA and surrounding moyamoya vessels in the cistern (a, b)
the mechanism of intra-arterial high signal
observed in acute cerebral infarction on FLAIR
[47]. The mechanism of FLAIR ivy sign is also
applicable to the deep medullary vessel area in
moyamoya disease patients, and linear high signal
ivy sign was also reported in the deep white
matter [48].
FLAIR ivy sign can change its appearance
responding to the change of hemodynamic status
and, therefore, is useful for the evaluation of
hemodynamic status in moyamoya disease
patients. Its presence correlates highly with
reduced CVR indicating misery perfusion area,
and regional arteriocapillary circulation time was
elongated [49] but was decreased after bypass
surgery, correlated with improved hemodynamic
status observed by 123I-IMP-SPECT [50, 51]
(Figs. 11, 12, 13 and 14).
Feasibility of 3-T MRI/MRA
in Moyamoya Disease
3-T MRI scanner has already been widely used in
clinical hospitals and has been reported to produce
predominance compared to 1.0-T/1.5-T MR
A. Yamamoto et al.
scanner in the diagnosis of various neurological
disorders. The capability of higher spatial
resolution and signal-to-noise ratio of 3-T MRI
attributable to higher static magnetic field
strength is also valuable for evaluating fine imag-
ing findings in moyamoya disease patients.
Moyamoya vessels have been reported to be
better delineated with MRA at 3-T than at
1.5-T [52]. For evaluation of petechial hemor-
rhage in moyamoya disease patients, which rep-
resents an important factor for prognosis,
susceptibility weighted imaging (SWI) has been
reported to offer better detection at 3-T than 1.5-T
MRI [53]. In the routine clinical situation,
MRI/MRA findings also have been used for the
diagnosis of moyamoya disease, such as cisternal
moyamoya vessel flow voids and FLAIR ivy
sign [46].
Among these fine imaging findings of
moyamoya disease, the cisternal moyamoya ves-
sels visualized as a distinct feature of moyamoya
disease on 3-T MR might be beneficial for diag-
nosing moyamoya disease. Cisternal moyamoya
vessels show low-signal flow voids on T2-WI
image and high signal on TOF-MRA [18]. Com-
parison between existing MRI/MRA diagnosis
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis) 837

Table 2 Classification and scoring based on the magnetic radiation exposure) of MRI/MRA than conven-
resonance angiography (MRA) findings [45] tional angiography [18] (Figs. 15, 16, and 17).
MRA findings Score
Internal carotid Normal 0
artery Stenosis of C1 1 Susceptibility Weighted Imaging
Discontinuity of the C1 2 in Moyamoya Disease
signal
Invisible 3
Cerebral hemorrhage is one of the main symptoms
Middle cerebral Normal 0
artery of moyamoya disease especially in adult patients.
Stenosis of M1 1
Petechial hemorrhage in moyamoya disease
Discontinuity of the M1 2
signal patients is considered as an important factor for
Invisible 3 prognosis. On MRI, local inhomogeneous suscep-
Anterior cerebral Normal A2 and its distal 0 tibility effect due to hemorrhagic component such
artery A2 and its distal signal 1 as hemosiderin can become a diagnostic clue for
decrease cerebral hemorrhage. Such local susceptibility
Invisible 2 inhomogeneity effect is more apparent on gradient
Posterior cerebral Normal P2 and its distal 0 echo T2-WI image than T2-WI image. Also, this
artery P2 and its distal signal 1 susceptibility effect is more apparent with 3-T MR
decrease
scanner than with 1.0-T/1.5-T MR scanner. In
Invisible 2
addition, recent development of MRI technology
A total score of ICA to PCA is calculated individually for
both sides including 3-T MR scanner and image acquisition
sequence produced more sophisticated sequence
MRA total score MRA stage for the detection of hemorrhage; susceptibility
0–1 1 weighted imaging (SWI) has been introduced. In
2–4 2 SWI calculation of its final image from gradient
5–7 3 echo T2WI magnitude image, component of
8–10 4 phase mask information is utilized, which can
Reprinted with permission from Guidelines for Diagnosis enhance local susceptibility inhomogeneity effect
and Treatment of Moyamoya Disease (Spontaneous Occlu- more prominent than conventional T2WI image
sion of the Circle of Willis). #The Japan Neurosurgical
Society [3]. MRA stage 1 identified using the above [53]. T2WI image at 3-T found cerebral
approach corresponds to stages I and II of the angiographic microbleeds in 44 % of MMD patients compared
classification, stage 2 corresponds to stage III, stage 3 cor- with 5.8 % in control subjects [54]. They are
responds to stage IV, and stage 4 corresponds to stages V detected significantly more on SWI than T2WI
and VI
image at 3-T [53]. SWI can detect more hemor-
rhagic lesion in moyamoya disease patients than
conventional gradient echo T2WI images.
Not only for hemorrhage detection, SWI also
criteria and the proposed criteria using cisternal can show fine deep medullary veins apparently in
moyamoya vessels was performed. Diagnostic the deep white matter, which has never been visu-
accuracy was 62.5 % with the existing MR criteria alized on other image sequences such as T2WI
and 97.5 % with the proposed MR criteria. The image [55]. Evaluation of deep medullary veins in
proposed MR criteria were more sensitive than the moyamoya disease patients by SWI has been
existing MR criteria, but less specific than the reported, and symptomatic patients show more
existing MR criteria. By using this proposed prominent appearance of deep medullary veins
MRI/MRA criteria, diagnostic accuracy of than asymptomatic patients. Reportedly, promi-
moyamoya disease will be improved, and this nent deep medullary vein appearance correlated
criterion should be suitable for pediatric patients with less cerebral blood flow and cerebral vascu-
because of the less invasiveness (without ionizing lar reserve (Figs. 18 and 19).
838 A. Yamamoto et al.

Fig. 11 3-T MR images of a pediatric moyamoya patient FLAIR ivy sign. Also, collateral vessels in the deep white
are shown. On FLAIR image (b, c), dilated collateral matter are noted (b). Infarction was noted in the bilateral
vessels in the cortical surface are noted as so-called parietal cortical area

Fig. 12 3-T MR images of

an adult moyamoya disease
patient are shown. (a) MIP
image of TOF-MRA shows
bilateral stenosis of ICA and
moyamoya vessels. (b)
FLAIR image shows intra-
sulcal high-signal-intensity
area representing dilated
collateral vessels. This
finding is known as “FLAIR
ivy sign.” (c) Also, FLAIR
image at lower level shows
linear high-signal-intensity
area in the deep white
matter. This deep medullary
vessel high signal on
FLAIR also represents
dilated collateral vessels.
Infarction is apparent in the
same slice

modality due to its higher spatial resolution and

Staging by Conventional Angiography capability
in Moyamoya Disease (Table 3).
When the original concept of moyamoya disease
was introduced, digital subtraction angiography
(DSA) is still the most reliable diagnostic
of hemodynamics information
However, as mentioned in the above sections,
as far as the depiction of the steno-occlusive
change of the circle Willis is concerned, other
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis) 839

Fig. 13 FLAIR ivy sign

and cerebral perfusion. A
man in his 60s lost
consciousness when
driving. Stroke was
suspected, and MRI/MRA
was conducted. (a) MRA
shows severe stenosis at
bilateral ICA tops. There is
no apparent asymmetry.
(b–d) FLAIR images at
different levels show no
asymmetry at a glance.
However, a careful look at
the right Sylvian fissure and
frontal sulci shows
serpentine high signal,
called “FLAIR ivy sign.” It
is considered to indicate
maximally dilated and
slowly flowing arterial
branches. (e) IMP-SPECT
at rest shows frontoparietal
hypoperfusion bilaterally
without apparent
asymmetry. (f) IMP-SPECT
after acetazolamide or
vasodilator infusion shows
reduced CBF at the right
frontal lobe where “FLAIR
ivy sign” is observed. (g)
Cerebrovascular reserve
map (CVR) or percent CBF
increase map decreased
flow reserve at the right
frontal lobe. Other areas
show normal CBF increase
of more than 30 %,
represented as yellow to red
colors

modalities such as TOF-MRA and three- circulation is not always clear even with state-of-
dimensional CTA with best quality are quite com- the-art MRA/MRI due to temporal and spatial
patible with DSA. However, the depiction of the resolution and signal-to-noise ratio limitation. In
fine moyamoya vessels and the collateral addition, 3D-CTA essentially shows the
840
Fig. 14 FLAIR ivy sign pre- and post-bypass surgery. A
man in his 40s had headache and paresthesia. (a) MRA
shows bilateral ICA and right PCA stenosis. (b) Rest and
(c) post-acetazolamide infusion IMP-SPECT shows
morphological aspect of the vasculature that is not
necessarily identical to the true circulation of the
blood flow (lack of information of
hemodynamics).
In addition, the collateral circulations such as
basal moyamoya and transdural anastomosis of
the meningeal artery including the vault
moyamoya and ethmoidal moyamoya may not
be well demonstrated on other imaging modali-
ties. In addition, a microaneurysm associated with
moyamoya vessels, which is supposed to be the
cause of the intracerebral and intraventricular
hemorrhage, may not be well demonstrated on
other imaging modalities. For the diagnosis of
these small microaneurysms, conventional digital
A. Yamamoto et al.
reduced CBF at the right temporo-occipital area. FLAIR
ivy sign is observed at the corresponding area (d), which
disappeared after right STA-MCA bypass surgery (e)
subtraction angiography may be necessary
(Fig. 20).
Angiographical Staging Progression
of Moyamoya Disease
Angiographically, moyamoya disease progresses
along with the six stages introduced by Suzuki
[1, 56]. Firstly, stenosis appears in the terminal
portion of the ICA and the proximal portion of the
ACA and MCA bilaterally. Development of an
extensive collateral network at the base of the
brain along with the classic “puff of smoke”
appearance on conventional angiography is seen
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis)
Fig. 15 3-T MR images of a pediatric moyamoya patient
are shown. On T2-WI axial image (a), moyamoya vessels
in the sylvian valley are apparent as collection of tiny flow
voids. The source image of time-of-flight MR angiography
(b) shows moyamoya vessels as high signal intensities.
in the intermediate stage. “Basal moyamoya”
includes abnormal dilatation of the perforating
arteries, such as the lenticulostriate artery and
the thalamoperforating artery, in the basal ganglia
and thalamus. The steno-occlusive process also
involves the posterior circulation, including the
basilar and posterior cerebral arteries. Total occlu-
sion of the ICA terminal results in the disappear-
ance of “basal moyamoya,” and simultaneous
development of other collateral circulation is
observed in the advanced stage. “Ethmoidal
moyamoya” involves dilatation of the anterior
and posterior ethmoidal arteries, which also func-
tion as a collateral pathway, mainly from the
841
MIP image of 3D-TOF-MRA shows moyamoya vessels
and steno-occlusive change of bilateral terminal portion of
the internal carotid arteries. (d) Chronic infarction was
apparent on both T2-WI image (a) and FLAIR image (c)
ophthalmic artery to the ACA branches
[57]. Finally, the collateral flow from the dural
arteries to the pial arteries develops, and this path-
way is known as “vault moyamoya” [57]. Most of
the patients present with intermediate stages and
are treated promptly; therefore, the advanced
stage is nowadays rarely seen.
Aneurysm in Moyamoya Disease
In adult moyamoya disease patients, prevalence of
aneurysm has been reported as 14 % [58]. How-
ever, in pediatric moyamoya disease patients,
842 A. Yamamoto et al.

Fig. 16 3-T MR images of

an adult moyamoya disease
patient are shown. T2-WI
image (a), TOF-MRA
image (b), and MIP image
of TOF-MRA show
moyamoya vessels in the
cistern and bilateral ICA
stenosis. Also, chronic
infarction is apparent in the
left temporal lobe (a)

aneurysm is rare. Aneurysms in moyamoya dis-

ease can be categorized as (1) aneurysm occurring
in the circle of Willis and (2) small aneurysm
occurring in the peripheral or moyamoya vessels.
In the former, contrary to the common prevalence
of aneurysm, higher prevalence has been reported
in the posterior circulation especially
vertebrobasilar artery systems bifurcation. It has
been reported that aneurysm was rarely seen in the
MCA and ACA. In the latter, spontaneous regres-
sion or size decrease after revascularization sur-
gery has been reported; therefore,
pseudoaneurysm due to the rupture of moyamoya
vessels has been hypothesized as a cause.
Roles of Morphological
and Hemodynamic Functional Imaging
for MMD
Diagnosis
Conventional angiography is the gold standard for
the diagnosis of moyamoya disease (MMD)
[48, 59]. It clearly visualizes arterial stenosis and
moyamoya vessels (MMVs) and enables correct
evaluation, but there may be some difficulty when
it is conducted in children. Moreover, conven-
tional angiography may have potentially serious
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis)
Fig. 17 3-T MR images of
an adult moyamoya disease
patient are shown. (a) Time-
of-flight MRA source
image and (b) T2-weighted
image show cisternal
moyamoya vessels
apparently. (c) MIP image
of TOF-MRA shows
bilateral ICA stenosis
adverse effects [60]. Noninvasive imaging diag-
nosis is preferable, and angiography using CT and
MR (CTA and MRA, respectively) has been much
investigated for the diagnosis of MMD. Recent
progress in noninvasive imaging methods has
improved diagnostic capability. Guidelines for
the diagnosis and treatment of spontaneous occlu-
sion of the circle of Willis (i.e., MMD) from the
Ministry of Health and Welfare, Japan, have
included MRA as a diagnostic method
[3, 19]. Time-of-flight (TOF) MR angiography at
1.5-T could accurately depict stenosis around
80 % with some overestimation. Moyamoya ves-
sels (MMVs) were better visualized with MRI
than MRA, and MRI diagnosis had 92 % sensi-
tivity and 100 % specificity at 1.5-T [61]. CTA is
accurate and has high diagnostic capability
843
comparable to conventional method [62] and col-
laterals [63], but it still requires infusion of con-
trast material and radiation exposure is inevitable.
More recently, a 3-T MRI system is widely
available and has superiority to 1.5-T for the diag-
nosis of various neurological disorders. The
higher spatial resolution and signal-to-noise ratio
(SNR) at 3-T are useful for evaluating the fine
vascular changes of MMD. MMVs can be better
delineated with MRA at 3-T than at 1.5-T
[52]. Sawada et al. [18] focused on cisternal
MMVs compared with MMVs at the basal
ganglia, which is adopted in the existing MR
criteria, and reported that detection of cisternal
MMVs on T2-WI and MRA could attain 97.5 %
sensitivity with slightly deteriorated specificity of
95 % at 3-T. MMD is a progressive disorder, and
844
Fig. 18 (a) CT scan image of an adult moyamoya patient
shows low-density area in the right basal ganglia. (b) 3-T
MRI of T2-WI image shows low signal intensity in the
same area, indicating that there might be hemosiderin due
to hemorrhage. (c) T2WI image shows this hemorrhage as
serial evaluation is mandatory. Imaging measure-
ments had better be noninvasive and repeatable
with high precision, hopefully without radiation
exposure or contrast material administration.
Staging of Moyamoya Disease by
MRI/MRA
For this purpose as well, conventional angiogra-
phy has been the reference standard, but MRA
scoring has been proposed [45]. Scores were
assigned from 0 (normal) to 10 (most severe) for
changes from normal, stenosis, discontinuity, and
invisibility of the ICA, the horizontal portion of
the MCA, and the anterior and the posterior
A. Yamamoto et al.
low-signal-intensity area more apparently. Magnitude
image (d) and SWI (e) and MiniP image (f) are shown.
On SWI and MiniP image, hemorrhage is the most appar-
ent. Also, on SWI, deep medullary veins are apparent as
low-signal-intensity stripe
cerebral arteries. MRA scores were significantly
correlated with the conventional angiographic
staging (coefficient = 0.77). Four grades based
on the MRA scoring system correlated well with
Suzuki’s stages with high sensitivity and specific-
ity, and is a reliable alternative to conventional
staging, especially for children.
At the same time, we have to be aware
that disparity is sometimes observed between
angiography staging and clinical severity. Even
when bilateral ICAs are heavily stenotic or
occluded, ample collateral vessels may keep the
patient asymptomatic, whereas patients with
angiographically confirmed mild stenosis may
have TIA attacks. Simple angiographic
evaluation may not be sufficient to estimate future
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis)
Fig. 19 Difference of appearance of hemorrhage among
different MRI sequences. T2-WI image (a), T2WI image,
and MiniP image of SWI of hemorrhage in the left tempo-
ral lobe are shown. T2WI image shows hemorrhage as
more pronounced low-signal area compared to T2-WI
image (b). This difference comes from the effect of 180
Table 3 Suzuki’s stage classification of moyamoya dis-
ease by conventional angiography [3]
I Narrowing of the carotid fork
II Initiation of the moyamoya (dilated major cerebral
artery and a slight moyamoya vessel network)
III Intensification of the moyamoya (disappearance of
the middle and anterior cerebral arteries and thick
and distinct moyamoya vessels)
IV Minimization of the moyamoya (disappearance of
the PCA and narrowing of individual moyamoya
vessels)
V Reduction of the moyamoya (disappearance of all
the main cerebral arteries arising from the internal
carotid artery system, further minimization of the
moyamoya vessels, and an increase in the
collateral pathways from the external carotid artery
system)
VI Disappearance of the moyamoya (disappearance of
the moyamoya vessels, with cerebral blood flow
derived only from the external carotid artery and
the vertebrobasilar artery systems)
Reprinted with permission from Guidelines for Diagnosis
and Treatment of Moyamoya Disease (Spontaneous Occlu-
sion of the Circle of Willis). #The Japan Neurosurgical
Society [3]
morbidity of the MMD patients. Perfusion status
has to be examined more in detail including infor-
mation on the affected regions and local
hemodynamics.
845
refocus pulse used in T2-WI image. In addition, MiniP
image of SWI shows hemorrhage as more pronounced
low-signal area compared to T2WI image (c). This differ-
ence comes from the fact that SWI uses phase information
to enhance the local susceptibility difference
Altered Perfusion Status in MMD
The cerebral artery has an autoregulation mecha-
nism, and CBF stays constant for changes in cere-
bral perfusion pressure (CPP) within the range of
50–150 mmHg. When stenosis or occlusion of a
cerebral vessel around the circle of Willis reduces
CPP beyond the autoregulatory capacity, CBF
begins to decrease. When CPP reduction is not
so severe, CBV increases and oxygen extraction
fraction (OEF) increases. This compensatory
reaction maintains cerebral metabolic rate of oxy-
gen (CMRO2) and brain function. Further reduc-
tion of CPP, however, cannot be compensated and
may result in infarction.
There is another index of compensatory mech-
anism, called cerebrovascular reserve (CVR). It
reflects the capacity to further dilate vessels
from static condition, when CPP is decreased or
vasodilator such as acetazolamide (ACZ) or CO2
is administered. Vessels are fully dilated in
severely ischemic condition, and little
additional capacity of dilatation remains,
which is reflected as decreased CVR. In the
worst, “steal phenomenon” is observed where
flow is diverted to dilated vessels within a
846
Fig. 20 MIP image of TOF-MRA shows stenosis of ICA
(a). Conventional digital subtraction angiography (b, c)
shows same stenosis more precisely. In addition, conven-
tional DSA image shows finer moyamoya vessels in the
non-impaired region. In patients with unilateral
ICA or MCA occlusion, strokes recurred in
35 % and 6 % of cases with and without regional
CVR reduction for a follow-up period of
24 months [64]. CVR is an important predictor
of future morbidity.
CBF and CVR difference in age has to be taken
into considerations for evaluation of perfusion
status. Normal CBF is known to correlate nega-
tively with age. When young, CBF at rest shows
hyperperfusion at the frontal region, and CVR is
also very high. Both CBF and CVR decline with
aging. The images and measured values depend
on the method used for the measurement (Fig. 21).
Evaluation of Perfusion Status
There are several conditions of cerebral perfusion
associated with MMD. Nariai et al. [65] sorted
MMD patients out into those with transient
A. Yamamoto et al.
basal area. Ethmoidal moyamoya vessels in the frontal
skull base area are also apparent only on conventional
DSA image
ischemic attacks (TIA), infarction associated
with TIA (I/TIA), permanent deficit with infarc-
tion (PD), and non-symptomatic (NS) and com-
pared these types with normal controls using 15O-
labeled PET tracers. In NS, cerebral blood flow
(CBF) was preserved, but cerebral blood volume
(CBV) was increased without increase in oxygen
extraction fraction (OEF). In patients with TIA,
CBF was preserved, but CBV showed further
increment. If they got infarction (I/TIA), CBF
was largely decreased by around 10 ml/min/
100 ml and OEF was increased. When they got
PD, all of OEF, CBV, and CBF were decreased
(matched perfusion status).
MMD is graded based on digital subtraction
angiography (DSA), which has limited clinical
availability. Czabanka et al. [66] proposed a grad-
ing system included the following: for DSA, ste-
nosis/occlusion = 1 point, stenosis/occlusion +
intracranial compensation = 2 points, and steno-
sis/occlusion + intracranial compensation +
extracranial-intracranial compensation = 3 points
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis)
Fig. 21 MRA and cerebral perfusion. Pre- and postoper-
ative changes in TOF-MRA (left column) and IMP-SPECT
perfusion at rest (right column) by bypass surgery. (a)
Before bypass surgery. Collateral circulation from the
right PCA and MCA, resulting in less perfusion on the
right hemisphere. (b) After right STA-MCA bypass sur-
gery. The bypass is well visualized, and hypoperfusion is
and, for MRI, no sign of ischemia = 0 point, signs
of ischemia = 1 point, and CVR capacity > 5 %
= 0 points and < 5 % = 2 points. MMD grade I
referred to 1–2 points, grade II to 3–4, and grade
III to 5–6 points. In an analysis of 80 hemispheres
of MMD patients, they were graded as mild
(grade I, 18 %), moderate (grade II, 44 %), and
severe (grade III, 39 %) with clinical symptoms in
21 %, 63 %, and 93 %, respectively. This grading
system allows stratification for clinical symptoms
in MMD patients.
Mismatched perfusion condition has to be cor-
rectly evaluated. Until recently, PET (positron
emission tomography), SPECT (single-photon
emission tomography), and Xe-CT have been
used. CT perfusion and MR imaging including
DSC and ASL can now be used for cerebral
hemodynamic analysis [67]. Each has its advan-
tages and limitations, which should be well
understood.
847
improved. (c) After successive left STA-MCA bypass sur-
gery. Although right bypass is better visualized this time
indicating further improvement in perfusion on the right
MCA area, the left MCA area shows higher perfusion than
the other side. Note the relative flow decrease in the cere-
bellum caused by the cerebral flow increase
Oxygen-15 PET
This modality is considered most accurate and has
comprehensive ability to estimate multiple hemo-
dynamic parameters: CBF, CBV, OEF (oxygen
extraction fraction), and therefore CMRO2 (cere-
bral metabolic rate of oxygen). This modality has
been widely used to study patients with MMD
[65, 68, 69]. However, 15O-PET examination sys-
tem is not readily available, and the short half-life
time of 15O (2 min) requires a cyclotron that is
very expensive.
SPECT
Availability of SPECT is much higher than PET.
However, quantitation of CBF is affected by scan
condition as well as the tracer used for measure-
ment, and spatial resolution is much lower than
848
PET. In spite of these limitations, MMD has been
much studied with SPECT [70–72]. Additionally
to rest CBF, CVR (cerebrovascular reserve) can
be measured by using acetazolamide (ACZ,
17 mg/kg) to dilate the cerebrovasculature.
Three tracers are currently used for brain
123
perfusion SPECT: I-N-isopropyl-p-
99m
iodoamphetamine (IMP), Tc-ethyl cysteinate
dimer (ECD), and 99mTc-hexamethyl propylene
amine oxime (HMPAO). Among them,
IMP is used for quantitative scan, called
“QSPECT” [73–75]. In this scan protocol, rest
and ACZ-loaded scans can be successively
conducted. Rest scan begins at the onset of IMP
infusion that lasts for a minute. After around
10 min, an arterial blood sample is collected to
calibrate arterial input function. ACZ is infused at
20 min, and scan lasts to 28 min later. After 2 min
intermission, i.e., at 30 min, a second scan is
initiated by infusing another dose of IMP and
lasts to 58 min. Quantitation is not conducted in
a pediatric case.
CT Perfusion
Xenon-enhanced CT was much used in the former
time, but more recently, perfusion CT is
conducted by dynamic intravenous infusion of
iodine contrast material using a first-pass
tracer model. With the advent of multi-detector
row CT that does not require table movement
to cover the central part or the entire brain,
this method is gaining popularity and has been
used in MMD cases [76–78]. Higher availability
of MDCT scanners than RI scanners is advanta-
geous. By using arterial input function (AIF)
and venous output function (VOF),
quantitative measurement of CBF, CBV, BAT
(bolus arrival time), or MTT (mean transit
time) can be conducted [79]. However,
cautions have to be paid, because the results
can be affected by the analysis methods [80].
Radiation exposure is the most critical issue,
but modern dose-reduction methods make
this method more attractive. CT perfusion
imaging and analysis of the whole brain
can now be conducted by introduction of
A. Yamamoto et al.
multi-detector row CT that can cover the whole
brain [81].
MR Perfusion
As in CT, MR perfusion can be conducted
using gadolinium contrast agent, called
dynamic susceptibility contrast (DSC) MR perfu-
sion imaging [67], where first-pass signal
change is analyzed and semiquantitative results
can be obtained. The area size of signal
change shows CBV, and peak height shows
CBF, and MTT is calculated as CBV/CBF. As in
CT, quantitation is conducted using AIF (and
VOF). Signal intensity after contrast administra-
tion is logarithmically proportional to about
5 mMol/L at 1.5-T, but this linear relationship is
not true when the concentration becomes higher
[82]. There are several models and
implementations for this analysis, and results
may be different with frequent effects of bolus
delay [80]. Although the administered amount is
much smaller than iodine CT contrast material,
free form of gadolinium is highly toxic and rarely
causes fatal event in patients with renal insuffi-
ciency called “nephrogenic systemic fibrosis” or
NSF [83].
Arterial Spin Labeling
There is another MR perfusion method called
“arterial spin labeling” or ASL [84, 85]. It sup-
presses the signal from stationary tissues and
detects blood signal that flows into the brain.
The blood is magnetically labeled upstream at
the skull base and used as an endogenous contrast
material, and no contrast material administration
is required. For better background signal suppres-
sion, brain images are acquired as a pair of images
with and without labeling, and subtraction is
conducted. Due to this process, motion causes a
large misregistration artifact. Recently, however,
ASL imaging can be conducted in as short as
1 min on a 3-T MRI system (Fig. 22). Because
there is no need to use exogenous contrast mate-
rial, ASL measurement can be conducted
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis)
Fig. 22 Arterial spin labeling (ASL) for perfusion mea-
surement. (a) Severe stenosis is observed at bilateral MCA
origin. (b) IMP-SPECT after acetazolamide infusion
shows reduced CBF at the left MCA area, which is hard
to infer from the MRA. There is another method to detect
repeatedly [86], which is beneficial especially for
the children. Labeling can also be conducted to a
specific artery. Saida et al. applied this method to a
bypassed artery and found relatively low flow in
the early postoperative period, which later
increased [87]. Post-bypass perfusion can also be
hyperemic in the early period, and noninvasive
ASL can be applied to multiphase observations.
CBF measured by ASL had relatively weak
correlation with that by rest IMP-SPECT. How-
ever, it was much higher with ACZ IMP-SPECT
and CVR, suggesting that ASL could reflect the
currently and potentially dangerous zones for
ischemia [88]. Determining the presence and ade-
quacy of collateral blood flow is important in
MMD. ASL can also identify collaterals in
MMD patients compared with DSA in high sen-
sitivity and specificity over 0.8 [89]. ASL is use-
ful, but one drawback is that ASL is highly
sensitive to delay in arterial arrival time. Labeled
blood may not reach the brain parenchyma and
stay within the proximal artery. It looks as if flow
is increased on the more severely impaired side.
Caution has to be paid on the serpentine high
signal within the artery (Fig. 23).
849
reduced perfusion called “arterial spin labeling” or ASL
(c). In this case, arterial signal is prominent on the left
MCA area as high signal surrounded by low signal,
which means that arterial flow is very slow on the
corresponding area
CBF Changes After Bypass Surgery
Quantitative hemodynamics enables better selec-
tion of patients for revascularization surgery and
also postoperative evaluation of the clinical
outcome. Cerebral arterial bypass surgery
improves CBF and CVR, and the latter is more
prominent [90]. It also decreases CBV and OEF
[68, 72]. Early increase in CBF after revasculari-
zation can be detected as an asymmetry ratio of
the operative side compared with the other side
using IMP-SPECT and ASL [91]. It can be symp-
tomatic hyperperfusion, which is a potential com-
plication of bypass surgery. In a review of 82 adult
MMD patients with 99 surgeries, 39 patients
(47 sides, 47 %) suffered from transient neurolog-
ical deterioration due to hyperperfusion from 1 to
9 days after operation (median 2 days), which was
sustained for 1–14 days (median 7 days) [92,
93]. If symptom is severe, MR scans with
T2WI or susceptibility weighted imaging (SWI)
would be the primary choice to exclude subarach-
noid hemorrhage, and addition of an ASL scan
would be helpful.
850 A. Yamamoto et al.

Fig. 23 ASL dependency

on post-labeling delay time
(PLDT). (a) An MMD
patient after bilateral
STA-MCA bypass surgery.
(b) ASL with PLDT of
1,800 ms. Intra-arterial
signal is prominent, and
severe CBF reduction can
be assumed. However, (c)
CVR measured by
IMP-SPECT shows no
apparent reduction. (d) ASL
with PLDT of 2,600 ms
suggests much better
perfusion status. ASL is
very easy to use, but caution
has to be paid to transit
artifact caused by prolonged
arterial transit time

In a retrospective review of 73 MMD children

who had surgical revascularization due to strokes
or TIAs, three children (4.1 %) had recurrent
strokes, and TIAs were improved in about 80 %
for follow-up periods of 34 (5–166) months
[94]. Pediatric MMD patients at 6–12 months
after indirect bypass operation, preserved CVR
observed by SPECT, had better outcome than
those with decreased CVR resulting in remaining
neurological deficit and ischemic attacks on
follow-up [71]. However, it is not necessarily
matched with clinical outcomes. When CVR
improvement is not observed, close monitoring
would be required especially when patients have
worsening symptoms.
Visualization of Lenticulostriate
Artery (LSA)
Dilatation of the LSA is visualized as moyamoya
vessels. Occlusion of this small artery causes
one-third of all strokes and infrequent hemorrhage
due to aneurysm formation [95]. For LSA
visualization, CT angiography can be used. It
allows subtle changes such as diminished visual-
ization of the LSA [96]. In MRA, TOF-MRA can
visualize some LSA branches. It visualizes LSA
as “white-blood”; there is another option of LSA
visualization as “black-blood.” One of such
method is called flow-sensitive black-blood
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis) 851

Fig. 24 TOF-MRA
vs. flow-sensitive black-
blood (FSBB) MRA. A case
of unilateral MMD. (a)
TOF-MRA shows severe
left MCA stenosis. (b)
FSBB-MRA at 1.5-T shows
focal stenosis with distal
patency. Note that the
lenticulostriate artery is
more prominent on the
left side

(FSBB) MRA, which can visualize more of the

LSA branches than TOF-MRA [97]. This
relatively new method found reduced visualiza-
tion of LSA branches in lacunar stroke
patients [98].
Another interesting option is a hybrid of
opposite-contrast (HOP) MRA that combines
TOF and FSBB contrasts using a dual-echo
three-dimensional gradient echo sequence
[99]. Black-blood signal is subtracted
from white-blood signal making better
depiction of arteries and suppression of the back-
ground. In cases with major trunk,
HOP-MRA better visualized branches distal to
steno-occlusion of the major trunk artery, which
was considered to represent collaterals and
would contribute much as well in MMD [100]
(Figs. 24 and 25).
Treatment of Moyamoya Disease
Treatment for Ischemic Moyamoya
Disease
Surgical Treatment
Indication
Surgical revascularization improves hemody-
namic impairment and reduces ischemic attacks
in patients with ischemic moyamoya disease
[101]. Therefore, indications of surgical treatment
should be first examined in the patients
manifesting with cerebral ischemic attacks. Chil-
dren and adults with moyamoya disease who suf-
fer from cerebral infarction or frequent TIAs
caused by definite hemodynamic impairment are
good candidates for surgery.
852
Fig. 25 FSBB-MRA of a
normal subject at 3-T. The
lenticulostriate artery is
much better depicted at 3-T
than at 1.5-T
Single-photon emission computed tomogra-
phy (SPECT) is widely used to measure regional
cerebral blood flow (rCBF) and detect hemody-
namic failure. SPECT studies obtained after intra-
venous administration of acetazolamide can
evaluate cerebrovascular reserve capacity which
is an excellent indicator of hemodynamic failure
[102]. Positron emission computer tomography
(PET) using 15O tracer can evaluate rCBF,
regional cerebral blood volume (rCBV), regional
cerebral metabolic rate of oxygen (rCMRO2), and
regional oxygen extraction fraction (rOEF). At
present, PET study is thought to be a gold standard
to detect cerebral hemodynamic failure.
Selection and Application of Surgical
Revascularization Procedures
Since the 1970s, numerous surgical techniques
have been proposed, which can be divided into
two categories: direct anastomotic bypass
and indirect bypass. Active debates on the selec-
tion of these bypass procedures were once com-
mon in the previous century, but a consensus on
this issue has recently emerged, as clearly stated in
the new guidelines published in 2009
(in Japanese) [103].
1. Pediatric cases: Either one of the following
approaches is effective at improving hemody-
namic impairment: direct bypass procedures or
indirect bypass procedures.
A. Yamamoto et al.
2. Adult cases: Either one of the following
approaches is effective: direct bypass proce-
dures or a combination of procedures including
direct bypass. Indirect bypass procedures only
are not recommended because of inferior
effectiveness.
Direct Bypass
Direct bypass is a procedure for directly
establishing the blood flow pathway from extra-
cranial arteries to intracranial cortical arteries
using the microvascular anastomotic technique
(Fig. 26). Superficial temporal artery-to-middle
cerebral artery (STA-MCA) anastomosis is the
most popular and has been proven to be effective
in improving cerebral circulation in moyamoya
disease [104, 105]. It can be combined with vari-
ous indirect procedures. Direct bypass provides an
immediate improvement of hemodynamic failure,
but it requires sophisticated skills, because the
walls of the recipient cortical arteries are
extremely fragile in moyamoya disease.
Indirect Bypass
Principally, indirect methods are based on the
idea that neovascularization can be induced
from the extracranial arteries to the cortical
arteries by putting the vascular-rich tissues.
A variety of indirect revascularization procedures
– such as encephalo-myo-synangiosis (EMS),
encephalo-galeo-synangiosis (EGS),
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis)
Fig. 26 Direct bypass procedures for a 3-year-old female
with moyamoya disease. (a) Skin incision and line of the
craniotomy. (b) STA is dissected from the inner surface of
the skin flap. (c) Brain surface is exposed by opening the
encephaloduroarteriosynangiosis (EDAS), and
omentum transplantation – have been described
[106–109]. Although indirect revascularization is
technically easy, the success of this strategy
depends on the natural neovascularization capa-
bility of the patient’s brain; in some cases, post-
operative collateral formation is insufficient and
the progression of cerebral ischemia cannot be
halted [105, 110]. It is well known that neovascu-
larization tends to be insufficient in adult patients.
Perioperative Management
Regardless of the kinds of the bypass procedure,
perioperative management is extremely impor-
tant. Throughout the period, adequate intravenous
drip is mandatory to avoid dehydration. Because
intraoperative hypercapnia, hypocapnia, and
hypotension can increase the risk of ischemic
complications, anesthesiologists are required to
strictly control PaCO2 levels within the normal
range and to avoid systemic hypotension during
853
dura mater. (d) Donor STA and recipient MCA. (e)
Microanastomosis is completed. (f) Confirmation of the
bypass flow (white arrows) using fluorescent dye
surgery. Repeated radiological studies are
recommended during the postoperative acute
period to detect subdural or epidural hematomas
because most patients with ischemic moyamoya
disease are under the effect of preoperative
antiplatelet drugs.
Transient neurological events mimicking TIAs
are frequently observed after direct bypass pro-
cedures with no evidence of infarction apparent in
MR imaging. Although their pathophysiology has
long remained a mystery, recent postoperative
SPECT studies have revealed that these events
can be caused by local hyperperfusion
[111]. Because management of hyperperfusion is
completely different from that of hypoperfusion
(ischemia), the postoperative routine CBF study is
essential (Figs. 27 and 28).
Medical Treatment
Patients only with mild hemodynamic failure
can be treated conservatively. However, it must
854 A. Yamamoto et al.

Fig. 27 Effect of STA-MCA bypass in ischemic
moyamoya disease. (a, b) The anteroposterior view (a)
and the lateral view (b) of the left external carotid angio-
gram obtained after left STA-MCA bypass. The blood flow
via the STA (black arrows) covers the wide area of the
MCA territory. (c) Preoperative 123I-IMP-SPECT showing
the reduced CBF in the bilateral frontal lobe. (d) Postop-
erative 123I-IMP-SPECT showing the increased CBF in the
left frontal lobe

Fig. 28 Postoperative hyperperfusion. (a) Intraoperative confirmation of the bypass flow (yellow arrows). (b) 123I-IMP-
SPECT on postoperative day 7 showing local hyperperfusion (white arrows)
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis)
Fig. 29 Intracranial bleeding patterns of hemorrhagic
moyamoya disease. (a) Thalamic hemorrhage. (b)
Putaminal hemorrhage with ventricular perforation. (c)
be remembered that moyamoya disease is a pro-
gressive disease and the hemodynamic failure can
deteriorate later. When medical treatment is
selected, oral administration of aspirin or
clopidogrel is recommended for both adults
and children. Care must be taken, however, to
avoid excessively long-term treatment with
antiplatelet drugs in adults because it can
increase the risk of intracranial bleeding. Hemor-
rhagic transformation of adult ischemic
moyamoya disease is not uncommon. Therefore,
these drugs should be discontinued when the
patients are free from ischemic attacks for
several years.
855
Intraventricular hemorrhage. (d) Subcortical hemorrhage.
(e) Subarachnoid hemorrhage
Treatment for Hemorrhagic Moyamoya
Disease
More than one-half of all adult patients with
moyamoya disease present with intracranial hem-
orrhage [112]. Such bleeding attacks, which are
potentially fatal, seriously affect the patient’s
prognosis [113]. Typically, the hemorrhage occurs
in the thalamus and basal ganglia and frequently
involves perforation to the lateral and third ven-
tricles. In rare instances, subcortical or subarach-
noid hemorrhage can also be observed (Fig. 29).
It is speculated that chronic hemodynamic
stress might induce pathological change in the
856
collateral moyamoya vessels and the following
phenomena cause bleeding attacks [114]:
1. Rupture of fragile and maximally dilated
moyamoya vessels
2. Rupture of microaneurysms formed in the
moyamoya vessels
Management of the Acute Period
The patient should be treated with special consid-
eration for the pathophysiology of moyamoya
disease. Although blood pressure should be
strictly controlled to prevent rebleeding during
the acute period, excessive reduction of blood
pressure and dehydration must be avoided
because hemodynamic failure can coexist. Acute
hydrocephalus should be treated by emergent ven-
tricular drainage to reduce the intracranial pres-
sure and maintain adequate cerebral perfusion
pressure. Massive intracerebral bleeding requires
removal of the hematoma through a craniotomy.
Ruptured microaneurysms in tiny collateral ves-
sels are usually treated conservatively because
direct excision and intravascular treatment are
both difficult.
Prevention of Rebleeding in the
Chronic Stage
Management of the hemorrhagic type of
moyamoya disease presents a serious challenge
Fig. 30 Reduction of moyamoya vessels after STA-MCA
bypass in adult patients (ischemic type). (a) A left internal
carotid angiogram obtained before surgery. The
moyamoya vessels are remarkably well developed. (b) A
left internal carotid angiogram obtained after STA-MCA
A. Yamamoto et al.
because the rate of rebleeding attacks is extremely
high [115, 116]. A survey by Nishimoto
et al. revealed that 33 % of 175 patients with
hemorrhagic moyamoya disease experienced a
rebleeding attack [115]. Moreover, Kobayashi
et al. reported the annual rebleeding rate to be
7.09 % [116].
Although these hemorrhagic attacks cause seri-
ous permanent neurological deficits and the mas-
sive bleeding can be life-threatening, no
therapeutic method for preventing rebleeding
attacks has been established. No evidence exists
that treatment of hypertension reduces the
rebleeding rate in hemorrhagic moyamoya disease.
At present, bypass surgery is the only promis-
ing strategy. It is well known that, in ischemic
moyamoya disease, reductions in moyamoya ves-
sels can often be detected by angiography after
direct bypass surgery (Fig. 30) [117]. Because tiny
moyamoya vessels are the source of the bleeding,
the rate of hemorrhagic attacks can possibly be
decreased by reducing this hemodynamic stress
and consequently reducing the size and number of
moyamoya vessels. As a result, a hypothesis has
emerged that direct anastomotic bypass surgery
prevents future rebleeding; in fact, some authors
have reported the effectiveness of direct bypass
for hemorrhagic moyamoya disease [118]. On the
other hand, some reports indicate no significant
reduction in the rebleeding rate following
bypass. The reduced size of the moyamoya vessels is
evident. (c) A left external carotid angiogram obtained
after STA-MCA bypass. The blood flow via the STA
(arrow) covers approximately two-thirds of the MCA ter-
ritory (red arrowheads)
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis)
revascularization surgery [113, 119]; therefore,
surgical treatment of adult hemorrhagic
moyamoya disease remains controversial.
To resolve these issues, the Japan Adult
Moyamoya (JAM) Trial was undertaken in Japan
in 2001 [2, 120]. This multicentered prospective
randomized controlled trial seeks to determine
whether “direct” bypass surgery such as
STA-MCA anastomosis affects the prognosis
and incidence of recurrent bleeding attacks.
Patients who fulfilled all the clinical and radiolog-
ical requirements underwent a computer-
generated randomization scheme and were
assigned to receive either the best medical care
to modify risk factors or the best medical care plus
extracranial-intracranial direct bypass. Eighty
patients were enrolled in the trial, and a 5-year
follow-up of all the patients was completed in
2013. Kaplan-Meier survival analysis revealed
that direct bypass surgery significantly decreased
the rate of both all adverse events and rebleeding
attacks during the following 5 years [2]. These
results strongly suggest that the newly established
bypass flow can change the hemodynamic state of
the collateral vessels and lessen their overstress.
While subanalyses of the JAM Trial are now
underway, this epoch-making study will establish
a guiding principle for treatment of hemorrhagic
moyamoya disease.
Special Topics Moyamoya Disease and
Pregnancy
It is not uncommon for such patients to become
pregnant and give birth. In general, pregnancy is
known to increase the risk of cerebrovascular dis-
ease [121]. No evidence exists that the risk of
stroke increases during pregnancy in patients with
moyamoya disease; however, there have been spo-
radic case reports and reviews of pregnant
moyamoya patients presenting with stroke. At pre-
sent, no guidelines have been issued for managing
pregnancy in patients diagnosed with moyamoya
disease, and even recommendations regarding
delivery method (natural labor, painless labor
with spinal epidural anesthesia, or caesarian sec-
tion) seem to differ from one hospital to the next.
In 2008, the first nationwide survey on the
management of pregnancy and delivery in
857
women with moyamoya disease was conducted
in Japan [122]. This survey targeted the experi-
ences of 270 perinatal medical centers during the
5 years. It cited 64 deliveries comprising 59 cases
of previously diagnosed moyamoya disease and
5 cases of moyamoya disease newly diagnosed as
a result of perinatal stroke. The results of this
survey indicate that the incidence of perinatal
stroke is low in pregnant women previously diag-
nosed with moyamoya disease. Bypass surgery
had been performed on 58 % of the patients before
pregnancy. Only one case of cerebral hemorrhage
during pregnancy was reported in which the
maternal prognosis was poor. Although caesarian
section was mainly employed for women
previously diagnosed with moyamoya disease
(76 %), no attacks were observed during either
caesarian sections or vaginal deliveries;
therefore, no evidence exists that vaginal delivery
should be avoided. On the other hand, serious
cerebrovascular events (three cases of
cerebral hemorrhage and two cases of cerebral
ischemia) occurred in patients who had not been
diagnosed with moyamoya disease before preg-
nancy, and one patient with intracerebral
hemorrhage died.
Although most pregnant patients known to
have moyamoya disease can deliver safely, the
risk of pregnancy and delivery in women with
moyamoya disease should not be underestimated.
Serious cerebrovascular events occur in patients
with moyamoya disease undiagnosed before preg-
nancy. Poor prognosis is mostly likely to result
from cerebral hemorrhage. All physicians have an
obligation to provide accurate information and
ensure appropriate management of patients with
moyamoya disease. Rapid formulation of guide-
lines on the management of pregnancy and deliv-
ery in women with moyamoya disease is
mandatory.
Summary
1. Moyamoya disease has an unknown origin and
shows steno-occlusive change of the terminal
portion of the ICA with development of
moyamoya vessels.
858
2. Diagnosis of moyamoya disease should be
done with conventional angiography; how-
ever, MRI/MRA can become a substitute for
it in some situations.
3. Bypass surgery is the essential strategy for
ischemic moyamoya disease.
4. Recent randomized controlled trial has
revealed the effectiveness of “direct” bypass
surgery.
5. Although most pregnant patients known to
have moyamoya disease can deliver safely,
the risk of pregnancy and delivery with
moyamoya disease should not be
underestimated.
References
1. Suzuki J, Takaku A (1969) Cerebrovascular
“moyamoya” disease. Disease showing abnormal
net-like vessels in base of brain. Arch Neurol 20
(3):288–299, PubMed PMID: 5775283
2. Miyamoto S, Yoshimoto T, Hashimoto N, Okada Y,
Tsuji I, Tominaga T et al (2014) Effects of
extracranial-intracranial bypass for patients with hem-
orrhagic moyamoya disease: results of the Japan
Adult Moyamoya Trial. Stroke 45(5):1415–1421,
PubMed PMID: 24668203. Eng
3. Research Committee on the Pathology and Treatment
of Spontaneous Occlusion of the Circle of Willis,
Health Labour Sciences Research Grant for Research
on Measures for Infractable Diseases, Hashimoto N
et al (2012) Guidelines for diagnosis and treatment of
moyamoya disease (spontaneous occlusion of the cir-
cle of Willis). Neurol Med Chir ;52(5):245–266.
PubMed PMID: 22870528. Epub 2012/08/09. eng
4. Wakai K, Tamakoshi A, Ikezaki K, Fukui M,
Kawamura T, Aoki R et al (1997) Epidemiological
features of moyamoya disease in Japan: findings from
a nationwide survey. Clin Neurol Neurosurg 99(Suppl
2):S1–S5, PubMed PMID: 9409395. Epub 1997/12/
31. eng
5. Kleinloog R, Regli L, Rinkel GJ, Klijn CJ (2012)
Regional differences in incidence and patient charac-
teristics of moyamoya disease: a systematic review.
J Neurol Neurosurg Psychiatry 83(5):531–536,
PubMed PMID: 22378916. Epub 2012/03/02. eng
6. Kuriyama S, Kusaka Y, Fujimura M, Wakai K,
Tamakoshi A, Hashimoto S et al (2008) Prevalence
and clinicoepidemiological features of moyamoya
disease in Japan: findings from a nationwide epide-
miological survey. Stroke 39(1):42–47, PubMed
PMID: 18048855. Epub 2007/12/01. eng
7. Baba T, Houkin K, Kuroda S (2008) Novel epidemi-
ological features of moyamoya disease. J Neurol
A. Yamamoto et al.
Neurosurg Psychiatry 79(8):900–904, PubMed
PMID: 18077479. Epub 2007/12/14. eng
8. Yamauchi T, Houkin K, Tada M, Abe H (1997) Famil-
ial occurrence of moyamoya disease. Clin Neurol
Neurosurg 99(Suppl 2):S162–S167, PubMed PMID:
9409429. Epub 1997/12/31. eng
9. Nanba R, Kuroda S, Tada M, Ishikawa T, Houkin K,
Iwasaki Y (2006) Clinical features of familial
moyamoya disease. Child’s Nerv Syst 22
(3):258–262, PubMed PMID: 16195874. Epub
2005/10/01. eng
10. Houkin K, Yoshimoto T, Abe H, Nagashima K,
Nagashima M, Takeda M et al (1998) Role of basic
fibroblast growth factor in the pathogenesis of
moyamoya disease. Neurosurg Focus 5(5):e2,
PubMed PMID: 17112205. Epub 2006/11/23. eng
11. Mineharu Y, Takenaka K, Yamakawa H, Inoue K,
Ikeda H, Kikuta KI et al (2006) Inheritance pattern
of familial moyamoya disease: autosomal dominant
mode and genomic imprinting. J Neurol Neurosurg
Psychiatry 77(9):1025–1029, PubMed PMID:
16788009. Pubmed Central PMCID: PMC2077755.
Epub 2006/06/22. eng
12. Kamada F, Aoki Y, Narisawa A, Abe Y,
Komatsuzaki S, Kikuchi A et al (2011) A genome-
wide association study identifies RNF213 as the first
moyamoya disease gene. J Hum Genet 56(1):34–40,
PubMed PMID: 21048783. Epub 2010/11/05. eng
13. Liu W, Morito D, Takashima S, Mineharu Y,
Kobayashi H, Hitomi T et al (2011) Identification of
RNF213 as a susceptibility gene for moyamoya dis-
ease and its possible role in vascular development.
PLoS One 6(7):e22542, PubMed PMID: 21799892.
Pubmed Central PMCID: PMC3140517. Epub 2011/
07/30. eng
14. Houkin K, Ito M, Sugiyama T, Shichinohe H,
Nakayama N, Kazumata K et al (2012) Review
of past research and current concepts on the etiology
of moyamoya disease. Neurol Med Chir 52
(5):267–277, PubMed PMID: 22688062. Epub
2012/06/13. eng
15. Kuroda S, Hashimoto N, Yoshimoto T, Iwasaki Y
(2007) Radiological findings, clinical course, and
outcome in asymptomatic moyamoya disease: results
of multicenter survey in Japan. Stroke 38
(5):1430–1435, PubMed PMID: 17395863. Epub
2007/03/31. eng
16. Kuroda S (2010) Asymptomatic moyamoya disease.
In: Cho B-K, Tominaga T (eds) Moyamoya disease
update. Springer, Tokyo, pp 336–340
17. Kuroda S, Ishikawa T, Houkin K, Nanba R,
Hokari M, Iwasaki Y (2005) Incidence and clinical
features of disease progression in adult moyamoya
disease. Stroke 36(10):2148–2153, PubMed PMID:
16179571. Epub 2005/09/24. eng
18. Sawada T, Yamamoto A, Miki Y, Kikuta K, Okada T,
Kanagaki M et al (2012) Diagnosis of moyamoya
disease using 3-T MRI and MRA: value of cisternal
moyamoya vessels. Neuroradiology 54
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis)
(10):1089–1097, PubMed PMID: 22349623. Epub
2012/02/22
19. Fukui M (1997) Guidelines for the diagnosis and
treatment of spontaneous occlusion of the circle of
Willis (‘moyamoya’ disease). Research Committee
on Spontaneous Occlusion of the Circle of Willis
(Moyamoya Disease) of the Ministry of Health and
Welfare, Japan. Clin Neurol Neurosurg 99(Suppl 2):
S238–S240, PubMed PMID: 9409446. Epub 1997/
12/31
20. Houkin K, Abe H, Yoshimoto T, Takahashi A (1996)
Is “unilateral” moyamoya disease different from
moyamoya disease? J Neurosurg 85(5):772–776,
PubMed PMID: 8893713. Epub 1996/11/01. eng
21. Kawano T, Fukui M, Hashimoto N, Yonekawa Y
(1994) Follow-up study of patients with “unilateral”
moyamoya disease. Neurol Med Chir 34
(11):744–747, PubMed PMID: 7533268. Epub
1994/11/01. eng
22. Kelly ME, Bell-Stephens TE, Marks MP, Do HM,
Steinberg GK (2006) Progression of unilateral
moyamoya disease: a clinical series. Cerebrovasc
Dis (Basel, Switzerland) 22(2–3):109–115, PubMed
PMID: 16685122. Epub 2006/05/11. eng
23. Cheong HI, Choi Y (2010) Systemic arterial involve-
ment in moyamoya disease. In: Cho B-K, Tominaga T
(eds) Moyamoya disease update. Springer, Tokyo, pp
126–131
24. Fukui M, Kono S, Sueishi K, Ikezaki K (2000)
Moyamoya disease. Neuropathology 20(Suppl):
S61–S64, PubMed PMID: 11037190. Epub 2000/
10/19. eng
25. Oka K, Yamashita M, Sadoshima S, Tanaka K (1981)
Cerebral haemorrhage in moyamoya disease at
autopsy. Virchows Arch A Pathol Anat Histol 392
(3):247–261, PubMed PMID: 7269227. Epub 1981/
01/01. eng
26. Takagi Y, Kikuta K, Sadamasa N, Nozaki K, Hashi-
moto N (2006) Caspase-3-dependent apoptosis in
middle cerebral arteries in patients with moyamoya
disease. Neurosurgery 59(4):894–900, discussion -1.
PubMed PMID: 17038954. Epub 2006/10/14. eng
27. Yamashita M, Oka K, Tanaka K (1983) Histopathol-
ogy of the brain vascular network in moyamoya dis-
ease. Stroke 14(1):50–58, PubMed PMID: 6823686.
Epub 1983/01/01. eng
28. Kuroda S, Houkin K (2008) Moyamoya disease: cur-
rent concepts and future perspectives. Lancet Neurol
7(11):1056–1066, PubMed PMID: 18940695. Epub
2008/10/23. eng
29. Karasawa J, Touho H, Ohnishi H, Miyamoto S,
Kikuchi H (1993) Cerebral revascularization using
omental transplantation for childhood moyamoya dis-
ease. J Neurosurg 79(2):192–196, PubMed PMID:
8331399. Epub 1993/08/01. eng
30. Miyamoto S, Kikuchi H, Karasawa J, Nagata I,
Ihara I, Yamagata S (1986) Study of the posterior
circulation in moyamoya disease. Part 2: visual dis-
turbances and surgical treatment. J Neurosurg 65
859
(4):454–460, PubMed PMID: 3760953. Epub 1986/
10/01. eng
31. Kuroda S, Houkin K, Ishikawa T, Nakayama N,
Ikeda J, Ishii N et al (2004) Determinants of intellec-
tual outcome after surgical revascularization in pedi-
atric moyamoya disease: a multivariate analysis.
Child’s Nerv Syst 20(5):302–308, PubMed PMID:
15045517. Epub 2004/03/27. eng
32. Kodama N, Aoki Y, Hiraga H, Wada T, Suzuki J
(1979) Electroencephalographic findings in
children with moyamoya disease. Arch Neurol 36
(1):16–19, PubMed PMID: 420596. Epub 1979/01/
01. eng
33. Kuroda S, Kamiyama H, Isobe M, Houkin K, Abe H,
Mitsumori K (1995) Cerebral hemodynamics and “re-
build-up” phenomenon on electroencephalogram in
children with moyamoya disease. Child’s Nerv Syst
11(4):214–219, PubMed PMID: 7621482. Epub
1995/04/01. eng
34. Irikura K, Miyasaka Y, Kurata A, Tanaka R, Fujii K,
Yada K et al (1996) A source of haemorrhage in adult
patients with moyamoya disease: the significance of
tributaries from the choroidal artery. Acta Neurochir
138(11):1282–1286, PubMed PMID: 8980730. Epub
1996/01/01. eng
35. Kuroda S, Houkin K, Kamiyama H, Abe H (2001)
Effects of surgical revascularization on peripheral
artery aneurysms in moyamoya disease: report of
three cases. Neurosurgery 49(2):463–467, PubMed
PMID: 11504126. discussion 7–8. Epub 2001/08/16.
eng
36. Kawaguchi S, Sakaki T, Morimoto T, Kakizaki T,
Kamada K (1996) Characteristics of intracranial
aneurysms associated with moyamoya disease. A
review of 111 cases. Acta Neurochir 138
(11):1287–1294, PubMed PMID: 8980731. Epub
1996/01/01. eng
37. Osanai T, Kuroda S, Nakayama N, Yamauchi T,
Houkin K, Iwasaki Y (2008) Moyamoya disease
presenting with subarachnoid hemorrhage localized
over the frontal cortex: case report. Surg Neurol 69
(2):197–200, PubMed PMID: 18261650. Epub 2008/
02/12. eng
38. Verret S, Steele JC (1971) Alternating hemiplegia in
childhood: a report of eight patients with complicated
migraine beginning in infancy. Pediatrics 47
(4):675–680, PubMed PMID: 5089756. Epub 1971/
04/01. eng
39. Shirane R, Fujimura M (2010) Headache in
moyamoya disease. In: Cho B-K, Tominaga T (eds)
Moyamoya disease update. Springer, Tokyo,
pp 110–113
40. Nogawa S, Suzuki N (2010) Involuntary movement.
In: Cho B-K, Tominaga T (eds) Moyamoya disease
update. Springer, Tokyo, pp 126–131
41. Gonzalez-Alegre P, Ammache Z, Davis PH,
Rodnitzky RL (2003) Moyamoya-induced paroxys-
mal dyskinesia. Mov Disord 18(9):1051–1056,
PubMed PMID: 14502675. Epub 2003/09/23. eng
860
42. Nishimoto A, Takeuchi S (1968) Abnormal cerebro-
vascular network related to the internal carotid arter-
ies. J Neurosurg 29(3):255–260, PubMed PMID:
5684407. Epub 1968/09/01. eng
43. Houkin K, Iihoshi S, Mikami T (2010) Overview of
image diagnosis of moyamoya disease. In: Cho B-K,
Tominaga T (eds) Moyamoya disease update.
Springer, Tokyo, pp 141–149
44. Sugino T, Mikami T, Ohtaki S, Hirano T, Iihoshi S,
Houkin K et al (2013) Assessment of moyamoya
disease using multidetector row computed tomogra-
phy. J Stroke Cerebrovasc Dis 22(5):644–649,
PubMed PMID: 22365710. Epub 2012/03/01. eng
45. Houkin K, Nakayama N, Kuroda S, Nonaka T,
Shonai T, Yoshimoto T (2005) Novel magnetic reso-
nance angiography stage grading for moyamoya dis-
ease. Cerebrovasc Dis 20(5):347–354, PubMed
PMID: 16131804. Epub 2005/09/01. eng
46. Maeda M, Tsuchida C (1999) “Ivy sign” on fluid-
attenuated inversion-recovery images in childhood
moyamoya disease. AJNR Am J Neuroradiol 20
(10):1836–1838, PubMed PMID: 10588105. Epub
1999/12/10
47. Toyoda K, Ida M, Fukuda K (2001) Fluid-attenuated
inversion recovery intraarterial signal: an early sign of
hyperacute cerebral ischemia. AJNR Am J
Neuroradiol 22(6):1021–1029, PubMed PMID:
11415892. Epub 2001/06/21. eng
48. Takanashi J (2011) Moyamoya disease in children.
Brain Dev 33(3):229–234, PubMed PMID:
20932697. Epub 2010/10/12. eng
49. Kawashima M, Noguchi T, Takase Y, Ootsuka T,
Kido N, Matsushima T (2009) Unilateral hemispheric
proliferation of ivy sign on fluid-attenuated inversion
recovery images in moyamoya disease correlates
highly with ipsilateral hemispheric decrease of cere-
brovascular reserve. AJNR Am J Neuroradiol 30
(9):1709–1716, PubMed PMID: 19713323. Epub
2009/08/29. eng
50. Kawashima M, Noguchi T, Takase Y, Nakahara Y,
Matsushima T (2010) Decrease in leptomeningeal ivy
sign on fluid-attenuated inversion recovery images
after cerebral revascularization in patients with
moyamoya disease. AJNR Am J Neuroradiol 31
(9):1713–1718, PubMed PMID: 20466798. Epub
2010/05/15. eng
51. Mori N, Mugikura S, Higano S, Kaneta T,
Fujimura M, Umetsu A et al (2009) The
leptomeningeal “ivy sign” on fluid-attenuated inver-
sion recovery MR imaging in moyamoya disease: a
sign of decreased cerebral vascular reserve? AJNR
Am J Neuroradiol 30(5):930–935, PubMed PMID:
19246527. Epub 2009/02/28. eng
52. Fushimi Y, Miki Y, Kikuta K, Okada T, Kanagaki M,
Yamamoto A et al (2006) Comparison of 3.0- and 1.5-
T three-dimensional time-of-flight MR angiography
in moyamoya disease: preliminary experience. Radi-
ology 239(1):232–237, PubMed PMID: 16467209.
Epub 2006/02/10. eng
A. Yamamoto et al.
53. Mori N, Miki Y, Kikuta K, Fushimi Y, Okada T,
Urayama S et al (2008) Microbleeds in moyamoya
disease: susceptibility-weighted imaging versus T2-
weighted imaging at 3 Tesla. Invest Radiol 43
(8):574–579, PubMed PMID: 18648257. Epub
2008/07/24. eng
54. Kikuta K, Takagi Y, Nozaki K, Hanakawa T, Okada T,
Mikuni N et al (2005) Asymptomatic microbleeds in
moyamoya disease: T2-weighted gradient-echo
magnetic resonance imaging study. J Neurosurg 102
(3):470–475, PubMed PMID: 15796381. Epub 2005/
03/31
55. Horie N, Morikawa M, Nozaki A, Hayashi K,
Suyama K, Nagata I (2011) “Brush Sign” on
susceptibility-weighted MR imaging indicates the
severity of moyamoya disease. AJNR Am J
Neuroradiol 32(9):1697–1702, PubMed PMID:
21799039. Epub 2011/07/30. eng
56. Hayashi K, Nagata I (2010) Progression of
moyamoya disease. In: Cho B-K, Tominaga T (eds)
Moyamoya disease update. Springer, Tokyo,
pp 118–125
57. Suzuki J, Kodama N (1983) Moyamoya disease – a
review. Stroke 14(1):104–109, PubMed PMID:
6823678. Epub 1983/01/01. eng
58. Kodama N, Suzuki J (1978) Moyamoya disease asso-
ciated with aneurysm. J Neurosurg 48(4):565–569,
PubMed PMID: 632880. Epub 1978/04/01. eng
59. Currie S, Raghavan A, Batty R, Connolly DJ,
Griffiths PD (2011) Childhood moyamoya disease
and moyamoya syndrome: a pictorial review. Pediatr
Neurol 44(6):401–413, PubMed PMID: 21555050.
Epub 2011/05/11
60. Maki Y, Enomoto T (1988) Moyamoya disease.
Child’s Nerv Syst 4(4):204–212, PubMed PMID:
3048657. Epub 1988/08/01
61. Yamada I, Suzuki S, Matsushima Y (1995)
Moyamoya disease: comparison of assessment with
MR angiography and MR imaging versus conven-
tional angiography. Radiology 196(1):211–218,
PubMed PMID: 7784569. Epub 1995/07/01. eng
62. Tsuchiya K, Makita K, Furui S (1994) Moyamoya
disease: diagnosis with three-dimensional CT angiog-
raphy. Neuroradiology 36(6):432–434, PubMed
PMID: 7991085. Epub 1994/08/01
63. Kikuchi M, Asato M, Sugahara S, Nakajima K,
Sato M, Nagao K et al (1996) Evaluation of surgically
formed collateral circulation in moyamoya
disease with 3D-CT angiography: comparison with
MR angiography and X-ray angiography.
Neuropediatrics 27(1):45–49, PubMed PMID:
8677026. Epub 1996/02/01
64. Ogasawara K, Ogawa A, Yoshimoto T (2002) Cere-
brovascular reactivity to acetazolamide and outcome
in patients with symptomatic internal carotid or mid-
dle cerebral artery occlusion: a xenon-133 single-
photon emission computed tomography study. Stroke
33(7):1857–1862, PubMed PMID: 12105366. Epub
2002/07/10
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis)
65. Nariai T, Matsushima Y, Imae S, Tanaka Y, Ishii K,
Senda M et al (2005) Severe haemodynamic stress in
selected subtypes of patients with moyamoya disease:
a positron emission tomography study. J Neurol
Neurosurg Psychiatry 76(5):663–669, PubMed
PMID: 15834024. Pubmed Central PMCID:
1739646. Epub 2005/04/19. eng
66. Czabanka M, Pena-Tapia P, Schubert GA, Heppner
FL, Martus P, Horn P et al (2011) Proposal for a new
grading of moyamoya disease in adult patients.
Cerebrovasc Dis (Basel, Switzerland) 32(1):41–50,
PubMed PMID: 21576942. Epub 2011/05/18
67. Tanaka Y, Nariai T, Nagaoka T, Akimoto H,
Ishiwata K, Ishii K et al (2006) Quantitative evalua-
tion of cerebral hemodynamics in patients with
moyamoya disease by dynamic susceptibility contrast
magnetic resonance imaging–comparison with posi-
tron emission tomography. J Cereb Blood Flow
Metab 26(2):291–300, PubMed PMID: 16049427.
Epub 2005/07/29
68. Ikezaki K, Matsushima T, Kuwabara Y, Suzuki SO,
Nomura T, Fukui M (1994) Cerebral circulation and
oxygen metabolism in childhood moyamoya disease:
a perioperative positron emission tomography study.
J Neurosurg 81(6):843–850, PubMed PMID:
7965114. Epub 1994/12/01
69. Kanno I, Uemura K, Higano S, Murakami M, Iida H,
Miura S et al (1988) Oxygen extraction fraction at
maximally vasodilated tissue in the ischemic brain
estimated from the regional CO2 responsiveness mea-
sured by positron emission tomography. J Cereb
Blood Flow Metab 8(2):227–235, PubMed PMID:
3125186. Epub 1988/04/01
70. Saito N, Nakagawara J, Nakamura H, Teramoto A
(2004) Assessment of cerebral hemodynamics in
childhood moyamoya disease using a quantitative
and a semiquantitative IMP-SPECT study. Ann Nucl
Med 18(4):323–331, PubMed PMID: 15359926.
Epub 2004/09/14
71. So Y, Lee HY, Kim SK, Lee JS, Wang KC, Cho BK
et al (2005) Prediction of the clinical outcome of
pediatric moyamoya disease with postoperative
basal/acetazolamide stress brain perfusion SPECT
after revascularization surgery. Stroke 36
(7):1485–1489, PubMed PMID: 15947261. Epub
2005/06/11
72. Touho H, Karasawa J, Ohnishi H (1996) Preoperative
and postoperative evaluation of cerebral perfusion
and vasodilatory capacity with 99mTc-HMPAO
SPECT and acetazolamide in childhood Moyamoya
disease. Stroke 27(2):282–289, PubMed PMID:
8571424. Epub 1996/02/01
73. Iida H, Nakagawara J, Hatazawa J, Suzuki M,
Takahashi M, Shimosegawa E et al (2012) Estimation
of the arterial blood radioactivity concentration using
the whole body-to-arterial blood partition coefficients
and the cross-calibration factor in 123I-
iodoamphetamine SPECT – towards a noninvasive
clinical protocol with the QSPECT-DTARG method.
861
Kaku Igaku 49(2):49–58, PubMed PMID: 22779347.
Epub 2012/07/12
74. Hatazawa J, Iida H, Shimosegawa E, Sato T,
Murakami M, Miura Y (1997) Regional cerebral
blood flow measurement with iodine-123-IMP auto-
radiography: normal values, reproducibility and sen-
sitivity to hypoperfusion. J Nucl Med 38
(7):1102–1108, PubMed PMID: 9225798. Epub
1997/07/01
75. Iida H, Itoh H, Nakazawa M, Hatazawa J,
Nishimura H, Onishi Y et al (1994) Quantitative map-
ping of regional cerebral blood flow using iodine-
123-IMP and SPECT. J Nucl Med 35
(12):2019–2030, PubMed PMID: 7989987. Epub
1994/12/01
76. Kang KH, Kim HS, Kim SY (2008) Quantitative
cerebrovascular reserve measured by acetazolamide-
challenged dynamic CT perfusion in ischemic adult
moyamoya disease: initial experience with angio-
graphic correlation. AJNR Am J Neuroradiol 29
(8):1487–1493, PubMed PMID: 18499785. Epub
2008/05/24
77. Rim NJ, Kim HS, Shin YS, Kim SY (2008) Which
CT perfusion parameter best reflects cerebrovascular
reserve?: correlation of acetazolamide-challenged CT
perfusion with single-photon emission CT in
Moyamoya patients. AJNR Am J Neuroradiol 29
(9):1658–1663, PubMed PMID: 18617583. Epub
2008/07/12
78. Sakamoto S, Ohba S, Shibukawa M, Kiura Y, Arita K,
Kurisu K (2006) CT perfusion imaging for childhood
moyamoya disease before and after surgical revascu-
larization. Acta Neurochir 148(1):77–81, discussion
PubMed PMID: 16184319. Epub 2005/09/27
79. Wintermark M, Maeder P, Verdun FR, Thiran JP,
Valley JF, Schnyder P et al (2000) Using 80 kVp
versus 120 kVp in perfusion CT measurement of
regional cerebral blood flow. AJNR Am J Neuroradiol
21(10):1881–1884, PubMed PMID: 11110541. Epub
2000/12/08
80. Kudo K, Christensen S, Sasaki M, Ostergaard L,
Shirato H, Ogasawara K et al (2013) Accuracy and
reliability assessment of CT and MR perfusion anal-
ysis software using a digital phantom. Radiology 267
(1):201–211, PubMed PMID: 23220899. Pubmed
Central PMCID: 3606546. Epub 2012/12/12
81. Zhang J, Wang J, Geng D, Li Y, Song D, Gu Y (2013)
Whole-brain CT perfusion and CT angiography
assessment of moyamoya disease before and after
surgical revascularization: preliminary study with
256-slice CT. PLoS One 8(2):e57595, PubMed
PMID: 23451248. Pubmed Central PMCID:
3579776. Epub 2013/03/02
82. Morkenborg J, Pedersen M, Jensen FT, Stodkilde-
Jorgensen H, Djurhuus JC, Frokiaer J (2003) Quanti-
tative assessment of Gd-DTPA contrast agent from
signal enhancement: an in-vitro study. Magn Reson
Imaging 21(6):637–643, PubMed PMID: 12915195.
Epub 2003/08/14
862
83. Nardone B, Saddleton E, Laumann AE, Edwards BJ,
Raisch DW, McKoy JM et al (2014) Pediatric
nephrogenic systemic fibrosis is rarely reported: a
RADAR report. Pediatr Radiol 44(2):173–180,
PubMed PMID: 24057195. Epub 2013/09/24
84. Detre JA, Subramanian VH, Mitchell MD, Smith DS,
Kobayashi A, Zaman A et al (1990) Measurement of
regional cerebral blood flow in cat brain using
intracarotid 2H2O and 2H NMR imaging. Magn
Reson Med 14(2):389–395, PubMed PMID:
2345518. Epub 1990/05/01
85. Kuwabara Y, Ichiya Y, Sasaki M, Yoshida T,
Masuda K, Matsushima T et al (1997) Response to
hypercapnia in moyamoya disease. Cerebrovascular
response to hypercapnia in pediatric and adult patients
with moyamoya disease. Stroke 28(4):701–707,
PubMed PMID: 9099182. Epub 1997/04/01
86. Aoyama K, Fushimi Y, Okada T, Miyasaki A, Taki H,
Shibamoto K et al (2012) Detection of symptomatic
vasospasm after subarachnoid haemorrhage: initial
findings from single time-point and serial measure-
ments with arterial spin labelling. Eur Radiol 22
(11):2382–2391, PubMed PMID: 22653285. Epub
2012/06/02
87. Saida T, Masumoto T, Nakai Y, Shiigai M,
Matsumura A, Minami M (2012) Moyamoya disease:
evaluation of postoperative revascularization using
multiphase selective arterial spin labeling MRI. J
Comput Assist Tomogr 36(1):143–149, PubMed
PMID: 22261785. Epub 2012/01/21. eng
88. Noguchi T, Kawashima M, Irie H, Ootsuka T,
Nishihara M, Matsushima T et al (2011) Arterial
spin-labeling MR imaging in moyamoya disease
compared with SPECT imaging. Eur J Radiol 80(3):
e557–e562, PubMed PMID: 21315533. Epub 2011/
02/15. eng
89. Zaharchuk G, Do HM, Marks MP, Rosenberg J,
Moseley ME, Steinberg GK (2011) Arterial spin-
labeling MRI can identify the presence and intensity
of collateral perfusion in patients with moyamoya
disease. Stroke 42(9):2485–2491, PubMed PMID:
21799169. Pubmed Central PMCID: 3164217. Epub
2011/07/30. eng
90. Lee HY, Paeng JC, Lee DS, Lee JS, Oh CW, Cho MJ
et al (2004) Efficacy assessment of cerebral arterial
bypass surgery using statistical parametric mapping
and probabilistic brain atlas on basal/acetazolamide
brain perfusion SPECT. J Nucl Med 45(2):202–206,
PubMed PMID: 14960636. Epub 2004/02/13. eng
91. Sugino T, Mikami T, Miyata K, Suzuki K, Houkin K,
Mikuni N (2013) Arterial spin-labeling magnetic res-
onance imaging after revascularization of moyamoya
disease. J Stroke Cerebrovasc Dis 22(6):811–816,
PubMed PMID: 22721824. Epub 2012/06/23
92. Zhao WG, Luo Q, Jia JB, Yu JL (2013) Cerebral
hyperperfusion syndrome after revascularization sur-
gery in patients with moyamoya disease. Br J
Neurosurg 27(3):321–325, PubMed PMID:
23461748. Epub 2013/03/07
A. Yamamoto et al.
93. Hwang JW, Yang HM, Lee H, Lee HK, Jeon YT, Kim
JE et al (2013) Predictive factors of symptomatic
cerebral hyperperfusion after superficial temporal
artery-middle cerebral artery anastomosis in adult
patients with moyamoya disease. Br J Anaesth 110
(5):773–779, PubMed PMID: 23274781. Epub 2013/
01/01
94. Ng J, Thompson D, Lumley JP, Saunders DE,
Ganesan V (2012) Surgical revascularisation for
childhood moyamoya. Child’s Nerv Syst 28
(7):1041–1048, PubMed PMID: 22570164. Epub
2012/05/10
95. Cai X, Han S, Feske SK, Chou SH (2013) Pearls and
oy-sters: small but consequential: intracerebral hem-
orrhage caused by lenticulostriate artery aneurysm.
Neurology 80(9):e89–e91, PubMed PMID:
23439707. Pubmed Central PMCID: 3598457. Epub
2013/02/27
96. Gotoh K, Okada T, Satogami N, Yakami M,
Takahashi JC, Yoshida K et al (2012) Evaluation of
CT angiography for visualisation of the
lenticulostriate artery: difference between normoten-
sive and hypertensive patients. Br J Radiol 85(1019):
e1004–e1008, PubMed PMID: 22744324. Pubmed
Central PMCID: 3500797. Epub 2012/06/30
97. Gotoh K, Okada T, Miki Y, Ikedo M, Ninomiya A,
Kamae T et al (2009) Visualization of the
lenticulostriate artery with flow-sensitive black-
blood acquisition in comparison with time-of-flight
MR angiography. J Magn Reson Imaging 29
(1):65–69, PubMed PMID: 19097113. Epub 2008/
12/20
98. Okuchi S, Okada T, Ihara M, Gotoh K, Kido A,
Fujimoto K et al (2013) Visualization of
lenticulostriate arteries by flow-sensitive black-blood
MR angiography on a 1.5 T MRI system: a compar-
ative study between subjects with and without stroke.
AJNR Am J Neuroradiol 34(4):780–784, PubMed
PMID: 23064597. Epub 2012/10/16
99. Kimura T, Ikedo M, Takemoto S (2009) Hybrid of
opposite-contrast MR angiography
(HOP-MRA) combining time-of-flight and flow-
sensitive black-blood contrasts. Magn Reson Med
62(2):450–458, PubMed PMID: 19526496. Epub
2009/06/16. eng
100. Tsuchiya K, Kobayashi K, Nitatori T, Kimura T,
Ikedo M, Takemoto S (2010) Hybrid of opposite-
contrast MRA of the brain by combining time-of-
flight and black-blood sequences: initial experience
in major trunk stenoocclusive diseases. J Magn Reson
Imaging 31(1):56–60, PubMed PMID: 20027573.
Epub 2009/12/23. eng
101. Miyamoto S, Akiyama Y, Nagata I, Karasawa J,
Nozaki K, Hashimoto N et al (1998) Long-term out-
come after STA-MCA anastomosis for moyamoya
disease. Neurosurg Focus 5(5):e5, PubMed PMID:
17112208
102. Nakagawara J, Takeda R, Suematsu K, Nakamura J
(1997) Quantification of regional cerebral blood flow
36 Moyamoya Disease (Spontaneous Occlusion of the Circle of Willis)
and vascular reserve in childhood moyamoya disease
using [123I]IMP-ARG method. Clin Neurol
Neurosurg 99(Suppl 2):S96–S99, PubMed PMID:
9409415. Epub 1997/12/31. eng
103. Research Comittee on Spontaneous Occlusion of the
Circle of Willis (Moyamoya Disease) (2009)
Research on intractable diseases of the Ministry of
Health LaW, Japan. Recommendations for the Man-
agement of Moyamoya Disease. A Statement from
Research Committee on Spontaneous Occlusion of
the Circle of Willis (Moyamoya Disease). Surg
Cereb Stroke 37(5):321–337. Japanese
104. Karasawa J, Kikuchi H, Furuse S, Kawamura J,
Sakaki T (1978) Treatment of moyamoya disease
with STA-MCA anastomosis. J Neurosurg 49
(5):679–688, PubMed PMID: 712390
105. Miyamoto S, Kikuchi H, Karasawa J, Nagata I,
Yamazoe N, Akiyama Y (1988) Pitfalls in the surgical
treatment of moyamoya disease. Operative tech-
niques for refractory cases. J Neurosurg 68
(4):537–543, PubMed PMID: 3351581
106. Karasawa J, Kikuchi H, Furuse S, Sakaki T, Yoshida
Y (1977) A surgical treatment of “moyamoya” dis-
ease “encephalo-myo synangiosis”. Neurol Med Chir
17(1 Pt 1):29–37, PubMed PMID: 74029. Epub 1977/
01/01. eng
107. Takeuchi S, Abe H, Ozawa T, Tanaka R (2000) Sur-
gical treatment for moyamoya disease. Surgical effect
of encephalo-galeo-synangiosis on moyamoya dis-
ease. Surg Cereb Stroke 28:98–103, Japanese
108. Matsushima Y, Fukai N, Tanaka K, Tsuruoka S,
Inaba Y, Aoyagi M et al (1981) A new surgical treat-
ment of moyamoya disease in children: a preliminary
report. Surg Neurol 15(4):313–320, PubMed PMID:
7245020
109. Karasawa J, Kikuchi H, Kawamura J, Sakai T (1980)
Intracranial transplantation of the omentum for cere-
brovascular moyamoya disease: a two-year follow-up
study. Surg Neurol 14(6):444–449, PubMed PMID:
7221855
110. Nakagawa Y, Abe H, Kamiyama H, Sawamura Y,
Gotoh S, Kashiwaba T (1988) Revascularization sur-
gery for 50 patients with moyamoya disease. In:
Suzuki J (ed) Advances in surgery for cerebral stroke.
Springer, Tokyo, pp 141–149
111. Fujimura M, Kaneta T, Mugikura S, Shimizu H,
Tominaga T (2007) Temporary neurologic deteriora-
tion due to cerebral hyperperfusion after superficial
temporal artery-middle cerebral artery anastomosis in
patients with adult-onset moyamoya disease. Surg
Neurol 67(3):273–282, PubMed PMID: 17320638.
Epub 2007/02/27. eng
863
112. Nishimoto A (1979) Moyamoya disease (author’s
transl). Neurol Med Chir 19(3):221–228, PubMed
PMID: 86959
113. Yoshida Y, Yoshimoto T, Shirane R, Sakurai Y (1999)
Clinical course, surgical management, and long-term
outcome of moyamoya patients with rebleeding after
an episode of intracerebral hemorrhage: an extensive
follow-up study. Stroke 30(11):2272–2276, PubMed
PMID: 10548656
114. Takahashi JC, Miyamoto S (2010) Moyamoya dis-
ease: recent progress and outlook. Neurol Med Chir
50(9):824–832, PubMed PMID: 20885116
115. Nishimoto A, Ueda K, Honma Y (eds) (1982)
Follow-up study on outcome of the occlusion of
the circle of Willis. Proceedings of the Research
Committee on Spontaneous Occlusion of the
Circle of Willis. Ministry of Health and Welfare,
Tokyo
116. Kobayashi E, Saeki N, Oishi H, Hirai S, Yamaura A
(2000) Long-term natural history of hemorrhagic
moyamoya disease in 42 patients. J Neurosurg 93
(6):976–980, PubMed PMID: 11117870
117. Houkin K, Kamiyama H, Abe H, Takahashi A,
Kuroda S (1996) Surgical therapy for adult
moyamoya disease. Can surgical revascularization
prevent the recurrence of intracerebral hemorrhage?
Stroke 27(8):1342–1346, PubMed PMID: 8711799.
Epub 1996/08/01. eng
118. Kawaguchi S, Okuno S, Sakaki T (2000) Effect of
direct arterial bypass on the prevention of future
stroke in patients with the hemorrhagic variety of
moyamoya disease. J Neurosurg 93(3):397–401,
PubMed PMID: 10969936
119. Fujii K, Ikezaki K, Irikura K, Miyasaka Y, Fukui M
(1997) The efficacy of bypass surgery for the patients
with hemorrhagic moyamoya disease. Clin Neurol
Neurosurg 99(Suppl 2):S194–S195, PubMed PMID:
9409436
120. Miyamoto S (2004) Study design for a prospective
randomized trial of extracranial-intracranial bypass
surgery for adults with moyamoya disease and hem-
orrhagic onset–the Japan Adult Moyamoya Trial
Group. Neurol Med Chir 44(4):218–219, PubMed
PMID: 15185765. Epub 2004/06/10. eng
121. Wiebers DO (1985) Ischemic cerebrovascular com-
plications of pregnancy. Arch Neurol 42
(11):1106–1113, PubMed PMID: 2864911
122. Takahashi JC, Ikeda T, Iihara K, Miyamoto S (2012)
Pregnancy and delivery in moyamoya disease: results
of a nationwide survey in Japan. Neurol Med Chir
52(5):304–310, PubMed PMID: 22688067
 Brain Death Imaging
37
M. Sawicki, Joanna Wojczal, Bozena Birkenfeld, and
Lech Cyrylowski

Contents Prerequisites for Blood Flow Studies

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 866
The Role of Imaging Tests in Determination
of Brain Death . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 867
Pathophysiological Mechanism of Cerebral
Circulatory Arrest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 867
Imaging of Structural Changes in
Brain Death . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 869
CT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 869
MRI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 869
Functional Imaging of Brain Death . . . . . . . . . . . . . . . 871
Imaging of Cerebral Circulatory Arrest . . . . . . . . . . 872
in the Diagnosis of Brain Death . . . . . . . . . . . . . . . . . . . . . 872
Catheter Cerebral Angiography . . . . . . . . . . . . . . . . . . . . . 872
CT Angiography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 874
Perfusion CT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 878
Xenon CT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 880
MR Angiography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 881
Transcranial Doppler Sonography . . . . . . . . . . . . . . . . . . . 881
Nuclear Medicine Methods in the Diagnosis
of Brain Death . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886
Limitations of Blood Flow Studies
in the Diagnosis of Brain Death . . . . . . . . . . . . . . . . . . . 889
Future Trends: Looking for an Ideal
Ancillary Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 889
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 892
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 893

M. Sawicki (*)
Department of Diagnostic Imaging and Interventional
Radiology, Pomeranian Medical University, Szczecin,
Poland
e-mail: msaw@pum.edu.pl; msaw3108@gmail.com
J. Wojczal
Department of Neurology, Medical University of Lublin,
Lublin, Poland
e-mail: jwojczal@poczta.onet.pl
B. Birkenfeld
Department of Nuclear Medicine, Pomeranian Medical
University, Szczecin, Poland
e-mail: birka@pum.edu.pl
L. Cyrylowski
CT and MRI Clinic, “Nasz Doktor” Health Care
Institution, Szczecin, Poland
MRI Clinic, “Euromedic” Health Care Institution,
Szczecin, Poland
e-mail: blecyr@hotmail.com

# Springer Science+Business Media New York 2016 865

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_26
866 M. Sawicki et al.

spectroscopy • Diffusion-weighted MRI •

Abstract
Functional MRI
Brain death is defined as the irreversible cessa-
tion of functioning of the entire brain, includ-
ing the brain stem. Brain death is principally
Introduction
established using clinical criteria including
coma, absence of brain stem reflexes, and loss
The concept of death is an important part of med-
of central drive to breathe, assessed with apnea
icine since the beginning of its history. In modern
test. In situations in which clinical testing can-
medicine, the death of a human is identified with
not be performed or when uncertainty exists
the death of his/her brain. The earliest description
about the reliability of its parts due to
of brain death was “coma dépassé” (beyond
confounding conditions, ancillary tests (i.e.,
coma) by Mollaret and Goulon in 1959. The first
imaging studies) may be useful. The objective
formalized definition of brain death referred to as
of ancillary tests in the diagnosis of brain death
irreversible coma was by the Ad Hoc Committee
is to demonstrate the absence of cerebral elec-
of the Harvard Medical School in 1968. The def-
trical activity (EEG and evoked potentials) or
inition has evolved over the next years, since there
cerebral circulatory arrest. In clinical practice
has been a greater understanding of neuronal
catheter cerebral angiography, perfusion scin-
function and mechanisms of brain cell injury.
tigraphy, transcranial Doppler sonography, CT
Nowadays, brain death is defined as the irrevers-
angiography, and MR angiography are used.
ible cessation of functioning of the entire brain,
Other methods, like perfusion CT, xenon CT,
including the brain stem with two pathways to
MR spectroscopy, diffusion-weighted MRI,
reach that point: (i) permanent absence of circula-
and functional MRI, are being studied as
tion and (ii) subsequent to a catastrophic brain
potentially useful in the diagnosis of brain
injury [1].
death.
The definition and diagnosis of brain death are
Among these tests, blood flow and perfu-
of particular importance in view of the medical
sion studies are most commonly applied. The
and technological advances in supporting,
diagnosis of brain death can be confirmed
repairing, or replacing failing organs. The confir-
when a lack of blood flow or tissue perfusion
mation of brain death, synonymous with the dec-
in the whole brain is documented.
laration of death, is a necessary condition for the
However, the application of blood flow and
termination of futile life support and potential
perfusion tests to the diagnosis of brain death is
procurement of organs for transplantation. There-
limited by variable vulnerability of individual
fore, diagnosis of brain death is one of the main
brain structures to ischemia. Besides there are
points of interest in intensive and critical care and
known factors predisposing to dissociation
transplantation medicine.
between brain death and total cessation of cere-
Brain death is principally established using
bral circulation. These factors include skull
clinical criteria defined as diagnostic testing
decompression, infratentorial localization,
based on direct, measurable observation or exam-
and hypoxic-ischemic mechanism of brain
ination of the patient. Because the clinical criteria
injury.
vary in detail between countries, the minimum
acceptable clinical standard has been agreed to
Keywords
test for the cessation of brain function [1] – see
Brain death • Cerebral circulatory arrest •
Table 1.
Ancillary test • Catheter angiography • Com-
To confirm brain death according to clinical
puted tomographic angiography • Perfusion
criteria, prerequisites should be fulfilled prior to
CT • Xenon CT • Transcranial Doppler sonog-
the application of the diagnostic tests – see
raphy • Perfusion scintigraphy • Magnetic res-
Table 2. Most countries mandate that the clinical
onance imaging • MR angiography • MR
37 Brain Death Imaging
Table 1 Minimum acceptable clinical standard for the
cessation of brain function
1. Coma (excluding spinal cord-mediated reflexes)
2. Absence of brain stem reflexes
(a) Pupils mid-position or greater and absent pupillary
light reflex (fixed dilated pupils)
(b) Corneal
(c) Gag/pharyngeal
(d) Cough/tracheal
(e) Vestibulo-ocular (“cold caloric”)
(f) Loss of central drive to breathe, assessed with apnea
test
Table 2 Preconditions: patient-related prerequisites that
should be fulfilled prior to application of diagnostic tests
1. Established etiology and/or structural lesion capable of
causing death by neurological criteriaa
2. Reduced consciousness (as measured by GCS 3–5)
3. Evidence for progressing loss of brain stem function
a
Neuroimaging studies are useful for assessing the severity
of brain injury
diagnosis be confirmed by more than one physi-
cian, either two or three.
The Role of Imaging Tests
in Determination of Brain Death
There are known circumstances during which a
clinical examination to confirm brain death may
become unreliable or cannot be completely
performed. These confounding conditions that
may impede or invalidate clinical diagnosis of
brain death are listed in Table 3.
In situations in which clinical testing cannot be
performed or when uncertainty exists about the
reliability of its parts due to confounding condi-
tions, ancillary tests (i.e., imaging studies) may be
useful. Ancillary tests are applied as complemen-
tary to clinical testing that otherwise, for any
reason, cannot be conducted or is unreliable. Fur-
thermore, the usage of ancillary tests can reduce
the time taken to confirm the diagnosis and
increase its confidence in all cases of suspected
brain death regardless of the presence of
confounding conditions. Early determination of
867
Table 3 Confounding conditions that may impede or
invalidate the clinical diagnosis of brain death
1. Trauma to the eyes
2. Trauma to the middle and/or inner ears
3. Nerve or muscle dysfunction or neuromuscular
blockade potentially accounting for unresponsiveness
4. Presence of atypical responses
5. Unresuscitated shock
6. Acquired or therapeutic hypothermia
7. Severe metabolic and endocrine abnormality that is
difficult to reverse (e.g., glucose, acid-base, electrolytes)
8. Clinically significant drug intoxications (e.g.,
alcohol, barbiturates, sedatives, hypnotics)
9. Inability to perform apnea test
10. High cervical spine injury
11. Primary infratentorial brain injury
12. Young children
brain death may allow for avoidance of protracted
stays in the intensive care unit (ICU) and poten-
tially expedite organ donation before tissue via-
bility becomes a concern. In some protocols,
ancillary tests are used to shorten the duration of
the observation period.
The objective of ancillary tests in the diagnosis
of brain death is to demonstrate the absence of
cerebral electrical activity (EEG and evoked
potentials) or cerebral circulatory arrest. Cur-
rently, only catheter cerebral angiography and
perfusion scintigraphy are widely accepted as
ancillary tests in the determination of cerebral
circulatory arrest. However, in clinical practice
transcranial Doppler sonography (TCD), CT angi-
ography (CTA), and MR angiography (MRA) are
used as well. Other methods, like perfusion CT
(PCT), xenon CT (XeCT), MR spectroscopy
(MRS), diffusion-weighted MRI (DWI), and
functional MRI (fMRI), are being studied as
potentially useful in the diagnosis of brain death.
Pathophysiological Mechanism
of Cerebral Circulatory Arrest
Cerebral blood flow (CBF) is continuously regu-
lated by cerebral perfusion pressure (CPP) and
cerebrovascular resistance (CVR). CPP is mea-
sured as a difference between the entry pressure,
868
mean arterial pressure (MAP), and the exit pres-
sure – intracranial pressure (ICP). According to
the Poiseuille equation:
CPP MAP  ICP
CBF ¼ ¼ :
CVR CVR
The normal CBF at rest is maintained within a
range of 45–60 ml/100 g of brain tissue/min. A
brain injury is accompanied by cerebral
edema. Because the skull bones limit the intracra-
nial volume, this “mass effect” leads to the
intracranial hypertension. The process spreads
from the region of primary injury gradually
involving the whole brain, and the intracranial
vessels become compressed but preserve their
patency. Initially, CVR is kept almost constant
by the activated autoregulation mechanisms
releasing blood vessel tone. However, this phe-
nomenon is capable to preserve a sufficient blood
supply as long as CPP exceeds approx. 50 mmHg
– see Fig. 1.
Further elevation of ICP results in an increase
of CVR. The capillaries as the most susceptible
vessels are primarily affected; then the process
Fig. 1 Relationship between cerebral blood flow (CBF)
and cerebral perfusion pressure (CPP). CPP is measured as
a difference between the entry pressure, mean arterial
pressure (MAP), and the exit pressure, intracranial pressure
(ICP). In increasing ICP the cerebral autoregulation is
M. Sawicki et al.
spreads proximally progressively involving big-
ger arteries. As ICP reaches the diastolic pressure,
the end-diastolic cerebral flow completely ceases.
At this stage CBF is deeply compromised,
although still sufficient to preserve a viability of
the brain. When ICP exceeds the diastolic pres-
sure approaching MAP, the circulatory arrest
occurs in the compressed cerebral capillaries and
venules. At this time larger arteries stay patent,
and their elastic walls are distended by the blood
pushed forward by the heartbeat in systolic phase
(antegrade flow). In diastole a similar blood vol-
ume is pressed out of the collapsed arteries (retro-
grade flow). Such an oscillating, to-and-fro
movement of the blood provides the zero-net
CBF. This is an early phase of cerebral circulatory
arrest leading to necrosis of neurons within 10–15
min. With further ICP elevation, the oscillations
of blood column in the proximal arteries become
weaker. If ICP exceeds the level of systolic pres-
sure, a blood movement in the cerebral vessels
ceases completely, while oscillation can be still
observed in extracranial segments of the internal
carotid arteries (ICAs) and vertebral arteries
(VAs).
capable to preserve almost constant CBF as long as CPP
exceeds approx. 50 mmHg. Further elevation of ICP
results in deterioration of cerebral blood flow leading to
cerebral circulatory arrest
37 Brain Death Imaging
Imaging of Structural Changes in
Brain Death
CT
Non-contrast-enhanced CT (NECT) of the head is
useful for assessing the severity of brain injury in
the suspicion of brain death. The study may reveal
the features of cerebral edema including a gener-
alized hypodensity of brain parenchyma,
decreased differentiation between the gray and
white matter (GM:WM ratio), sulcal effacement,
absence of the basal cisterns, and signs of brain
herniation – see Fig. 2. It was shown that a midline
shift >10 mm, absence of the ambient cistern, and
GM:WM ratio <1.18 predict progression to brain
death. Although the study is ubiquitous, noninva-
sive, non-time-consuming, and non-cost consum-
ing, it lacks sufficient diagnostic value as an
ancillary test.
MRI
The advent of magnetic resonance imaging has
improved diagnostic possibilities in brain death, due
to several sophisticated techniques offered by MRI.
In the early 1990s, first reports demonstrated
the potential role of MRI in the diagnosis of brain
death, including MR spectroscopy [2, 3].
Fig. 2 CT findings in a
23-year-old man after
cardiac arrest who
presented with clinical signs
of brain death: (a, b) 5 mm
MPRs in axial plane show
diffuse edema, sulcal
effacement, ventricular
narrowing, low GM/WM
differentiation, and
diminution of the basal
cisterns
869
Common MRI findings in echo-planar scans
are variable edema, including diffuse gyral swell-
ing, most prominent in the acute phase, diffuse
cerebral white matter injury, diffuse cortical high
signal intensity, and tonsillar herniation
[4–6]. Some authors observed also transtentorial
herniation [4, 7, 8]. Less frequent finding is focal
edema in the upper spinal cord [5].
Poor gray/white matter differentiation after the
diagnosis of brain death is not commonly
observed. Matsumura et al. [8] found significant
increasing of the ratio of signal intensities
between white and gray matters on PD- and
T2-WI images but decreased on T1-WI images.
These phenomena may be related to different vul-
nerabilities of white and gray matter to anoxic
injury – see Fig. 3a, c.
They also showed the absence of normal flow
voids in MR echo-planar images. Moreover, with
serial MR before and after the diagnosis of brain
death, they demonstrated disappearance of flow
voids in cavernous portions of the ICAs after the
diagnosis. The disappearance of flow voids in the
ICA correlates with nonfilling of the ICA revealed
with intra-arterial digital subtraction arteriogra-
phy [8]. Sohn et al. [9] using 3-T MRI demon-
strated that tonsillar herniation and loss of intra-
arterial flow signal voids on T2-WI are highly
sensitive and specific findings in patients with
brain death.
870
Fig. 3 MR findings in a 42-year-old man after cardiac
arrest who presented with clinical signs of brain death.
First examination performed several hours after resuscita-
tion: (a) 5 mm axial T2-WI image (TR = 3,713 ms, TE =
104 ms) shows moderate increase in signal of lentiform
nuclei consistent with edema; (b) DWI (TR = 5,000 ms,
TE = 86 ms) shows restricted diffusion in lentiform nuclei.
Second examination performed the next day: (c) 5 mm
axial T2-WI image (TR = 3,713 ms, TE = 104 ms)
shows massive edema involving the basal ganglia and
cortex with narrowing of ventricles; (d) DWI (TR =
Additional signs include the bilateral
transcerebral vein sign, defined as branching
structures extending through cerebral hemi-
spheres parallel or perpendicular to the outer
wall of a lateral ventricle, and the bilateral cortical
vein sign, defined as visualization of cortical veins
of cerebral hemispheres. Both signs can be easily
demonstrated with T2*WI gradient-recalled echo
(GRE) and susceptibility-weighted images (SWI)
[9] – see Fig. 3e, f.
M. Sawicki et al.
5,000 ms, TE = 86 ms) shows restricted diffusion in
lentiform nuclei and cortex. Cortical and deep GM are
more frequently affected because neurons are more vulner-
able to hypoxia than oligodendroglia or astrocytes; (e, f)
SWI (TR = 73 ms, TE = 47 ms) shows multiple and
branching low signal intensities extending through the
cerebral hemisphere parallel or perpendicular to the outer
wall of both lateral ventricles (arrowhead, transcerebral
vein sign) and abnormal low signal intensities in both
cerebral hemisphere cortical areas (arrows, bilateral corti-
cal vein sign)
Diffusion-Weighted Imaging
Lövblad et al. first demonstrated the usefulness of
diffusion-weighted images (DWI) in the diagnosis
of brain death. They found a significant drop in the
apparent diffusion coefficient (ADC) values com-
pared with those of normal controls, and these obser-
vations have been later confirmed by other authors
[7, 10, 11]. With DWI and ADC mapping, it is
possible to identify areas corresponding to cytotoxic
edema and ischemic damage – see Fig. 3b, d.
37 Brain Death Imaging
Selcuk et al. [11] observed greater decrease in
ADC values in white matter than in gray
matter for both the cerebral and cerebellar
hemispheres. The sensitivity and specificity as
well as positive and negative predictive
cutoff values of ADC to distinguish patients
with brain death from controls were 100 % in
their study, but they recommend to
determine ADC cutoff values for different
MRI scanners and DWI protocols. They
believe that a 100 % sensitivity and specificity in
their results indicates a potential role of ADC
maps as confirmatory test in diagnosis of brain
death.
However, Luchtmann et al. evaluating the
reliability of DWI in the diagnosis of brain
death concluded that effects of pseudonorma-
lization of ADC values (after 5–28 days after
brain ischemic damage) and possible susceptibil-
ity artifacts due to microhemorrhages preclude a
leading role of DWI in the diagnosis of brain
death [10].
MR Spectroscopy
Other sophisticated technique offered by MRI is
MR spectroscopy (MRS), allowing to study
metabolic changes in brain tissue. In previously
used method of 31P MRS, the most striking
feature was a complete absence of adenosine
triphosphate (ATP) and domination of intense
inorganic phosphate signal [2, 3]. Kato et al. [3]
observed also an absence of phosphocreatine and
stated that MRS may be helpful in diagnostics of
brain death particularly in children, when other
examinations are not sufficient to establish
defined diagnosis.
Falini et al. [12] performing sequential MR
imaging and proton (1H) MRS in a patient
with severe hypoxic-ischemic brain injury
found that observed structural and
biochemical changes reflected the known evolu-
tion of brain degeneration after asphyxial
injury, particularly the massive decrease of corti-
cal N-acetylaspartate in the acute phase,
suggesting the severity of the neuronal damage,
and the subsequent progressive increase of cho-
line related to the delayed degeneration of white
matter.
871
To our best knowledge, no other study on
application of MRS in human brain death is
available.
Functional Imaging of Brain Death
The future perspectives of imaging of the brain in
comatose patients may be linked to functional
MRI (fMRI) with blood oxygen level-dependent
(BOLD) signal. In 2009 Boly et al. [13] first
reported that fMRI showed no significant func-
tional connectivity in a brain death patient in
contrary to a vegetative state patient with pre-
served cortico-cortical connectivity. In 2010
Monti et al. [14] successfully elaborated new
methodology, which can be applied in fMRI in a
patient without conscience. With imagery tasks
they were able to establish functional and interac-
tive communication with a minority of patients
who met the behavioral criteria for a vegetative
state after serious brain injury, detecting residual
cognitive function in them. For this reason fMRI
probably will provide a method for detecting min-
imal cognitive ability in patients without a behav-
ioral response and could be considered a
complementary tool for existing diagnostic
methods. Hence, it can be believed that fMRI will
increase significantly our understanding of con-
science disorders in patients with severe brain dam-
age and differentiate with the highest accuracy
brain death from a vegetative state or chronic coma.
From the theoretical point of view, MRI might
seem an ideal confirmatory test in brain-dead
patients because it is able to prove the cessation
of cerebral blood flow (MRA) and to demonstrate
the irreversible loss of cerebral bioelectrical activ-
ity (fMRI). However, practical application of MRI
in these patients is limited because transportation
and monitoring of ventilated patients to an MRI
scanner can be challenging, and the procedure is
time-consuming. On the other hand, in chosen
cases MRI may add certainty to the diagnosis of
brain death in addition to clinical findings. Further
extensive studies are needed to establish a poten-
tial role of MRI techniques, including MRA,
DWI, MRS, and fMRI, in the diagnostic algo-
rithm for brain death.
872
Imaging of Cerebral Circulatory Arrest
Prerequisites for Blood Flow Studies
in the Diagnosis of Brain Death
It is recommended that blood flow studies be
performed when it is impossible to complete the
clinical criteria defined in Table 1. Before these
tests are performed, at least two particular clinical
criteria must be met:
1. An established etiology capable of causing
neurological death in the absence of reversible
conditions capable of mimicking neurological
death
2. Deep unresponsive coma
Prior to the testing, unresuscitated shock and
hypothermia must be corrected.
The blood flow studies are not confounded by
drugs or metabolic disorders, so they are reliable
in cases of barbiturate overdose.
Catheter Cerebral Angiography
Catheter cerebral angiography is considered as a
reference method in the diagnosis of majority of
cerebrovascular pathologies including cerebral
circulatory arrest. For the diagnosis of brain death,
it was used as the first imaging modality in 1953.
The study involves a follow-up of iodinated
contrast dynamics after its intravascular injection
on time-resolved images made with the usage of
the digital subtraction technique (DSA).
Catheter Cerebral Angiography Findings
in Intracranial Hypertension and Brain
Death
At raising ICP the cerebral arteries and veins fill
with contrast with increasing delay in comparison
to extracranial vessels of the head. At the early
stage of cerebral circulatory arrest, nonfilling of
the cortical arterial branches and the deep cerebral
veins (the internal cerebral vein, ICV; the great
cerebral vein, GCV; and the straight sinus) occurs.
At this stage an oscillation of contrast column in
the proximal intracranial arteries may be detected.
M. Sawicki et al.
The pendulum movement of the contrast is
accompanied by its slow antegrade propagation
and intracranial stasis. The contrast is gradually
eliminated from the intracranial vessels through
the arteriovenous shunts and the interrupted
blood-brain barrier. Noteworthy the superior sag-
ittal sinus and the transverse sinuses may fill from
the extracranial vessels through the emissary
veins. With further increasing ICP, the contrast
column is stopped in more proximal arterial seg-
ments. A total intracranial nonfilling occurs when
the ICP reaches systolic blood pressure.
Technique of Catheter Cerebral
Angiography in the Diagnosis of Brain
Death
Catheter cerebral angiography for the determination
of brain death is performed with an injection of
iodinated contrast material to reach both anterior
and posterior cerebral circulation. The propagation
of contrast is assessed in a filming sequence acquired
with the usage of DSA technique. The duration of
the acquisition varies among national guidelines, but
it should not be shorter than 15 s to confirm cerebral
circulatory arrest (personal experience).
The national guidelines permit two different
techniques of contrast injection:
1. Aortocervical arteriography – injection from
the aortic arch
2. Selective 4-vessel cerebral angiography,
performed with a selective catheterization of
both ICAs or the common carotid arteries
(CCAs) and VAs
While performing the selective angiography,
one must take into account that the contrast
injected under high pressure into the ICA or VA
may be pushed forward to intracranial arteries
resulting in an artifactual intracranial filling. This
can be avoided if the injection from the aortic arch
is applied. Moreover, the selective technique
requires higher manual skills and experience
than aortic arch angiography. Finally, the risk of
procedure-related complications is higher for
selective technique. Because both techniques are
characterized by equally high accuracy, the aortic
arch angiography is a preferred option.
37 Brain Death Imaging
Fig. 4 Catheter angiography findings in a 74-year-old
woman after traumatic brain injury who presented with
clinical signs of brain death. Aortic arch injection of
30 ml of contrast at flow rate of 15 ml/s. (a) 3 s after
injection branches of the external carotid arteries are filled
(empty arrows), while filling of the internal carotid arteries
(arrowheads) and the vertebral arteries (arrows) is limited
to their proximal extracranial parts; (b) 10 s after injection
Intravenous angiography can be applied in
the determination of cerebral circulatory
arrest as well [15]. However, when performing
the exam with i.v. injection, one must consider
the jugular and dural venous sinus reflux of
contrast. Such finding is not indicative of pre-
served cerebral circulation [15]. Another disad-
vantage of this technique is a low concentration
of contrast achieved in the head and neck vessels
due to the effect of dilution. This could make the
detection of possible intracranial opacification
equivocal. Therefore arterial injection is
preferred.
Criteria for the Diagnosis of Brain Death
in Catheter Cerebral Angiography
Cerebral circulatory arrest can be documented by
catheter cerebral angiography if all of the follow-
ing conditions are met:
873
slow propagation of contrast is visible in extracranial parts
of the ICAs (arrowheads) and the VAs (arrows); (c) 15 s
after injection contrast reaches cavernous segments of the
ICAs (arrowheads). Venous outflow through the external
jugular veins is observed (arrows). No intracranial filling is
noted. These findings are consistent with the diagnosis of
brain death
1. Filling of external carotid arteries confirming a
proper contrast delivery
2. No filling of ICAs beyond the level of the
anterior clinoid process
3. No filling of VAs beyond their dural
penetration
4. No filling of the internal cerebral veins [16]
(see Fig. 4)
However, delayed opacification of proximal
segments of the cerebral arteries with contrast
stasis and without deep venous filling does not
preclude the diagnosis of brain death – see Fig. 5.
Usefulness of Catheter Cerebral
Angiography in the Diagnosis of Brain
Death
Catheter angiography is widely accepted as a
reference test in the diagnosis of brain death.
874
Fig. 5 Catheter angiography findings in an 18-year-old
man after traumatic brain injury who presented with clin-
ical signs of brain death. Selective injection of 8 ml of
contrast at flow rate of 6 ml/s to the left common carotid
artery. (a) 3 s after injection branches of the external carotid
artery are filled (arrow), while filling of the ICA
Despite this several disadvantages limit its applica-
tion. The catheterization of aortic arch, especially
ICAs and VAs, requires high expertise. The exam
must be performed in the angiographic lab; thus,
transportation of a critically ill patient is necessary.
The procedure is invasive, accompanied by the risk
of serious complications, such as an injury of a
catheterized vessel or vasospasm. Besides catheter
angiography is the most costly procedure among
the blood flow tests. Potential risk of a damage of
transplantable organs caused by iodinated contrast,
although postulated, has not been confirmed yet.
Moreover, it was revealed that contrast medium
administration to donors does not affect kidney
graft function after transplant [17]. No false-
positive results of catheter angiography in the diag-
nosis of brain death have been reported (i.e., angi-
ography revealed cerebral circulatory arrest, yet the
patient survived), which means the 100 % specific-
ity. The sensitivity is close to 100 % as well. There
are casuistic reports of false-negative results in
cases in which residual preservation or restoration
of intracranial blood flow was observed after ICP
was relieved through some decompressive mecha-
nism, e.g., craniectomy or skull fractures (see
“Limitations” section) [15]. There are several
M. Sawicki et al.
(arrowhead) is limited to its proximal extracranial part;
(b) 12 s after injection stasis of contrast is visible in the
M1 segment of the middle cerebral artery (arrowhead) and
A1 and A2 segments of the anterior cerebral artery (arrow).
No intracranial venous outflow is noted. These findings are
consistent with the diagnosis of brain death
reports of preserved blood flow in the posterior
fossa with cerebral circulatory arrest in the anterior
circulation [15, 18]. This points to the protective
function of the cerebellar tentorium, which causes
an uneven increase in ICP. In all false-negative
exams, patients met all clinical criteria of brain
death, and in all cases intracranial circulation was
markedly slowed.
CT Angiography
CTA for the determination of brain death was
proposed in 1998 [19]. Despite the fact that tech-
nical protocol or evaluation criteria have not been
standardized yet, CTA is considered the most
promising ancillary test.
CT Angiography Findings in Brain Death
Unfortunately CTA findings were not studied in
non-brain-dead patients with intracranial hyper-
tension. The earliest sign of cerebral circulatory
arrest in CTA is a lack of opacification of deep
veins – the ICVs and the GCV. The sensitivity of
this finding in the diagnosis of brain death in CTA
is 98–100 % [19–23]. A lack of opacification of
37 Brain Death Imaging
Fig. 6 CTA findings in a 50-year-old man with traumatic
brain injury and right-sided craniectomy presented with
signs of brain death: 10 mm MIP in coronal plane (a) and
VRT (b) with a delay of 60 s after injection show
opacification of the basilar artery (arrows). No filling of
the anterior cerebral circulation or deep venous outflow is
Fig. 7 CTA findings in a 22-year-old woman with a brain
stem ischemic stroke and a right-sided craniectomy who
presented with signs of brain death on clinical examina-
tion. (a) 10 mm MIP in coronal plane with a delay of 60 s
after injection shows opacification of M1 and M2 segments
of the middle cerebral arteries (arrowheads) and A1 seg-
ments of the anterior cerebral arteries (arrow); (b) sagittal
cortical branches of the middle cerebral arteries
(MCA-M4) is slightly less sensitive indicator of
brain death with the sensitivity of 86–100 % [19,
21, 22]. The basilar artery (BA) and cortical
branches of the posterior cerebral arteries
(PCA-P2) are more frequently opacified in brain-
dead patients than MCA-M4 – see Fig. 6. Their
875
noted. Opacification of the superficial temporal artery
(empty arrow) indicates that the contrast was injected
properly. In some cases of brain death, the stasis of contrast
in the posterior circulation may be observed due to the
protective function of cerebellar tentorium
VRT shows opacification of the right pericallosal artery
(arrow). No filling of the basilar artery or deep venous
outflow is noted. In some cases of brain death, the stasis
of contrast in the pericallosal arteries may be observed
because of higher CPP in these arteries in comparison to
cortical branches of the middle cerebral arteries
sensitivity is 85–94 % and 79 %, respectively [19,
21]. The least sensitive finding of brain death is a
lack of opacification of cortical branches of the
anterior cerebral arteries (ACA-A3, the
pericallosal arteries) with the sensitivity of 64 %
[21] – see Fig. 7. Such a sequence is explained by
the highest susceptibility of cortical branches of
876
Fig. 8 CTA findings consistent with the diagnosis of brain
death: (a) VRT 20 s after contrast injection shows filling of
the ICAs (arrows) and the VAs (arrowheads) limited to
their proximal extracranial parts. Opacification of extracra-
nial arteries (empty arrows) indicates that the contrast was
injected properly; (b) VRT 60 s after contrast injection
the MCA to intracranial hypertension and the
lowest CPP in these arteries. In the course of
cerebral circulatory arrest, proximal segments of
the cerebral arteries frequently show delayed
opacification for some time. These vessels appear
as thin and faint. The intracranial stasis of contrast
is observed with its extremely slow elimination
through the arteriovenous shunts and extravasa-
tion due to the interrupted blood-brain barrier. The
latest sign of cerebral circulatory arrest is intracra-
nial nonfilling – see Fig. 8.
Technique of CT Angiography
in the Diagnosis of Brain Death
The technique of CTA involves a rapid, intrave-
nous administration of iodinated contrast in the
bolus at a flow rate of 3–5 ml/s. Volume of con-
trast medium used for CTA for the diagnosis of
cerebral circulatory arrest is 60–120 ml. The con-
trast can be followed by the infusion of 30–40 ml
of normal saline at the same rate (i.e., saline flush)
to push the contrast medium forward and to opti-
mize contrast enhancement.
For the diagnosis of brain death, at least three
acquisitions should be performed:
M. Sawicki et al.
shows slow propagation of contrast to the level of petrous
segment of the right ICA and cavernous segment of the left
ICA (arrows). Vertebral arteries are opacified to the level
of foramen magnum, before their dural penetration (arrow-
heads). Venous outflow through external jugular veins is
observed (empty arrows). No intracranial filling was noted
1. Non-enhanced scanning as a reference for
assessing vascular opacification.
2. Early post-contrast scanning for assessing
intra- and, more important, extracranial vascu-
lar opacification. This phase is started approx.
20 s after the beginning of injection.
Opacification of branches of external carotid
arteries – the superficial temporal or the facial
arteries – indicates that the contrast was
injected correctly, and there are no hemody-
namic abnormalities causing a delay of con-
trast delivery to the vessels of head and neck.
To ensure scanning is performed during the
phase of optimal contrast enhancement, sev-
eral strategies can be applied. A fixed delay is
the easiest but least favorable option since it
does not adjust for individual variations in
contrast dynamics. Two more advanced strate-
gies, test bolus injection and bolus tracking
technique, can be applied to tailor the contrast
injection to the individual patient.
3. Late post-contrast scanning for assessing intra-
cranial vascular opacification. This phase is
started 60 s after the beginning of injection,
with a delay of 40 s to the early phase. The
37 Brain Death Imaging 877

necessity of performing the late phase of CTA Table 4 Criteria for the diagnosis of brain death in CT
in the diagnosis of cerebral circulatory arrest angiography
is motivated by a possible delayed Sensitivity
intracranial opacification in increased ICP. Ter- Criteria Lack of opacification of (%)
mination of the study on the early phase may Intracranial ICA beyond the level of 86–92a
nonfilling the anterior clinoid
result in a failure to recognize delayed intra- [24, 25] process
cranial filling producing a false-positive result. VA beyond their dural
penetration
Intermediate post-contrast scans may be added ICV, GCV, and the
between the early and the late phase for a better straight sinus
follow-up of the contrast dynamics. 10 point [26] BA 54–70
Right and left PCA-P2
Practical Tips Right and left ACA-A3
(pericallosal artery)
The superficial temporal artery (STA) is routinely
Right and left MCA-M4
used for assessing sufficient contrast delivery. But Right and left ICV
in some cases, both STAs may be surgically GCV
ligated during bilateral temporal craniectomy. In 7 point Right and left ACA-A3 52–100
such cases the facial artery can be used as a good [19–21, 27] (pericallosal artery)
alternative. Right and left MCA-M4
In patients with suspected brain death, CT Right and left ICV
images often show subarachnoid hemorrhage GCV
(SAH) or high-attenuation areas along the basal 4 point [21, Right and left 86–97
cisterns and cortical sulci, mimicking SAH. The 22] MCA-M4b
Right and left ICV
latter is described as pseudo-SAH phenomenon,
which is a synergistic result of distention of In the 10-, 7-, and 4-point scale, one point is noted for each
nonopacified vessel in the late phase. Cerebral circulatory
congested superficial veins due to elevated intra- arrest is diagnosed with the score 10, 7, or 4 points
cranial pressure and severe brain edema. Both true accordingly
a
SAH and pseudo-SAH impede the assessment of Based on the opacification in the early phase
b
intracranial filling considering that cerebral ves- According to the 4-point scale, opacification of 1 or 2 cor-
tical branches of the MCA on the same side does not
sels in intracranial hypertension are usually abnor- exclude the diagnosis of cerebral circulatory arrest, pro-
mally thin and faintly opacified. Therefore vided a lack of opacification of the ICVs
cerebral vessels should be assessed on pre- and
post-contrast scans.
Another potential difficulty associated with Usefulness of CT Angiography
CTA assessment is a consequence of frequently in the Diagnosis of Brain Death
found midline shift and marked ventricular dis- CTA is far more widely available than any other
placement. This may cause problems with identi- blood flow test. The test is noninvasive, techni-
fication of the dislocated deep cerebral veins on cally uncomplicated, and non-time-consuming.
routinely used axial images. These vessels can be However, CTA cannot be applied at the bedside,
far easier localized on sagittal maximum intensity while a transport of the ICU patient is always
projections (MIPs). insecure. Potential risk of a damage of transplant-
able organs caused by iodinated contrast was
Criteria for the Diagnosis of Brain Death addressed above. The accuracy of CTA in the
in CT Angiography diagnosis of brain death has not been reliably
There are no widely accepted criteria of CTA in studied yet. In particular, specificity of the method
the diagnosis of brain death. Four methods of has not been evaluated with non-brain-dead
evaluation have been proposed so far; they are patients with intracranial hypertension as
summarized in Table 4 and Fig. 9. controls. However, there are no reports of the
878
Fig. 9 Evaluation criteria of CTA in the diagnosis of
cerebral circulatory arrest. A point is given for each vessel
without opacification: (a) positive result in the 10-point
scale (score = 10) confirming the diagnosis of brain death
is recorded when the following vessels are not opacified,
the bilateral PCA-P2, the BA, the bilateral ACA-A3, the
bilateral MCA-M4, the bilateral ICV, and the GCV. Scores
from 0 to 9 are classified as negative results excluding the
false-positive CTAs involving a multiphase scan-
ning. Skull decompression, primary infratentorial
process, and hypoxic-ischemic injury after car-
diac arrest are predisposing factors to false-
negative results of CTA (see “Limitations” sec-
tion) [21, 25, 27, 28].
Perfusion CT
PCT is commonly used to detect brain ischemia in
different clinical conditions, i.e., stroke, TIA, vas-
culitis, vasospasm, carotid stenosis, and traumatic
M. Sawicki et al.
diagnosis of brain death; (b) in the 7-point scale, positive
result (score = 7) is recorded with a lack of opacification of
the bilateral ACA-A3, the bilateral MCA-M4, the bilateral
ICV, and the GCV. Scores from 0 to 6 are classified as
negative results; (c) positive result in the 4-point scale
(score = 4) is recorded when the bilateral MCA-M4 and
the bilateral ICV are not opacified. Scores from 0 to 3 are
classified as negative results
injury. The diagnosis of brain death is a new
application of PCT [28, 29].
PCT studies are obtained by monitoring the
first pass of an iodinated contrast agent through
the cerebral vasculature. Changes in tissue atten-
uation that occur in the brain after the contrast
injection are measured. Post-processing of the
PCT data allows the generation of color-coded
maps of various perfusion parameters, including
cerebral blood flow (CBF), cerebral blood volume
(CBV), mean transit time (MTT), and the time to
peak (TTP) – the time from the start of contrast
injection to the time of maximum enhancement.
37 Brain Death Imaging
Perfusion CT Findings in Brain Death
The normal mean CBF is 45–60 ml/100 g/min.
This value is approx. 50 % higher for the gray
matter and 50 % lower for the white matter. The
cerebral autoregulation mechanisms maintain the
blood flow at almost constant level, provided the
CPP exceeds 50 mmHg. When an elevation of the
ICP is high enough to cause a decrease of CPP
below 50 mmHg, CBF gradually decreases. With
CBF 20 ml/100 g/min, disturbances of neuronal
electrical function develop. Neurological symp-
toms occur at this stage as a result of physiologic
paralysis of neurons and EEG becomes flatlined.
In such conditions near-normal concentrations of
extracellular potassium and tissue ATP are
detected that reflect preservation of energy metab-
olism and ion pumping thereby viability of neu-
rons. This state of ischemia is potentially
reversible if the normal blood flow is restored
within several hours. CBF 10 ml/100 g/min
causes a failure of energy metabolism and ion
pumping represented by an increase of extracel-
lular potassium and diminishing of tissue ATP
concentration. This is consistent with cerebral
circulatory arrest leading to necrosis of neurons
within several minutes. It should be emphasized
that there are significant differences in CBF
between the gray and white matter as well as
differences in vulnerability to ischemia between
individual brain structures.
Technique of Perfusion CT
in the Diagnosis of Brain Death
Several technical issues need to be resolved when
performing PCT for the diagnosis of brain death:
Arterial input function (AIF) selection
To ensure the accuracy of results, one needs
to select an artery with early and high contrast
enhancement. This may be impossible in some
brain-dead patients without detectable intracra-
nial filling. Wrong selection of AIF may result
in an overestimation of MTT and an underes-
timation of CBF.
Venous output function (VOF) selection
AIF is normalized with respect to the
venous output function. The venous region of
interest (ROI) is usually placed over the
879
superior sagittal sinus or torcular Herophili.
In some brain-dead patients, filling of the intra-
cranial veins may be insufficient to provide a
proper VOF. Poor placement of venous out-
flow region can result in an overestimation of
CBV and CBF.
Delay of arterial inflow and venous outflow
PCT protocols are usually designed for
acquisition duration of 40–45 s, principally to
minimize radiation dose. However, in brain-
dead patients, intracranial time-attenuation
curve (TAC), if detectable, is flat and low. In
such cases the standard series would not cover
the entire TAC. This may preclude calculation
of diagnostic PCT maps.
Scan range
In the diagnosis of cerebral circulatory
arrest, the scanned volume must include the
whole brain. Standard CT scanners (16–64
detector rows) cover 2–9,6 cm in z-axis,
which is not enough for the diagnosis of
whole-brain death. Such a limitation does not
apply to cutting-edge scanners with 128–320
detector rows offering the possibility of
9,6–16 cm coverage with acceptable temporal
resolution.
Criteria for the Diagnosis of Brain Death
in Perfusion CT
The threshold values of PCT parameters for brain
infarct are well known. It is widely accepted that
CBF <10 ml/100 g/min, CBV <2 ml/100 g, and
MTT >8 s are insufficient to preserve metabolic
processes of neurons and lead to neuronal necrosis
if maintained above 5–8 min – see Fig. 10. How-
ever, the PCT diagnostic criteria for cerebral cir-
culatory arrest have not been established yet. An
extrapolation of the values for stroke on the deter-
mination of brain death may not be appropriate
because of the technical issues listed above.
Usefulness of Perfusion CT
in the Diagnosis of Brain Death
The major advantage of PCT is the ability to
quantify a capillary blood flow, which is a key
factor in the diagnosis of brain death. Angio-
graphic tests and TCD are capable to show the
blood flow in large and medium vessels, while it
880
Fig. 10 Perfusion CT findings in a patient presenting
clinical signs of brain death: (a) CBF map in axial plane
and (b) CBF map in sagittal plane shows blood flow
limited to torcular Herophili (arrowhead), sigmoid sinus
does not always correspond with true blood sup-
ply to the brain parenchyma. The cerebral arterio-
venous shunts and the phenomenon of ineffective
reperfusion in hypoxic-ischemic brain injury after
cardiac arrest are known examples of such dis-
crepancy. The test is ubiquitous, non-time-con-
suming, not costly, and noninvasive.
On the other hand, the study cannot be
performed at the bedside, and the unsafe transport
of the ICU patient is necessary. Besides this, sev-
eral principal technical issues (mentioned above)
must be resolved before the wide implementation
of PCT as an ancillary test. The post-processing of
PCT data is not as simple as in angiographic
methods and requires expertise in interpretation
of perfusion studies. Finally, the test is not stan-
dardized, and there are no reliable case-control
studies assessing its diagnostic value.
Xenon CT
Xenon (Xe) is an inert gas that is soluble in both
water and lipid. X-ray attenuation by Xe is similar
to that of iodine, so Xe can be used as a contrast
agent in conjunction with CT (XeCT). The gas
dissolves rapidly in blood after inhalation and
M. Sawicki et al.
(arrow), and extracranial part of the ICA, proximal to the
petrous part (empty arrow). Reduction of blood flow below
10 ml/100 g/min is noted in an entire volume of the brain
consistent with neuronal necrosis
diffuses rapidly into the brain. Diffusion into the
brain is determined by blood flow and solubility
of the Xe within different brain compartments.
Patients inhale a mixture of 26–32 % Xe mixed
with O2 over a period of approximately 4 min.
Processing of the data allows generation of CBF
maps, from which quantitative data can be
extracted.
It is known that CBF values less than
10 ml/100 g/min correlate with eventual neuronal
necrosis. CBF between 10 and 20 ml/100 g/min
has been associated with potentially reversible
neurological deficits. It was shown that CBF <5
ml/100 g/min is required in order to confirm the
diagnosis of brain death [30, 31]. However, there
is insufficient evidence to determine the accuracy
of XeCT as an ancillary test in the determination
of brain death.
The method is fast and easy to perform and
simple in interpretation, and the accessory
equipment for commercial CT scanner is inexpen-
sive. A minor drawback is that XeCT quantifies
only a single parameter – CBF. A major disadvan-
tage is the lack of standardization and
unavailability of xenon for medical applications.
The test is available only in few centers around the
world.
37 Brain Death Imaging
MR Angiography
Magnetic resonance imaging of vessels (MRA,
MR angiography) can be performed with several
techniques, from which time-of-flight (TOF)
imaging is widely accepted as a technique for
the examination of the intracranial circulation. It
is based on the flow-related signal enhancement of
the blood entering into the volume of interest.
TOF is completely a noninvasive method because
there is no need to administer contrast agents.
Other advantages include high spatial resolution,
relative short imaging time, and high signal-to-
noise ratio. TOF imaging, however, has some
technical limitations, e.g., reduced sensitivity to
slow blood flow as well as saturation of the signal
when vessels run parallel to the imaging slice
orientation, what subsequently makes the blood
flow undetectable, even when it is not severely
impaired.
On MR angiograms in brain-dead patients, a
unique finding is the absence of any cerebral
blood flow above the level of the supraclinoid
portions of the ICAs and in the vertebrobasilar
arteries [4, 5].
Karantanas et al. [5], who performed also MR
venography, demonstrated lack of visualization of
veins and sinuses in the most but not all patients
with defined brain death. On the other hand, veins
and sinuses were visualized in all MR venograms
performed in comatose patients in whom the
follow-up did not reveal brain death in the next
12 months.
Luchtmann et al. [6] also documented that
another technique, contrast-enhanced angio-MR,
may be considered conclusive in detecting of ces-
sation of cerebral blood flow due to its extremely
high sensitivity and specificity.
Transcranial Doppler Sonography
Transcranial Doppler in Increased
Intracranial Pressure
Extensive death of brain tissue causes extreme
increase of ICP. When the ICP equals the diastolic
arterial pressure, the brain is perfused only in
systole, and with further increase of ICP over the
881
systolic arterial pressure, cerebral perfusion will
cease. Due to elasticity of the arterial wall and the
compliance of the vasculature distal to the record-
ing site, such cerebral circulatory arrest is associ-
ated with Doppler evidence of oscillatory
movement of blood in the large arteries at the
base of the brain. However, the net forward flow
volume is zero. With time the oscillations
decrease in amplitude of spectral spikes until no
pulsations are detectable. This development cor-
relates with a more proximal demonstration of the
angiographic flow arrest [32]. At the time of
angiographic flow arrest at the internal carotid
artery (ICA), the TCD shows an oscillating flow
pattern in the middle cerebral artery, because the
contrast medium progresses slowly toward the
brain. From the clinical experience from cardiac
arrest, such cerebral ischemia of about 10–15 min
in vivo at normal body temperature leads to irre-
versible total loss of brain function [33].
Transcranial Doppler Findings
in Cerebral Circulatory Arrest
1. Increasing pulsatility of the velocity wave-
form: if the velocity at the end of diastole is
zero, then the ICP has reached the diastolic
blood pressure. Forward flow persists in sys-
tole. Therefore, this stage does not correspond
to cerebral circulatory arrest.
2. Oscillating flow: cessation of cerebral perfu-
sion has been reached when forward and
reverse flows are nearly equal (biphasic, oscil-
lating, and net zero flow, i.e., the ICP equals or
exceeds the systolic blood pressure). Equality
of both flow components is reached if the area
under the envelope of the positive and negative
deflection is the same. This finding correlates
with the angiographic appearance of cerebral
circulatory arrest [34].
3. Systolic spikes: with further reduction of the
blood movement, only a very short velocity
peak can be seen (systolic spike). The systolic
spike pattern is highly characteristic for cere-
bral circulatory arrest. In this stage it can be
assumed that the slower and longer-lasting
reverse flow component (see oscillating flow)
is obscured due to high-pass filtering. All
Doppler devices use high-pass filters in order
882
Fig. 11 Progressive deterioration of TCD waveform observed during changes leading to cerebral circulatory arrest
(Modified from Hassler et al. [32]); DBP diastolic blood pressure, VB vascular bed
to eliminate artifacts from wall movements.
For the diagnosis of cerebral circulatory arrest,
this filter must be set at its lowest level (e.g.,
50 Hz).
4. No flow signals: as the intracranial pressure
increases further and the flow obstruction
becomes more proximal, no flow signals in
the basal cerebral arteries can be detected. Fail-
ure to detect flow signals can also result from
ultrasonic transmission problems. In such a
case the extracranial findings, including those
of the vertebral arteries (VA), represent an
important criterion for the diagnosis, or one
must confirm the disappearance of the flow
signal (the baseline examination before cere-
bral circulatory arrest with sufficient bone win-
dow must be available) [32, 33, 35] – see
Fig. 11.
Clinical Prerequisites of Transcranial
Doppler
Besides the prerequisites that apply to all
other blood flow tests, there are some special
clinical conditions for Doppler examination
[33, 36–38]:
M. Sawicki et al.
1. Tachycardia must not exceed 120/min.
2. Systolic blood pressure must not be less than
90 mmHg (in some countries mean arterial
blood pressure must not be less than 80 mmHg).
3. The patient must be normocapnic (PCO2 must
not be below the normal range).
These conditions may vary in some minor
details in national guidelines; however, they are
justified by influence on a Doppler spectrum
shape.
The technical prerequisites of TCD in cerebral
circulatory arrest include [38]:
1. The examination must be performed by highly
skilled certified examiner.
2. High-pass filter should be 50 Hz.
3. It is advised that gain be increased.
4. It is advised that sample volume be set
15 mm [38] (in guidelines in some countries,
sample volume should be lower than 10 mm
[36, 37], and in the opinion of the authors, the
sample volume depends on the device used
[transcranial Doppler or transcranial duplex]
and the bone window penetrability).
37 Brain Death Imaging
Fig. 12 Systolic spikes in
the middle cerebral artery,
insonation depth 55 mm.
Spikes’ duration is
<200 ms and velocity <50
cm/s
5. Envelope should be switched off.
6. A portable, 2 MHz, pulsed-wave Doppler
ultrasonographic instrument is preferred [34,
39]; however, in some countries transcranial
duplex scanner is permitted [36, 37].
Criteria of Sonographic Findings
to Confirm Cerebral Circulatory Arrest
According to the World Federation
of Neurology [33]
Cerebral circulatory arrest can be confirmed if the
following extra- and intracranial Doppler sono-
graphic findings have been recorded and
documented both intra- and extracranially and
bilaterally on two examinations at an interval of
at least 30 min.
1. Systolic spikes or oscillating flow in any cere-
bral artery which can be recorded by bilateral
transcranial insonation of the ICA and MCA,
respectively, and any branch or other artery
which can be recorded (anterior and posterior
circulation) – Fig. 12.
2. Oscillating flow is defined by signals with
forward and reverse flow components in one
cardiac cycle exhibiting almost the same area
under the envelope of the waveform (to-and-fro
movement) [33]. Some German guidelines
permit the velocity of the retrograde (diastolic)
component of biphasic (oscillating) flow to not
be greater than one third of the antegrade
(systolic) velocity of the flow [38] – Fig. 13.
883
3. Systolic spikes are sharp unidirectional veloc-
ity signals in early systole of less than 200 ms
duration and less than 50 cm/s peak systolic
velocity and without a flow signal during the
remaining cardiac cycle. Transitory patterns
between oscillating flow and systolic spikes
may be seen [33].
4. The diagnosis established by the intracranial
examination must be confirmed by the extra-
cranial bilateral recording of the common
carotid artery (CCA), ICA, and vertebral artery
(VA); they are allowed to be performed by
duplex device – Fig. 14.
5. The lack of a signal during transcranial
insonation of the basal cerebral arteries is not
a reliable finding because this can be due to
transmission problems. But the disappearance
of intracranial flow signals in conjunction with
typical extracranial signals can be accepted as
proof of circulatory arrest [33, 40].
6. Ventricular drains or large openings of the skull
like decompressive craniectomy possibly
interfering with the development of the ICP
are not present.
Once reverberating flow is identified, it should
be confirmed in the both MCAs and the BA and
monitored for 30 min to exclude effects of tran-
sient intracranial pressure increase. In such a case
the repetition of TCD examination on monitored
artery after 30 min is not necessary [40].
884 M. Sawicki et al.

Fig. 13 Oscillating flow in

the middle cerebral artery.
Insonation depth 55 mm.
All diastolic flow is in
opposite direction to
systolic antegrade flow

Fig. 14 Biphasic
(oscillating) flow observed
in extracranial segment of
the internal carotid artery.
Usually it overlaps partly
with spectrum from the
external carotid artery
(4 MHz CW probe)

Practical Tips Some kind of oscillating flow can be seen in

Middle cerebral artery examination should be
performed in depth 55–65 mm, avoiding internal
carotid artery bifurcation area, where overlapping
spectra of middle cerebral and anterior cerebral
arteries can be observed. It is especially important
because oscillating flow in both arteries can
mimic normal bifurcation two-directional flow
[38] – Fig. 15.
It is common that systolic spikes can vary in
velocity with respiratory action [41] – Fig. 16.
Sometimes triphasic flow can be seen, with
small early diastole antegrade flow. It is not con-
sidered to be a sign of irreversible circulatory
arrest – Fig. 17.
the aorta or common carotid artery dissections or
in severe aortal valve insufficiency. It can be eas-
ily distinguished in all branches of the aorta or
common carotid artery by longer duration of the
systolic complex than in systolic spikes [36, 38].
Usefulness of TCD in Cerebral Circulatory
Arrest Confirmation
TCD can be used as a screening tool to determine
the timing for other confirmatory tests such as
nuclear brain scanning or angiography. It is of
particular value in patients with high barbiturate
levels when clinical examination and electroen-
cephalography are unreliable [39, 42]. A variety
37 Brain Death Imaging 885

Fig. 15 Bifurcation of the

internal carotid artery at the
depth of 65 mm. This area
should be avoided in
confirmation of cerebral
circulatory arrest, because
overlapping spectra with
oscillating flow in the
middle cerebral artery and
the anterior cerebral artery
can mimic “normal flow”

Fig. 16 Systolic spikes

often change velocity with
respiratory movements. The
middle cerebral artery.
Insonation depth 55 mm

Fig. 17 Triphasic flow is

not consistent with cerebral
circulatory arrest. It is often
observed in patients with
large skull defects. The
middle cerebral artery.
Insonation depth 55 mm
886
of clinical conditions may lead to brain death.
TCD is indicated when the clinical examination
indicates brain death or is unreliable. The test may
determine positive end-diastolic flow and rule out
cerebral circulatory arrest or may demonstrate
reverberating flow and confirm clinical findings.
It is extremely important to document arterial
blood pressure during TCD examination to rule
out transient arrest of cerebral circulation due to
hypotension [36, 40].
Patients at risk of brain death may need base-
line TCD examination before clinical changes
become apparent. A baseline TCD will help to
establish the presence of temporal windows and
end-diastolic flow. Subsequent TCD studies can
be performed hours or days later with greater
confidence, and intracranial flow changes will be
easier to interpret. For example, absent
end-diastolic flow in both MCAs and the BA in
a patient with clinical brain death may
indicate incomplete arrest (extremely high resis-
tance to flow with some residual brain perfusion)
or complete arrest (vessel wall distention
during systole with no brain perfusion). Although
velocities may be minimal with the latter,
confusion may arise if temporal windows were
not established beforehand. In this situation, it is
prudent to wait and repeat TCD 1–2 h later. Pro-
gression to arrest will manifest as reverberating
flow, and a complete arrest will make identifica-
tion of any signals more difficult (asonic
arteries) [40].
TCD can effectively shorten the time to con-
firm the clinical diagnosis of brain death that is of
great importance when organ donation is planned
[36]. It is a preferable test due to noninvasiveness
and portability.
Main disadvantages of TCD include lack of
sufficient bone window that occurs in about
10 % patients, confounding noises in intensive
care unit setting during bedside examination, dif-
ficulties with keeping parameters of patient’s
blood pressure and heart rate in sufficient levels
to perform TCD examination, and lack of well-
trained, experienced examiners [36, 38, 42].
Although the sensitivity of TCD for the diag-
nosis of brain death is 91.2–100 %, the specificity
M. Sawicki et al.
is reported to be virtually 100 % [35, 43,
44]. There have been several case reports of
patients with demonstrable CBF on TCD who
were clinically brain-dead. In these cases the
pathology was limited to the cerebellum or brain
stem leaving blood flow in the anterior cerebral
circulation relatively intact. The cases of a small
reperfusion in anterior cerebral circulation after
hypoxic-ischemic injury due to cardiac arrest
were reported after some days. For this reason it
is important that patients show clinical signs of
brain death before performing TCD [35], personal
experience]. The same is true in patients with
great skull defects, in whom the preservation of
orthograde diastolic flow can persist, despite of
fulfilling all clinical criteria of brain death
diagnosis [33].
There are no reports in the available literature
of a child or adult patient “surviving” who dem-
onstrated bilateral signals of oscillating flow or
systolic spikes in the MCA and ICA for at least
half an hour [33].
In the meta-analysis [44] including ten studies
of the usage of TCD in brain death, two false-
positive results were reported, but in both patients
brain stem function did show brain death shortly
thereafter. In the first case respiratory movements
in response to an apnea test persisted for some
hours after the first TCD signals of cerebral circu-
latory arrest were detected. The second clinically
brain-dead patient, with cerebral circulatory arrest
by TCD in the basilar artery and the circle of
Willis, confirmed by angiography, became iso-
electric only several hours later.
Nuclear Medicine Methods
in the Diagnosis of Brain Death
Brain scintigraphy is one of the reference methods
among others used for the diagnosis of brain
death. Dynamic brain scintigraphy with non-
brain-binding agent Tc99m-DTPA has been used
for a long time. After bolus intravenous injection
of the tracer brain vascular flow is estimated.
DTPA does not cross the blood-brain intact barrier
so intracranial blood flow is seen in normal
37 Brain Death Imaging
Fig. 18 Planar brain
Tc99m-HMPAO images.
Absence of tracer
accumulation in the brain
tissue consistent with brain
death
patients. The absence of intracranial blood flow is
considered brain death. The disadvantage of this
tracer is that it provides rather low-resolution
brain vascular flow [45].
Development of new radiopharmaceuticals
Tc99m-HMPAO (hexamethylpropyleneamine
oxime) and Tc99m-ECD (ethyl cysteinate dimer),
which are brain specific, gave new input to the role
of nuclear medicine methods in the determination
of brain death. Both radiopharmaceuticals are lipo-
philic agents and after intravenous injection cross
the blood-brain barrier and are accumulated pro-
portionally to the blood flow in normal gray matter
– brain cells of cerebrum, cerebellum, and brain
stem. So not only blood flow but brain parenchyma
is seen in normally functioning brain. If there is no
blood flow, there is no radiotracer accumulation in
the brain tissue (“hollow skull phenomenon”).
After tracer injection there is a very quick accumu-
lation of the tracer in a functioning brain. It makes
Tc99m-HMPAO/Tc99m-ECD brain study much
more reliable than investigations estimating only
blood flow. The injected radiopharmaceuticals are
safe, with no side effects to the patients, and do not
interact with medications used for patients’
treatment [46].
887
Tc99m-HMPAO/Tc99m-ECD brain study
should be performed according to the guidelines
[47, 48]. Proper labeling of the tracer and injec-
tion time according to manufacturer’s instruction
is crucial. Before injection quality control of the
tracer labeling is necessary. Yield of labeling
for Tc99m-HMPAO and for Tc99m-ECD should
be very high to avoid misinterpretation of the
study. The mode of acquisition can be planar or
SPECT (single-photon emission computed
tomography). Immediately after the tracer injec-
tion dynamic acquisition can be done, and it can
add additional information about brain blood
flow. Planar (anterior, posterior and lateral) scin-
tigraphy is enough to make the diagnosis, but
additional SPECT or SPECT/CT is preferable
[49] – see Figs. 18 and 19.
Sometimes delayed images are helpful to visu-
alize if brain blood flow is present or absent.
SPECT acquisition can give better visualization
of perfusion in posterior fossa and brain stem
structures. There can be different patterns of
radiotracer accumulation in the brain. According
to existing standards, the only pattern which
allows diagnosing brain death is lack of accumu-
lation in the cerebrum and cerebellum. Some
888 M. Sawicki et al.

Fig. 19 SPECT/CT brain Tc99m-HMPAO images of a patient with intracerebral hemorrhage with ventricular involve-
ment and subsequent brain death. CT component, upper row; SPECT, lower row

authors report that lack of tracer accumulation in physiologically accumulated in viable brain tissue
the cerebrum is enough to diagnose brain death. can be considered as excellent tracer for estima-
Positron emission tomography (PET) agent tion of brain viability. The quality and spatial
18-fluorodeoxyglucose (FDG) which is resolution of images are superior to SPECT but
37 Brain Death Imaging
lack of instant availability of the tracer, logistic
problems made it did not gain the widespread
use [49].
Brain perfusion scintigraphy in most patients
investigated due to suspicion of brain death gives
proper diagnosis; is reliable, safe, reproducible,
and noninvasive; and can be repeated in
inconclusive cases [24]. There was no false-
positive study reported. The study should be
performed in accredited nuclear medicine depart-
ment having experience in performing brain per-
fusion scintigraphy. Brain death scintigraphy is
helpful in patients considered as possible organ
donors to document the lack of blood flow. But
according to guidelines, the brain perfusion scin-
tigraphy alone is not sufficient to diagnose the
brain death even if there is absence of perfusion
[47]. It should be correlated with other findings
like clinical examination, cerebral angiography,
EEG, or others.
Limitations of Blood Flow Studies
in the Diagnosis of Brain Death
The limitations of blood flow ancillary tests can be
divided into the technical and clinical ones. The
technical limitations were discussed above.
In the early days of research on brain death,
there was a prevailing notion based on the autop-
sies that the clinical diagnosis of brain death is
associated with totally necrotic brain described as
the so-called respirator brain. However, recent
neuropathologic findings show widespread areas
of normal to near-normal neuronal areas in one
third to two thirds of patients who meet the clin-
ical criteria of brain death [50]. Thus, normally or
hypoperfused areas frequently exist in the brain
which lost its all functions and thereby is defined
as the dead brain. Therefore, total cerebral circu-
latory arrest is not a necessary condition for a loss
of all brain functions. Given that the blood flow
test must demonstrate global cerebral circulatory
arrest to confirm brain death, this goal may be
unattainable in some cases.
Furthermore, it is known that individual brain
structures show variable vulnerability to ische-
mia. This makes it difficult (presumably
889
impossible) to define uniform diagnostic criteria
of cerebral circulatory arrest for the whole brain.
There are known factors predisposing to disso-
ciation between brain death and total cessation of
cerebral circulation. Their common feature is that
they cause a disproportional increase of the ICP,
which in certain areas of the brain may be insuf-
ficient to cease the cerebral circulation. There are
several mechanisms causing a local decrease of
the ICP including craniectomy, ventricular drain-
age, skull fracture, open fontanels, and unfused
sutures in neonates and infants. The regionally
decreased ICP may result in preservation or res-
toration of residual cerebral blood flow at the site
of decompression, while in the remaining parts of
the brain, the cessation of blood flow is observed,
as presented in Fig. 20 [15, 25]. Similar problem
can be met in some cases of infratentorial injuries
(e.g., thrombosis of the BA or brain stem hemor-
rhage) in which the rising of ICP in infratentorial
compartment precedes its increase in
supratentorial region due to protective function
of the cerebellar tentorium. In such cases the
blood flow in cerebral hemispheres ceases with
marked delay in comparison to posterior circula-
tion [35] – see Fig. 21. The opposite sequence is
known as well, when cessation of blood flow in
supratentorial region precedes cerebral circulatory
arrest in the basilar artery [15, 18]. Diagnostic
difficulties are also encountered in cases of
hypoxic-ischemic brain injuries after cardiac
arrest when the ineffective restoration of cerebral
blood flow is observed [28, 35] – see Fig. 22. It
should be noted that in all abovementioned situa-
tions, clinical criteria of brain death are met. It is
obvious that the situation when the residual per-
fusion is revealed in a patient meeting the clinical
criteria of brain death leads to diagnostic confu-
sion and delays the diagnosis of brain death.
Future Trends: Looking for an Ideal
Ancillary Test
Although a number of ancillary tests were intro-
duced to the diagnosis of brain death, none of
them is ideal. A potentially ideal method should
890
Fig. 20 CT angiography and perfusion findings in a
71-year-old man after aneurysmal SAH and a bilateral
craniectomy who presented with signs of brain death on
clinical examination: (a, b) VRTs 60 s after contrast injec-
tion show filling of cortical branches of the MCAs (arrow-
heads) and the ACAs (arrows). No opacification of the
basilar artery or intracranial parts of the vertebral arteries is
noted (empty arrows); (c) VRT 60 s after contrast injection
Fig. 21 CT angiography and perfusion findings in a
34-year-old woman with brain stem hemorrhage with a
frontal craniotomy and ventricular drainage who presented
with signs of brain death on clinical examination: (a) 5 mm
NECT in axial plane shows massive edema, sulcal efface-
ment, ventricular narrowing, low GM/WM differentiation,
and hematoma in the brain stem (arrow); (b) VRT 60 s after
contrast injection shows filling of cortical branches of the
MCAs and the ACAs (arrows). No opacification of the
basilar artery or intracranial parts of the vertebral arteries is
M. Sawicki et al.
shows filling of the deep cerebral veins (arrowhead). CTA
findings preclude the diagnosis of brain death. Perfusion
CT-CBF (d) and CBV (e) maps show CBF of 15–20
ml/100 g/min and CBV 2–2.5 ml/100 g in cortical areas
at the regions of craniectomies (arrows) and in the right
basal ganglia (empty arrows). These values are above the
thresholds for necrosis, consistent with deep ischemia
noted (arrowheads). CTA findings preclude the diagnosis
of brain death. (c) CT perfusion. CBF map shows values
consistent with necrosis in the brain stem (asterisk), mid-
brain, cerebellum, and occipital lobes. However, the blood
flow is preserved in the basal ganglia and temporal and
frontal lobes. In infratentorial brain injuries, the blood flow
ceases initially in the posterior circulation, while perfusion
in the supratentorial compartment may be preserved for
some time
37 Brain Death Imaging 891

Fig. 22 CTA findings in a

58-year-old man after
cardiac arrest who
presented with signs of
brain death on clinical
examination. NECT: (a)
5 mm MPR in axial plane
shows diffuse edema and
hypodense basal ganglia
(arrows) and midbrain
(arrowhead); (b) 5 mm
MPR in axial plane shows
sulcal effacement and
hypodense cortex (arrows).
Cortical and deep GM is
more frequently affected
because neurons are more
vulnerable to hypoxia than
oligodendroglia or
astrocytes. CT angiography:
(c) VRT 60 s after contrast
injection shows filling of
cortical branches of the
MCAs and the ACAs
(arrowheads); (d) filling of
the deep cerebral veins is
preserved (arrow). CTA
findings preclude the
diagnosis of brain death

Table 5 Comparison of the selected blood flow studies in the diagnosis of brain death
Technical Interpretation Specificity Sensitivity
Test Availability Feasibility simplicity simplicity Noninvasiveness (%) (%)
Catheter  +  +  100 100
angiography
CTA +++ + ++ + + 100b 52–97a
PCT ++ + ++ ++ + 100b 89–100a
XeCT  + + ++ + 100 67–100
MRA +  ++ + + 100 100
DWI +  + + + 100 89–100
TCD  +/   + 100 91–100
Nuclear  + + ++ + 100 100
scan
a
Depending on the evaluation criteria used
b
Reliable case-control studies are missing
892
Fig. 23 Time-resolved dynamic CT angiography
(4D CTA) in a patient with clinical signs of brain death:
(a) VRT 10 s after injection shows opacification of extra-
cranial arteries (arrows) indicating proper contrast deliv-
ery. No filling of the ICAs or the VAs is visible; VRT 25 s
(b) and 45 s (c) after injection show extremely slow
be characterized by 100 % specificity (the test
does not reveal cerebral circulatory arrest in non-
brain-dead cases), sensitivity close to 100 %,
ubiquity, noninvasiveness, technical and interpre-
tational simplicity, and portability – see Table 5.
It seems that catheter angiography and TCD
reached the peak of their technical capabilities in
the diagnosis of brain death. This does not apply
to CT considering the recent introduction of
ultrafast, wide-detector scanners. Such machines
are capable to perform a time-resolved dynamic
CTA (4D CTA) – see Fig. 23. Combination of the
4D CTA with the whole-brain PCT using a single
contrast injection seems to be the most promising
alternative for the future. The issue of portability
of CT may be resolved in the next years as first
portable CT scanners are already on the market.
M. Sawicki et al.
propagation of contrast with weak opacification of
MCA-M1 (arrowheads) and the basilar artery (empty
arrow). Extracranially contrast outflow through the super-
ficial temporal veins is noted (arrows). No intracranial
venous outflow is noted
In recent years new modalities were proposed,
e.g., the bispectral index scale monitor (mathe-
matical algorithm of EEG), measuring of venous
oxygen saturation in the jugular bulb and brain
tissue oxygenation.
Summary
Brain death is defined as the irreversible cessa-
tion of functioning a the brain stem. Brain death
is principally established using clinical criteria
including coma, absence of brain stem reflexes,
and loss of central drive to breathe, assessed with
apnea test. In situations in which clinical testing
cannot be performed or when uncertainty exists
about the reliability of its parts due to
37 Brain Death Imaging
confounding conditions, ancillary tests
(i.e., imaging studies) may be useful. The objec-
tive of ancillary tests in the diagnosis of brain
death is to demonstrate the absence of cerebral
electrical activity (EEG and evoked potentials) or
cerebral circulatory arrest. In clinical practice
catheter cerebral angiography, perfusion scintig-
raphy, transcranial Doppler sonography, CT
angiography, and MR angiography are used.
Other methods, like perfusion CT, xenon CT,
MR spectroscopy, diffusion-weighted MRI, and
functional MRI, are being studied as potentially
useful in the diagnosis of brain death.
Among these tests, blood flow and perfusion
studies are most commonly applied. The diagno-
sis of brain death can be confirmed when a lack of
blood flow or tissue perfusion in the whole brain is
documented.
However, the application of blood flow and
perfusion tests to the diagnosis of brain death is
limited by variable vulnerability of individual
brain structures to ischemia. Besides there are
known factors predisposing to dissociation
between brain death and total cessation of cerebral
circulation. These factors include skull decom-
pression, infratentorial localization, and hypoxic-
ischemic mechanism of brain injury.
References
1. WHO (2012) International guidelines for the determina-
tion of death – phase I. Canadian Blood Services http://
www.who.int/patientsafety/montreal-forum-report.pdf
2. Aichner F, Felber S, Birbamer G, Luz G, Judmaier W,
Schmutzhard E (1992) Magnetic resonance: a nonin-
vasive approach to metabolism, circulation, and mor-
phology in human brain death. Ann Neurol 32(4):
507–511. doi:10.1002/ana.410320405
3. Kato T, Tokumaru A, O’Uchi T, Mikami I, Umeda M,
Nose K, Eguchi T, Hasegawa M, Okuyama K (1991)
Assessment of brain death in children by means of P-31
MR spectroscopy: preliminary note. Work in progress.
Radiology 179(1):95–99. doi:10.1148/radiology.179.
1.2006312
4. Ishii K, Onuma T, Kinoshita T, Shiina G,
Kameyama M, Shimosegawa Y (1996) Brain death:
MR and MR angiography. AJNR Am J Neuroradiol
17(4):731–735
5. Karantanas AH, Hadjigeorgiou GM, Paterakis K,
Sfiras D, Komnos A (2002) Contribution of MRI and
MR angiography in early diagnosis of brain death. Eur
893
Radiol 12(11):2710–2716. doi:10.1007/s00330-002-
1336-z
6. Luchtmann M, Beuing O, Skalej M, Kohl J, Serowy S,
Bernarding J, Firsching R (2014) Gadolinium-
enhanced magnetic resonance angiography in brain
death. Sci Rep 4:3659. doi:10.1038/srep03659
7. Lövblad K-O, Basssetti C (2000) Diffusion-weighted
magnetic resonance imaging in brain death. Stroke J
Cereb Circ 31(2):539–542. doi:10.1161/01.STR.31.2.539
8. Matsumura A, Meguro K, Tsurushima H, Komatsu Y,
Kikuchi Y, Wada M, Nakata Y, Ohashi N, Nose T
(1996) Magnetic resonance imaging of brain death.
Neurol Med Chir 36(3):166–171
9. Sohn CH, Lee HP, Park JB, Chang HW, Kim E, Kim E,
Park UJ, Kim HT, Ku J (2012) Imaging findings of
brain death on 3-tesla MRI. Korean J Radiol 13(5):
541–549. doi:10.3348/kjr.2012.13.5.541
10. Luchtmann M, Bernarding J, Beuing O, Kohl J, Bondar I,
Skalej M, Firsching R (2013) Controversies of diffusion
weighted imaging in the diagnosis of brain death. J Neu-
roimaging 23(4):463–468. doi:10.1111/jon.12033
11. Selcuk H, Albayram S, Tureci E, Hasiloglu ZI,
Kizilkilic O, Cagil E, Kocer N, Islak C (2012)
Diffusion-weighted imaging findings in brain death.
Neuroradiology 54(6):547–554. doi:10.1007/s00234-
011-0912-9
12. Falini A, Barkovich AJ, Calabrese G, Origgi D,
Triulzi F, Scotti G (1998) Progressive brain failure
after diffuse hypoxic ischemic brain injury: a serial
MR and proton MR spectroscopic study. AJNR Am J
Neuroradiol 19(4):648–652
13. Boly M, Tshibanda L, Vanhaudenhuyse A,
Noirhomme Q, Schnakers C, Ledoux D, Boveroux P,
Garweg C, Lambermont B, Phillips C, Luxen A,
Moonen G, Bassetti C, Maquet P, Laureys S (2009)
Functional connectivity in the default network during
resting state is preserved in a vegetative but not in a
brain dead patient. Hum Brain Mapp 30(8):
2393–2400. doi:10.1002/hbm.20672
14. Monti MM, Vanhaudenhuyse A, Coleman MR,
Boly M, Pickard JD, Tshibanda L, Owen AM, Laureys
S (2010) Willful modulation of brain activity in disor-
ders of consciousness. N Engl J Med 362(7):579–589.
doi:10.1056/NEJMoa0905370
15. Braun M, Ducrocq X, Huot JC, Audibert G,
Anxionnat R, Picard L (1997) Intravenous angiogra-
phy in brain death: report of 140 patients. Neuroradi-
ology 39(6):400–405
16. Heran MK, Heran NS, Shemie SD (2008) A review of
ancillary tests in evaluating brain death. Can J Neurol
Sci 35(4):409–419
17. Grosse K, Brauer B, Kucuk O, Kruger R, Polster F,
Grauhan O, Wesslau C (2006) Does contrast medium
administration in organ donors affect early kidney graft
function? Transplant Proc 38:668–669. doi:10.1016/j.
transproceed.2006.01.034
18. Kricheff II, Pinto RS, George AE, Braunstein P, Korein
J (1978) Angiographic findings in brain death. Ann N
Y Acad Sci 315:168–183
894
19. Dupas B, Gayet-Delacroix M, Villers D, Antonioli D,
Veccherini MF, Soulillou JP (1998) Diagnosis of brain
death using two-phase spiral CT. AJNR Am J
Neuroradiol 19(4):641–647
20. Bohatyrewicz R, Sawicki M, Walecka A, Walecki J,
Rowinski O, Bohatyrewicz A, Kanski A,
Czajkowski Z, Krzysztalowski A, Solek-Pastuszka J,
Zukowski M, Marzec-Lewenstein E, Wojtaszek M
(2010) Computed tomographic angiography and per-
fusion in the diagnosis of brain death. Transplant Proc
42(10):3941–3946. doi:10.1016/j.transproceed.2010.
09.143
21. Frampas E, Videcoq M, de Kerviler E, Ricolfi F,
Kuoch V, Mourey F, Tenaillon A, Dupas B (2009) CT
angiography for brain death diagnosis. AJNR Am J
Neuroradiol 30(8):1566–1570. doi:10.3174/ajnr.
A1614
22. Leclerc X, Taschner CA, Vidal A, Strecker G,
Savage J, Gauvrit JY, Pruvo JP (2006) The role of
spiral CT for the assessment of the intracranial circula-
tion in suspected brain-death. J Neuroradiol J
Neuroradiol 33(2):90–95
23. Sawicki M, Bohatyrewicz R, Safranow K, Walecka A,
Walecki J, Rowinski O, Solek-Pastuszka J,
Czajkowski Z, Guzinski M, Burzynska M, Wojczal J
(2014) Computed tomographic angiography criteria in
the diagnosis of brain death—comparison of sensitivity
and interobserver reliability of different evaluation
scales. Neuroradiology 56(8):609–620. doi:10.1007/
s00234-014-1364-9
24. Berenguer CM, Davis FE, Howington JU (2010) Brain
death confirmation: comparison of computed tomo-
graphic angiography with nuclear medicine perfusion
scan. J Trauma 68(3):553–559. doi:10.1097/
TA.0b013e3181cef18a
25. Escudero D, Otero J, Marques L, Parra D, Gonzalo JA,
Albaiceta GM, Cofino L, Blanco A, Vega P, Murias E,
Meilan A, Roger RL, Taboada F (2009) Diagnosing
brain death by CT perfusion and multislice CT angiog-
raphy. Neurocrit Care 11(2):261–271. doi:10.1007/
s12028-009-9243-7
26. Combes JC, Chomel A, Ricolfi F, d’Athis P, Freysz M
(2007) Reliability of computed tomographic angiogra-
phy in the diagnosis of brain death. Transplant Proc
39(1):16–20. doi:10.1016/j.transproceed.2006.10.204
27. Quesnel C, Fulgencio JP, Adrie C, Marro B, Payen L,
Lembert N, El Metaoua S, Bonnet F (2007) Limitations
of computed tomographic angiography in the diagnosis
of brain death. Intensive Care Med 33(12):2129–2135.
doi:10.1007/s00134-007-0789-6
28. Shankar JJ, Vandorpe R (2013) CT perfusion for con-
firmation of brain death. AJNR Am J Neuroradiol
34(6):1175–1179. doi:10.3174/ajnr.A3376
29. Sawicki M, Bohatyrewicz R, Safranow K, Walecka A,
Walecki J, Rowinski O, Solek-Pastuszka J,
Czajkowski Z, Marzec-Lewenstein E, Motyl K,
Przybyl W, Czarnecka A (2013) Dynamic evaluation
of stasis filling phenomenon with computed
M. Sawicki et al.
tomography in diagnosis of brain death. Neuroradiology
55(9):1061–1069. doi:10.1007/s00234-013-1210-5
30. Ashwal S, Schneider S, Thompson J (1989) Xenon
computed tomography measuring cerebral blood
flow in the determination of brain death in children.
Ann Neurol 25(6):539–546. doi:10.1002/ana.
410250603
31. Pistoia F, Johnson DW, Darby JM, Horton JA,
Applegate LJ, Yonas H (1991) The role of xenon CT
measurements of cerebral blood flow in the clinical
determination of brain death. AJNR Am J Neuroradiol
12(1):97–103
32. Hassler W, Steinmetz H, Pirschel J (1989) Transcranial
Doppler study of intracranial circulatory arrest. J
Neurosurg 71(2):195–201. doi:10.3171/jns.1989.71.
2.0195
33. Ducrocq X, Hassler W, Moritake K, Newell DW, von
Reutern GM, Shiogai T, Smith RR (1998) Consensus
opinion on diagnosis of cerebral circulatory arrest
using Doppler-sonography: task force group on cere-
bral death of the Neurosonology Research Group of the
World Federation of Neurology. J Neurol Sci 159(2):
145–150
34. Wijdicks EF (1995) Determining brain death in adults.
Neurology 45(5):1003–1011
35. Hadani M, Bruk B, Ram Z, Knoller N, Spiegelmann R,
Segal E (1999) Application of transcranial Doppler
ultrasonography for the diagnosis of brain death. Inten-
sive Care Med 25(8):822–828
36. Lovrencic-Huzjan A, Vukovic V, Gopcevic A,
Vucic M, Kriksic V, Demarin V (2011) Transcranial
Doppler in brain death confirmation in clinical practice.
Ultraschall Med 32(1):62–66. doi:10.1055/s-0029-
1245237 (Stuttgart, Germany: 1980)
37. Segura T, Calleja S, Irimia P, Tembl JI, Spanish Society
of Neurosonology (2009) Recommendations for the
use of transcranial Doppler ultrasonography to deter-
mine the existence of cerebral circulatory arrest as
diagnostic support for brain death. Rev Neurosci 20
(3–4):251–259
38. Widder B, Görtler M, Widder G (2004) Doppler- und
Duplexsonographie der hirnversorgenden Arterien, 6th
edn. Springer, Berlin
39. Wijdicks EF, Varelas PN, Gronseth GS, Greer DM,
American Academy of Neurology (2010) Evidence-
based guideline update: determining brain death in
adults: report of the Quality Standards Subcommittee
of the American Academy of Neurology. Neurology 74
(23):1911–1918. doi:10.1212/
WNL.0b013e3181e242a8
40. Alexandrov AV, Sloan MA, Tegeler CH, Newell DN,
Lumsden A, Garami Z, Levy CR, Wong LK,
Douville C, Kaps M, Tsivgoulis G, American Society
of Neuroimaging Practice Guidelines C (2012) Practice
standards for transcranial Doppler (TCD) ultrasound.
Part II. Clinical indications and expected outcomes. J
Neuroimaging 22(3):215–224. doi:10.1111/j.1552-
6569.2010.00523.x
37 Brain Death Imaging
41. Ducrocq X, Braun M, Debouverie M, Junges C,
Hummer M, Vespignani H (1998) Brain death and
transcranial Doppler: experience in 130 cases of brain
dead patients. J Neurol Sci 160(1):41–46
42. White H, Venkatesh B (2006) Applications of
transcranial Doppler in the ICU: a review. Intensive
Care Med 32(7):981–994. doi:10.1007/s00134-006-
0173-y
43. de Freitas GR, Andre C (2006) Sensitivity of
transcranial Doppler for confirming brain death: a
prospective study of 270 cases. Acta Neurol Scand
113(6):426–432. doi:10.1111/j.1600-0404.2006.
00645.x
44. Monteiro LM, Bollen CW, van Huffelen AC,
Ackerstaff RG, Jansen NJ, van Vught AJ (2006)
Transcranial Doppler ultrasonography to confirm
brain death: a meta-analysis. Intensive Care Med 32
(12):1937–1944. doi:10.1007/s00134-006-0353-9
45. Munari M, Zucchetta P, Carollo C, Gallo F, De
Nardin M, Marzola MC, Ferretti S, Facco E (2005)
Confirmatory tests in the diagnosis of brain death:
895
comparison between SPECT and contrast angiography.
Crit Care Med 33(9):2068–2073
46. Zuckier L, Kolano J (2008) Radionuclide studies in the
determination of brain death: criteria, concepts, and
controversies. Semin Nucl Med 38(4):262–273.
doi:10.1053/j.semnuclmed.2008.03.003
47. Donohoe KJ, Agrawal G, Frey KA, Gerbaudo VH,
Mariani G, Nagel JS, Shulkin BL, Stabin MG, Stokes
MK (2012) SNM practice guideline for brain death
scintigraphy 2.0. J Nucl Med Technol 40(3):198–203.
doi:10.2967/jnmt.112.105130
48. Donohoe KJ, Frey KA, Gerbaudo VH, Mariani G,
Nagel JS, Shulkin B (2003) Procedure guideline for
brain death scintigraphy. J Nucl Med 44(5):846–851
49. Sinha P, Conrad GR (2012) Scintigraphic confirmation
of brain death. Semin Nucl Med 42(1):27–32.
doi:10.1053/j.semnuclmed.2011.07.007
50. Wijdicks EF, Pfeifer EA (2008) Neuropathology of
brain death in the modern transplant era. Neurology
70(15):1234–1237. doi:10.1212/01.wnl.0000289762.
50376.b6
 Part VII
Veins Imaging
 Jugular Vein Thrombosis
38
Nicholas A. Koontz, Richard H. Wiggins III, and
Lubdha M. Shah

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 900 Jugular vein thrombosis (JVT) is a not uncom-
mon and serious finding encountered by radi-
Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 900
ologists. While identifying thrombus within
Pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902 the internal jugular vein (IJV) is usually not a
Clinical Presentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 903 diagnostic dilemma, a thorough evaluation of
the surrounding structures is critical due to the
Etiologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 904
Iatrogenic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 904 sundry etiologies, potentially devastating com-
Neoplastic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 906 plications (including pulmonary embolism,
Infectious . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 908 septic emboli, and dural sinus thrombosis),
Traumatic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 915
and high association with neoplasm. In this
Hematologic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 915
Rheumatologic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 916 chapter, we discuss the typical anatomy of the
Gynecologic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 917 IJV and the pathophysiology, clinical presenta-
Endocrine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 918 tion, and potential etiologies of JVT (including
Idiopathic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 918
iatrogenic, neoplastic, infectious, traumatic,
JVT in Pediatric Patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 918
hematologic, rheumatologic, gynecologic,
Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 919 endocrine, and idiopathic causes, as well as
CT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 919
US . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 921 pediatric-specific considerations). Multimodality
MRI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 923 imaging findings of JVT (computed tomography,
Nuclear Medicine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 924 magnetic resonance imaging, ultrasound, nuclear
Angiography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 927 medicine, and angiography) and the diagnostic
Diagnostic Pitfalls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 928 pitfalls are also reviewed, utilizing an image-rich,
False-Negative Misinterpretation . . . . . . . . . . . . . . . . . . . . 929 comprehensive review of the medical literature.
False-Positive Misinterpretation . . . . . . . . . . . . . . . . . . . . . 929
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 931 Keywords
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 931 CT • MRI • Vascular venous imaging • Jugular
vein

N.A. Koontz • R.H. Wiggins III • L.M. Shah (*)

Department of Radiology, University of Utah Health
Sciences Center, Salt Lake City, UT, USA
e-mail: Nick.Koontz@hsc.utah.edu; Richard.
Wiggins@hsc.utah.edu; Lubdha.shah@hsc.utah.edu

# Springer Science+Business Media New York 2016 899

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_42
900 N.A. Koontz et al.

Introduction Table 1 Causes of jugular vein thrombosis

Iatrogenic
Jugular vein thrombosis (JVT) is a not uncommon Central venous access device (CVAD)
pathologic process characterized by thrombus for- Cardioverter defibrillator/pacemaker
mation within the internal jugular vein (IJV). This Ventriculoatrial shunt
review will focus on the extracranial IJV. Postoperative
Although JVT is a relatively straightforward Cervical manipulation
diagnosis to make with imaging, the challenge Neoplastic
lies in determining the etiology and identifying Mass effect
associated pathology and complications. While Primary tumor
Lymphadenopathy
most commonly encountered in ill patients
Direct invasion
with thrombogenic central venous catheters or
Migratory thrombophlebitis (Trousseau syndrome)
postoperatively in patients following neck
Posttreatment (chemotherapy and/or radiation therapy)
dissections, sundry etiologies exist beyond the
Infectious
obvious iatrogenic sources and are Intravenous drug abuse (IVDA)
frequently dependent upon the patient profile Human necrobacillosis (Lemierre syndrome)
(Table 1). Otomastoiditis
Given the many etiologies of JVT, it is impor- Postauricular abscess
tant for radiologists to consider the clinical con- Cytomegalovirus
text with this imaging finding, as well as the Traumatic
appearance on different imaging modalities. In Blunt trauma
doing so, one can suggest an appropriate, Penetrating trauma
patient-specific, clinically relevant differential Hematologic
diagnosis and add real clinical value to a radio- Activated protein C (APC) resistance
logic report. A comprehensive search for the eti- Congenital
ology of JVT will often rely on a thorough review Factor V Leiden thrombophilia
of the electronic medical record and clinical con- Factor V Cambridge mutation
sultation, as well as scrutinizing other radiologic Acquired
Antiphospholipid antibodies (APLAs)
studies for imaging clues. The diagnosis of JVT,
Oral contraceptives
like so many other diagnoses encountered in diag-
Paraneoplastic
nostic imaging, is an investigational starting point
Thalidomide
and not simply a dead-end observation.
Hemophilia therapies
Factor VIII inhibitor bypass activity (FEIBA)
Activated recombinant factor VII (rVIIa)
Rheumatologic
Anatomy Rheumatoid arthritis
Systemic lupus erythematosus
The paired IJVs serve as the primary conduit for Silk road disease (Behçet disease)
venous return not only from the superficial parts Gynecologic
of the face and neck but also from the intracranial Assisted reproductive techniques (ART)
circulation to the right atrium of the heart (Fig. 1). In vitro fertilization
Functionally, the IJVs are a continuation of the Ovarian hyperstimulation syndrome (OHSS)
sigmoid sinuses below the level of the jugular Pregnancy
foramina in the skull base. The left IJV is gener- Endocrine
ally smaller than the right, and there is a unilateral Tertiary or secondary hypercortisolism (Cushing
syndrome)
valve in most cases [1].
Thyroid carcinoma
At its cranial terminus is a normal, focal dila-
Goiter
tation of the IJV, which is known as the superior
(continued)
38 Jugular Vein Thrombosis 901

Table 1 (continued)
Idiopathic
Pediatric patients
Iatrogenic
CVAD
Intravenous immunoglobulin (IVIg) infusion
Infectious
Human necrobacillosis (Lemierre syndrome)
Mastoiditis
Neoplastic
Hematologic
Traumatic
Congenital heart disease
Cystic fibrosis

Fig. 2 Typical appearance of the superior jugular

bulb (black open arrow), a normal focal dilatation of
the IJV

bulb (Figs. 1 and 2). The superior jugular bulb

Fig. 1 Coronal reconstruction of CECT of the neck dem-
onstrates a normal appearance of the right IJV (black
arrow), which extends from the confluence of the subcla-
vian and brachiocephalic veins (white asterisk) centrally to
the sigmoid sinus peripherally. Normal, focal dilatations at
the cranial and caudal extremes of the IJV are known as the
superior jugular bulb (black open arrow) and inferior jug-
ular bulb (white open arrow), respectively
typically lies below the floor of the internal audi-
tory canal (IAC), but those extending above the
floor of the IAC are considered “high-riding.”
Although a high-riding jugular bulb is a clinically
silent finding, it may be at increased risk of injury
during mastoidectomy. In comparison, a dehis-
cent jugular bulb occurs when rarefaction of the
osseous covering allows extension of the jugular
bulb into the middle ear cavity. One may also
encounter a jugular bulb diverticulum variant,
which is a tubular projection of the jugular bulb
into the adjacent skull base without extension into
the middle ear.
Below the skull base, the IJV resides within the
carotid space, deep into the sternocleidomastoid
muscle. In the suprahyoid neck, the IJV is typi-
cally located posterolateral to the internal carotid
artery (ICA) (Fig. 3), overlying the rectus capitis
lateralis, and the glossopharyngeal and hypoglos-
sal nerves pass between the ICA and IJV, while
the vagus nerve lies posteriorly. In the infrahyoid
neck, the IJV is typically located lateral and
anterolateral to the internal carotid and common
carotid arteries (Fig. 4). However, the exact
902 N.A. Koontz et al.

Fig. 3 Normal anatomy of

the IJV in the suprahyoid
neck. Axial CECT (a) and
axial T1-WI +C (b) at the
level of the nasopharynx
demonstrate the IJV (white
arrow) to be located lateral
to the ICA (white open
arrow)

Fig. 4 Normal anatomy of the IJV in the infrahyoid neck.

Axial CECT at the level of the glottic larynx demonstrates
the IJVs (white arrows) to be located lateral to the internal
carotid arteries (white open arrows). While typically
located lateral or anterolateral to the ICA in the infrahyoid
neck, the location of the IJV can be variable

location of the IJV with respect to the carotid

artery is somewhat variable, particularly at the
extremes of age [2]. At its caudal terminus, the
IJV joins the subclavian vein and forms the
brachiocephalic vein. Just proximal to its termi-
nation is a second dilatation, the inferior jugular
bulb (Figs. 1 and 5). Numerous venous tributaries
commonly feed into the IJV, including the facial
vein, retromandibular vein, superior thyroid vein,
and middle thyroid vein, although there is great
variation and overlap of tributary venous return.
Fig. 5 Coronal CECT image demonstrates the inferior
jugular bulb (white open arrow), a normal, focal dilatation
of the IJV just before its confluence with the jugular vein
Pathophysiology
The pathophysiology of JVT is rooted in the well-
known Virchow triad. This is named after, but
probably not actually described (at least in its
38 Jugular Vein Thrombosis
Static Blood
Flow
Hyper- Endothelial
coaguability Injury
Fig. 6 An abnormality in one or more of these variables
increases the risk of developing a thrombus. Areas of
intersection of the circles in this Venn diagram reflect
increased risk of thrombosis. At highest risk are patients
who fall into all three categories
current understanding), by Rudolf Ludwig Karl
Virchow (1821–1902), the so-called father of
modern pathology; this model posits that three
broad categories of factors are at play in the devel-
opment of vascular thrombosis: static blood flow,
endothelial injury, and hypercoagulability [3]
(Fig. 6). An abnormality of one or more of these
variables increases the risk of developing a throm-
bus. In its native state, the IJV, like all other veins
of the head and neck, is an unlikely culprit to
develop thrombus, due in large part to its paucity
of valves and the aid of gravity in venous return
while the body is in an upright state [4]. While in
the supine position, the IJVs serve as the primary
venous drainage pathway for the brain; in the erect
position, these veins collapse and flow via alter-
nate vertebral venous pathways [5]. For these
reasons, spontaneous JVT is rare and must raise
clinical concern for an underlying culprit causing
static blood flow, endothelial injury, or
hypercoagulability.
However, certain patient populations are at
increased risk of altering the dynamics of the
Virchow triad, shifting them toward a
prothrombotic state and subsequent JVT
(Table 2). For example, patients with central
903
Table 2 High-risk patient populations for development
of JVT
Central venous access device (CVAD)
Intravenous drug abuse (IVDA)
Malignancy
venous access devices (CVADs) are at increased
risk of JVT due to static venous flow along the
catheter as well as due to endothelial injury from
catheter placement. Patients with a history of
intravenous drug abuse (IVDA) are at increased
risk of JVT from either caustic, chemical endo-
thelial injury, or endothelial injury from IVDA-
associated blood-borne infection. Patients with
underlying neoplasm may be at increased risk
for JVT formation due to a systemic, cancer-
induced hypercoagulable state or from compres-
sion of the vessel lumen by cervical lymphade-
nopathy or primary tumor, resulting in static flow
and impaired venous return. These few examples
illustrate how the components of the Virchow
triad contribute to thrombosis of this ordinarily
resilient vessel. Keeping these basic principles in
mind during image interpretation can be quite
helpful in deciphering the etiology of JVT in
specific clinical settings.
Clinical Presentation
Just as the etiologies of JVT are diverse and
numerous, so too can be the clinical presentation
(Table 3). Presentation may range from
completely asymptomatic (incidental) to palpable
neck lump, dysphagia, cervical edema, upper
extremity edema, erythrocyanosis, varicose
superficial collateral veins, pulmonary embolism,
or frank sepsis, depending on the etiology, acuity,
and extent of disease [6, 7]. Patients more com-
monly present with unilateral disease, although
bilateral disease is not infrequently encountered
(Fig. 7). During the acute and subacute phases
(typically within 10–14 days of presentation) of
jugular vein thrombophlebitis, patients may expe-
rience diffuse edema and tenderness overlying the
sternocleidomastoid muscle. Patients are often
904 N.A. Koontz et al.

Table 3 Clinical presentation of jugular vein thrombosis

Acute phase (first 2 weeks)
Asymptomatic
Palpable lump
Neck pain
Dysphagia
Cervical or upper extremity edema
Erythrocyanosis
Varicose collateral veins
Pulmonary embolism
Sepsis
Chronic phase (>2 weeks)
Palpable, cord-like mass
Varicose collateral veins

Fig. 8 Postoperative NECT in a 31-year-old man with

recent left retrosigmoid craniectomy (black open arrow)
for resection of a vestibular schwannoma demonstrates
high density within the left IJV (black arrow),
corresponding to “downhill thrombus” propagating from
the thrombosed sigmoid sinus

Etiologies

In order to systematically cover the many possible

etiologies for JVT, it is helpful to first consider the
diagnosis from a “bird’s-eye view,” compartmen-
talizing the potential sources into several general
categories – iatrogenic, neoplastic, infectious,
traumatic, hematologic, rheumatologic, gyneco-
logic, endocrine, and idiopathic – which will be
discussed individually. Potential etiologies of JVT
in the pediatric population will also be discussed.

Fig. 7 Sixty-year-old man with left ventricular access Iatrogenic

device (LVAD) who presents with altered mental status
concerning for acute cerebral infarct. Coronal reformat of
CTA of the neck demonstrates occlusive thrombus in the
left IJV (white arrow) and near-complete occlusive throm-
bus in the right IJV (white open arrow)
suspected of having an abscess. In the chronic
phase after the inflammation subsides, thrombus
may persist within the internal jugular vein,
resulting in a palpable, cord-like mass and in
many cases total occlusion.
Given the widespread use of CVADs, the preva-
lence of head and neck surgery, and the increased
use of transvenous medical devices, iatrogenic
sources are the most common causes of JVT in
modern medicine [8–10]. IJV occlusion (typically
non-thrombotic) is infrequently seen postopera-
tively following head and neck surgery
(Figs. 8–15). Among iatrogenic causes, the vast
majority of JVT is due to central venous catheters,
38 Jugular Vein Thrombosis
Fig. 9 Volumetric T1-WI +C in the same patient as Fig. 8
demonstrates a low signal intensity filling defect (white
open arrow) within the enlarged left jugular bulb,
corresponding with acute postoperative “downhill” JVT
that had propagated from the sigmoid sinus
Fig. 10 Coronal 2D TOF image in the same patient as
Figs. 8 and 9 demonstrates apparent filling defect (white
arrow) in the right IJV which is due to turbulent blood flow
and should not be confused with true thrombus. In equiv-
ocal cases, further evaluation with dynamic post-contrast
MR venography or second-look ultrasound can be consid-
ered for clarification
905
Fig. 11 Coronal maximal intensity projection (MIP) from
2D time-of-flight (TOF) MRV in the same patient as
Figs. 8, 9, and 10 shows no flow-related enhancement
within the distal left transverse sinus, sigmoid sinus, jugu-
lar bulb, and visualized IJV, corresponding with “down-
hill” thrombus following left retrosigmoid craniectomy
often resulting in “uphill thrombosis” that propa-
gates into the IJV from the downstream subcla-
vian or brachiocephalic veins (Fig. 16). At-risk
patient populations include those with chronic
indwelling central venous catheters, as might be
utilized for chemotherapy or frequent blood
draws, and critically ill neonates requiring central
venous access. JVT as a complication of other
transvenous biomedical devices, such as
cardioverter defibrillators, pacemakers, and
ventriculoatrial shunts, is seen less commonly but
has been reported in the medical literature [9, 10].
As previously described, most iatrogenic causes
of JVT center around catheter-/device-induced
venous stasis and endothelial injury, although it is
important to consider superimposed hypercoa-
gulability in these typically ill patients.
A large group of patients with increased sus-
ceptibility to iatrogenic JVT is hemodialysis
(HD) patients, as they tend to have numerous,
long-term CVADs placed over a lifetime
(Fig. 17). Studies have demonstrated that in HD
906
Fig. 12 Axial diffusion-
weighted imaging (a) and
ADC map (b) in the same
patient as Figs. 8, 9, 10, and
11 demonstrate the JVT to
be hyperintense to
surrounding osseous
structures on both
sequences
Fig. 13 Axial T1-WI C demonstrates the left JVT
(white open arrow) to be isointense to brain parenchyma
and slightly hyperintense to skeletal muscle
patients, JVT risk increases with both duration of
catheter placement and total lifetime number of
catheters placed, with subsequent JVT reported in
2–64 % of cases based on older studies [8, 11].
The underlying renal insufficiency itself is asso-
ciated with hemostatic abnormalities, and HD
adds to these disturbances with turbulent flow,
high shear stress, and contact of blood to artificial
surfaces. This leads to activation of the coagula-
tion cascade, which is the target of anticoagulation
agents, such as unfractionated heparin and low-
molecular-weight heparin.
N.A. Koontz et al.
Another subset of patients at risk for IJV
thrombosis or occlusion is that group who has
undergone neck surgery, most commonly follow-
ing neck dissection for head and neck cancer.
While IJV patency is dependent on the type and
extent of neck dissection (with intentional sacri-
fice of the IJV performed in some cases), true
thrombosis or complete occlusion following
IJV-sparing neck dissection is uncommon
[12]. However, a diminution in IJV caliber is
expected following neck dissection, reaching a
nadir in the immediate postoperative period and
frequently returning to near-normal caliber by
3 months [12]. Uncommonly, JVT can be seen
following carotid endarterectomy, presumably
due to prolonged intraoperative retraction of
the IJV.
Lastly, there are reports of JVT following cer-
vical manipulation, including chiropractic and
deep tissue massage. Other than case reports, the
literature evidence for or against cervical
manipulation-induced IJV thrombosis is certainly
lacking. The mechanism of thrombus formation in
these rare cases is not well elucidated, although it
might theoretically reflect a combination of
pressure-induced venous stasis and traumatic
endothelial injury.
Neoplastic
Like lower extremity deep vein thrombosis
(DVT), JVT is commonly associated with
38 Jugular Vein Thrombosis
Fig. 14 Axial T1-WI +C (a) and axial T2-WI FS (b) in a
31-year-old man with recent left retrosigmoid craniectomy
for resection of a vestibular schwannoma demonstrates
“downhill” JVT (white open arrow) that has propagated
from iatrogenic thrombosis of the left sigmoid sinus. Note
Fig. 15 Sagittal SPGR-C
imaging through the
bilateral IJVs in the same
patient demonstrates the
normal, low signal (white
arrow) within the patent
right IJV (a) and increased
signal (white open arrow)
within the thrombosed left
IJV (b)
malignancy, with coincidence reported up to
10–35 % in some series [7, 13] (Figs. 18–21).
Mechanistically, neoplasm-related thrombosis of
the IJV results from (a) compression/invasion of
the vessel lumen by lymphadenopathy or primary
tumor, resulting in static blood flow, (b) a com-
plex, incompletely understood cancer-induced
hypercoagulable state (often referred to as migra-
tory thrombophlebitis or Trousseau syndrome)
that results in increased levels or activity of sev-
eral procoagulant factors (including tissue factor,
fibrinogen, factor VIIa, factor VIII, factor IXa,
907
that the thrombus is isointense to slightly hyperintense to
brain parenchyma on T1-WI +C and T2-WI FS and dem-
onstrates a thin rim of peripheral enhancement on post-
contrast imaging due to enhancement of engorged vasa
vasorum
factor Xa, C-reactive protein, carcinoma
mucins, tumor-associated cysteine proteinase, or
oncogene activation), or (c) iatrogenic conse-
quences of cancer treatment (cf. Iatrogenic)
[4, 14, 15].
Among all cancer patients, treatment-related
causes are the most common etiology of JVT;
these include CVADs, prolonged immobility,
intravascular administration of caustic chemother-
apeutic agents, postsurgical vascular injury, and
postradiation therapy vascular complications.
Nonetheless, the presence of JVT, especially in
908
Fig. 16 Digital subtraction angiography (DSA) of the
right upper extremity in a 45-year-old woman with “uphill
thrombosis” of the right IJV demonstrates filling defects
(black arrows) within the right IJV, subclavian vein, right
brachiocephalic vein, left brachiocephalic vein, and supe-
rior vena cava. A potentially thrombogenic central venous
catheter is present (black open arrow)
cancer patients, should prompt the interpreting
radiologist to carefully assess the cervical lymph
node stations for metastatic lymphadenopathy and
to assess the deep spaces of the neck for a lesion
exerting mass effect on the thrombosed IJV. While
most often due to underlying head and neck can-
cer (such as squamous cell carcinoma of the
aerodigestive tract or thyroid malignancy), cervi-
cal chain lymphadenopathy causing JVT can also
be seen in widely disseminated metastatic disease
of any number of primary malignancies
[16–18]. The most common of these primary
malignancies include pancreatic, lung, stomach,
and ovarian cancer, which may contribute to JVT
due to increased factor VIII and thromboplastin
levels. Additionally, nonmalignant neoplasms,
such as lymphangioleiomyomatosis, can lead to
JVT due to local mass effect and resulting venous
stasis. Both contrast-enhanced CT (CECT) and
ultrasound (US) are useful modalities for assess-
ment of the neck masses, with US offering the
opportunity for image-guided fine needle aspira-
tion or core needle biopsy of suspicious masses or
lymphadenopathy.
N.A. Koontz et al.
Fig. 17 Coronal MIP from CTA of the neck in a 60-year-
old critically ill patient requiring hemodialysis delineates
the nonocclusive thrombus in the right IJV (white open
arrow), as well as a potentially thrombogenic right internal
jugular venous catheter (white arrowhead)
Infectious
Prior to the widespread use of antibiotics in the
1940s, infection was by far the most common
etiology of JVT. However, in current medical
practice infectious sources of JVT are quite rare,
attributed mainly to IVDA (Figs. 22–24) and spo-
radic cases of typically oropharyngeal infections
resulting in Lemierre syndrome. IVDA increases
the risk of JVT for several reasons: (1) shooters
typically do not employ sterile technique, thus
risking blood-borne infection; (2) traumatic endo-
thelial injury from the needle encourages throm-
bosis; (3) infectious (mycotic) endothelial damage
by blood-borne organisms results in
prothrombotic infiltration and congestion of the
adventitia (and subsequently the other layers of
38 Jugular Vein Thrombosis 909

Fig. 20 Sagittal reconstruction from CECT in the same

patient as Figs. 18 and 19 demonstrates thrombus (white
Fig. 18 Coronal reconstruction from a CECT of the cer- open arrow) within the left IJV, as well as the bulky,
vical soft tissues in a 47-year-old man with laryngeal necrotic, conglomerate nodal metastatic disease (white
squamous cell carcinoma (SCCa) and necrotic, conglom- arrow)
erate nodal metastases (white arrow) demonstrates marked
mass effect upon the left IJVand direct invasion with tumor
thrombus (white open arrow)

Fig. 19 Axial CECT in the same patient as Fig. 18 dem- Fig. 21 Axial CECT in the same patient as Figs. 18, 19,
onstrates the necrotic conglomerate nodal metastases and 20 demonstrates soft-tissue density tumor thrombus
(white arrow) with extranodal spread of tumor, which invading the left IJV (white open arrow) and encasing the
completely obliterates the left IJV ICA (white arrowhead)
910
Fig. 22 Axial CECT in a 50-year-old woman with acute
catheter-associated JVT and septic thrombophlebitis, who
presented with 6 days of right-sided neck pain, swelling, and
fever. Image demonstrates complete occlusion of an enlarged
right IJV with central low-attenuation filling defect (white
arrow), prominent enhancement of the vasa vasorum (white
open arrow), perivascular edema (white asterisks), subtle intra-
muscular enhancement (black arrowhead) within an enlarged,
edematous right sternocleidomastoid muscle (black arrow),
and multiple venous bypass collaterals (white arrowheads)
Fig. 23 Coronal reformat of CECT in the same patient as
Fig. 22 demonstrates complete occlusion of an enlarged right
IJV with central low-attenuation filling defect (white arrow),
prominent enhancement of the vasa vasorum (white open
arrow), perivascular edema (white asterisks), intramuscular
phlegmon/abscess (black arrowhead) within an enlarged,
edematous right sternocleidomastoid muscle (black arrow),
and prominent venous bypass collateral (white arrowhead)
N.A. Koontz et al.
Fig. 24 Sagittal reconstruction of CECT in the same
patient as Figs. 22 and 23 demonstrates complete occlusion
of an enlarged right IJV with central low-attenuation filling
defect (white arrow), prominent enhancement of the vasa
vasorum (white open arrow), and perivascular edema
(white asterisks)
the vein wall) by inflammatory cells; and
(4) IVDA-associated deep neck infection contrib-
utes to altered blood flow dynamics, including
venous stasis [19]. Other infectious causes of the
deep spaces of the head and neck, such as post-
surgical infection (Figs. 25–33), acute
otomastoiditis, or postauricular abscess, may
also contribute to the development of JVT, albeit
much less commonly. Additionally, an increased
risk of JVT in immunocompetent patients with
severe acute cytomegalovirus infection has been
described, a link which may be under-recognized
and underreported.
Lemierre Syndrome
A unique subset of JVT that warrants special
consideration is Lemierre syndrome (LS), also
known as human necrobacillosis or anaerobic
postanginal sepsis. Although not originally
described by French bacteriologist André
Lemierre (1875–1956), this eponymous distinc-
tion was awarded to him as a result of his detailed
38 Jugular Vein Thrombosis 911

Fig. 25 Axial CECT in a

76-year-old woman with
recent anterior cervical
fusion presents with neck
pain and fever. Image
reveals central filling defect
in the left IJV (white
arrowhead) compatible
with JVT. Additional
thrombus was seen in the
left ventral epidural venous
plexus (black arrow) and
left-sided muscular
collateral veins (white open
arrow)

Fig. 26 Coronal reconstruction of the NECT data in the

same patient as Fig. 25 better delineates the irregular,
low-attenuation filling defect (white asterisk) seen eccen-
trically within the left IJV as well as peripheral wall
enhancement corresponding to engorged vasa vasorum
(white arrowhead)
Fig. 27 Coronal reconstruction of the NECT data in the
same patient as Figs. 25 and 26 better delineates the irreg-
ular, low-attenuation filling defect (white asterisk) seen
eccentrically within the left IJV as well as peripheral wall
enhancement corresponding to engaged vasa vasorum
(white arrowhead)

description of systemic manifestations of Bacillus

funduliformis (now called Fusobacterium septic embolic disease (typically to the lungs),
necrophorum) in 1936. Classically, LS is charac- internal jugular vein thrombophlebitis, and isola-
terized by a constellation of clinical findings that tion of bacteria (typically F. necrophorum) from
include pharyngitis within the preceding 4 weeks, the blood or another sterile body site.
912
Fig. 28 Axial CECT at the
C1 level demonstrates a
filling defect (black arrow)
within the left ventral
epidural space corresponds
with additional thrombus
Fig. 29 Sagittal T1-WI +C demonstrates filling defect
within a left-sided muscular collateral vein (white arrow)
as well as peripheral enhancement within the vessel
resulting from enhancement of the vasa vasorum
Although most commonly associated with
F. necrophorum (57 %) or other Fusobacterium
species (33 %), LS has also been reported with
Staphylococcus aureus (including methicillin-
resistant), Streptococcus, Peptostreptococcus,
Proteus, Bacteroides, Klebsiella, and
Porphyromonas species [20]. While LS is most
often associated with a head and neck site of
infection (~75 %), including the tonsils (37 %),
pharynx/upper respiratory tract (30 %), middle
N.A. Koontz et al.
Fig. 30 Axial T1-WI +C demonstrates filling defect
(white arrow) within the ventral left epidural space at
C1–C2, compatible with thrombus in the epidural venous
plexus
ear/mastoid (2 %), larynx (2 %), dental (1 %),
paranasal sinuses (1 %), and orbit (1 %), approx-
imately 25 % of cases originate in the chest or
lower respiratory tract [20]. Extraordinarily rare
cases of LS originating from the gastrointestinal
tract, genitourinary tract, and musculoskeletal
system have also been reported. Presenting symp-
toms typically relate to site of involvement, most
commonly sore throat, neck pain, or neck mass.
Several series have described LS to primarily
affect previously healthy children, adolescents,
38 Jugular Vein Thrombosis
Fig. 31 Axial T1-WI +C in the same patient as Fig. 30
demonstrates peripherally enhancing fluid collections
(black open arrows) in the prevertebral soft tissues, com-
patible with small abscesses
Fig. 32 Axial T1-WI +C demonstrates a filling defect
within the left IJV (white arrowhead), which is isointense
to skeletal muscle. The patient has recently undergone
anterior cervical fixation (white arrow)
and young adults, with an equal to slightly
increased male-to-female predominance [20, 21].
LS is characterized by four distinct pathophys-
iologic phases: primary infection, invasion of the
parapharyngeal (lateral pharyngeal) space, inter-
nal jugular vein thrombophlebitis, and metastatic
(septic embolic) disease [22]. While disease
913
Fig. 33 Axial T2-WI in the same patient as Fig. 32 illus-
trates a filling defect within the left IJV (white arrowhead)
which is slightly hyperintense to skeletal muscle. The
patient has recently undergone anterior cervical fixation
(white arrow)
progression often advances in a predictable, step-
wise manner through these phases, it is not unex-
pected to find cases where metastatic
complications, such as pulmonary disease, pre-
sent before JVT or JVT appears before
parapharyngeal disease. Early intervention with
prolonged (multi-week) course of antibiotics is
the current therapeutic standard [21, 22]. Thera-
peutic anticoagulation remains controversial,
often reserved for patients with underlying
prothrombotic disorder, secondary cerebral
infarction, or associated cavernous sinus throm-
bosis [21]. Surgical intervention is typically lim-
ited to abscess drainage with IJV ligation rarely
performed in current clinical practice.
Perhaps as a result of the widespread use of
antimicrobial agents starting in the mid-1940s, LS
became a vanishingly rare clinical entity through
the 1980s, receiving the moniker of “forgotten
disease” by many physicians [20, 21, 23].
However, since then, there has been a progressive
increase in the reported LS cases in the medical
literature. If there is an increased incidence
of LS, it may be the result of increased anti-
biotic resistance, evolving antibiotic prescribing
patterns of physicians, or a reduced number of
914 N.A. Koontz et al.

tonsillectomies. However, it is unclear whether

Fig. 34 Axial CECT at the level of the oropharynx in a
20-year-old man who presents with right-sided neck pain,
fever, and odynophagia demonstrates classic imaging find-
ings of Lemierre syndrome, including nonocclusive throm-
bus in the right IJV (white open arrow), edema and
phlegmon replacing the normal triangular-shaped fat of
the right parapharyngeal space (white arrowhead), and
tonsillopharyngitis with enlarged bilateral palatine tonsils
(white arrows) with striated enhancement, the so-called
“tiger stripe” sign
the increase in reported cases is a true increase in
the incidence of LS or increased detection due to
improved medical imaging [20, 21].
First-line imaging evaluation of suspected LS
is best accomplished with CECT of the neck and
chest, which provides a rapid, comprehensive
evaluation of the primary site of infection (usually
the tonsils), the adjacent deep tissues of the neck,
the IJVs, associated cervical chain lymphadenop-
athy, and assessment for metastatic/septic embolic
disease involving the pulmonary parenchyma
(Figs. 34 and 35). CECT can accurately identify
tonsillar and peritonsillar edema, phlegmon, and
abscess associated with the primary site of infec-
tion as well as elucidate filling defects within the
IJVs and tributary vessels and perivascular edema
(thrombophlebitis). Magnetic resonance imaging
(MRI) and magnetic resonance venography
(MRV) can provide useful adjunctive imaging,
particularly for evaluation of intracranial exten-
sion of disease. Additionally, duplex Doppler
ultrasound can rapidly identify JVT but is limited
for assessment of secondary complications.

Fig. 35 Coronal
reconstruction of the CECT
in the same 20-year-old
patient with Lemierre
syndrome in Fig. 34
demonstrates the
nonocclusive thrombus in
the right IJV (black open
arrow) as well as associated
enlarged, reactive
jugulodigastric lymph
nodes (white arrowhead)
38 Jugular Vein Thrombosis
Fig. 36 Axial CECT in a 25-year-old man with gunshot
wound to the left neck demonstrates posttraumatic JVT
(black open arrow) as well as obvious signs of penetrating
trauma, including subcutaneous emphysema (white arrow-
head) and multiple ballistic fragments (black arrows)
Traumatic
Reports of posttraumatic JVT in the medical liter-
ature are exceedingly rare with only a few case
reports documented, more likely due to the mun-
dane nature of traumatic etiologies rather than a
truly low incidence (Figs. 36–38). The few
reported cases have all involved nonpenetrating
trauma to the head or neck, one of which demon-
strated concurrent thrombosis of the sigmoid
sinus (suggesting propagation of dural sinus
thrombosis into the downstream IJV) [24, 25]. In
the authors’ experience, JVT associated with pen-
etrating trauma is not uncommon. The true inci-
dence of posttraumatic JVT is unknown but is
likely higher than the paucity of literature reports
might suggest. The presence of JVT in a trauma
patient certainly warrants close inspection of the
dural venous sinuses, especially in the setting of
skull base fractures (Fig. 39).
Hematologic
As previously described (cf. Pathophysiology),
the propensity to develop JVT is entrenched in
915
Fig. 37 Sagittal reconstruction of CECT of the neck in
the same patient as Fig. 36 demonstrates
nonocclusive thrombus in the left IJV (black
open arrow), diffuse subcutaneous emphysema (white
arrowhead), and extensive streak artifact from ballistic
fragments (black arrow)
alterations of the Virchow triad that shift patients
toward a prothrombotic state. Underlying
hypercoagulability of various etiologies can play
a critical role in tipping this balance toward JVT,
which has been reported in patients with activated
protein C (APC) resistance. APC resistance,
which is thought to account for somewhere
between 20 % and 50 % of all thrombotic events
in general, may be either congenital (hereditary
factor V Leiden or factor V Cambridge gene
mutations) or may be acquired (resulting from
antiphospholipid antibodies, oral contraceptive
use, paraneoplastic syndromes, or due to thalido-
mide use) [26]. Interestingly, JVT has also been
reported in patients with hemophilia, which seems
counterintuitive given their underlying bleeding
disorder. While thrombotic disease in hemophilia
patients is typically related to concomitant
916 N.A. Koontz et al.

Fig. 38 Coronal
reconstruction from CECT
of the neck in the same
patient as Figs. 36 and 37
performed approximately
6 h later demonstrates
interval progression of
posttraumatic JVT, now
with complete occlusion
and gradual tapering (black
open arrow) of the column
of refluxed iodinated
contrast that retrograde fills
the proximal left IJV

Fig. 39 Axial CT in bone algorithm (a) reveals an oblique extension through the jugular foramen (black arrow).
fracture through the right aspect of the occipital bone in a Axial CECT (b) shows filling defects in the right IJV
pediatric patient who fell from a jungle gym. There is (open black arrows), which are compatible with
diastasis of the occipitomastoid suture (white arrow) and nonocclusive, posttraumatic JVT

vascular catheters, it can also be treatment related, Rheumatologic

seen following therapeutic administration of fac-
tor eight (VIII) inhibitor bypassing activity Uncommonly, rheumatologic disease may result
(FEIBA) and activated recombinant factor VII in JVT. While not strictly limited to patients with
(rFVIIa). underlying rheumatologic disease, the presence of
38 Jugular Vein Thrombosis
prothrombotic antiphospholipid antibodies
(APLAs), such as anticardiolipin antibody and
lupus anticoagulant (a misnomer given its
procoagulant effects in vivo), is associated with
an increased risk of thromboses (venous greater
than arterial) [27]. While seen commonly in
patients with rheumatoid arthritis and systemic
lupus erythematosus, the prevalence of APLAs
within the general population is estimated
between 1 % and 5 %; moreover, they are
observed in 10 % of patients with DVT or pulmo-
nary embolism and in up to 42 % of patients with
stroke [28]. Whether APLAs are truly a cause or
an effect (or both) of thrombosis is not entirely
clear. Nonetheless, the presence of APLAs should
be considered when venous thrombosis occurs in
atypical sites like the IJV.
An extremely rare rheumatologic disorder asso-
ciated with thrombosis is Behçet disease, which
demonstrates vascular involvement in approxi-
mately 25 % of cases. Rarely, Behçet disease has
been implicated in spontaneous JVT [29]. Behçet
patients are at an increased risk of catheter- or
venipuncture-associated endothelial injury and
may benefit from vascular evaluation with nonin-
vasive magnetic resonance angiography (MRA) or
MRV in order to prevent iatrogenic vascular com-
plications (such as posttraumatic pseudoaneurysm
or catheter-related venous thrombosis).
Gynecologic
Nonneoplastic gynecologic etiologies make up a
small, but important source of JVT and are
observed in three patient subgroups: (1) women
undergoing in vitro fertilization or other assisted
reproductive techniques (ART) therapy,
(2) women with ovarian hyperstimulation syn-
drome (OHSS), and (3) and pregnant women.
One of the striking features of JVT of gynecologic
etiology is that it frequently occurs in otherwise
healthy women of reproductive age rather than the
stereotypical chronically ill (cancer, hemodialy-
sis, etc.) or elderly patients that more commonly
develop JVT.
Firstly, it is important to note that JVT or upper
extremity deep vein thrombosis (UEDVT) is
917
incredibly rare in patients undergoing ART,
occurring in 0.08–0.11 % of ART treatment cycles
[30]. As part of ART, patients receive exogenous
gonadotropins and gonadotropin-releasing drugs
in order to bolster the development and number of
ovarian follicles, which can be complicated by
OHSS. The OHSS may be severe in 1–2 % of
patients, resulting in significant ovarian enlarge-
ment, ascites, abdominal bloating, nausea, pleural
effusions, hypoalbuminemia, hypercoagulability,
venous thromboembolism (VTE), renal failure,
hypovolemic shock, respiratory distress, and
even death [31].
While the majority of ART patients who develop
JVT or UEDVT have preceding OHSS, studies
have found that a sizable minority (up to 29 % in
one series) do not have clinically apparent OHSS
[30]. The exact mechanism by which ART contrib-
utes to hypercoagulability (and subsequent JVT) is
not completely understood and is further compli-
cated by the variability in concurrent OHSS. The
pathogenesis of thrombotic disease with OHSS
likely results from hormonal effects of human cho-
rionic gonadotropin (hCG) and estrogen, which
lead to increased vascular permeability, leukocyte
margination, hemoconcentration, thrombocytosis,
and reduced fibrinolysis [31, 32]. Coupled
with immobility and recumbent positioning (due
to abdominal discomfort from ascites), use
of CVADs, as well as any underlying inherited
thrombophilic disorders (cf. Hematologic), it
becomes readily apparent how these women may
be tipped toward a prothrombotic state that results in
JVT [32].
Unassisted pregnancy (without ART) can cer-
tainly be viewed as a hypercoagulable state, and
spontaneous JVT has been reported rarely in oth-
erwise healthy pregnancies [30]. Nonetheless, the
presence of thrombotic disease within the IJV,
subclavian vein, or upper extremity deep veins
in pregnancy should warrant further investigation
into underlying thrombophilia.
Lastly, the diagnosis of JVT of gynecologic
etiology warrants careful evaluation of thrombus
extension, as well as assessment for complica-
tions. For reasons that are not entirely clear, preg-
nant women with JVT appear to have a
predilection for thrombus propagation or
918
pulmonary embolism even in the face of “appro-
priate” therapy with heparin or low-molecular-
weight heparin (LMWH), a risk reported between
14 % and 20 % [30]. The mortality rate of
gynecologic-related JVT is unknown, but given
the risk of complications, particularly in gravid
women, inpatient hospital management is
advised. In this special patient population, therapy
requires the use of heparin or LMWH, as warfarin
crosses the placental barrier and is contraindicated
in pregnancy.
Endocrine
True endocrine causes of JVT are rare. As previ-
ously discussed, OHSS and ART are important
endocrine-related gynecologic causes of JVT,
which can affect women of childbearing age. Far
less commonly, it has been reported that patients
with Cushing syndrome (CS) have an increased
risk of VTE, including JVT, which is probably
due to increased levels of factor VIII, factor IX,
von Willebrand factor, fibrinogen, and plasmino-
gen activator inhibitor-1 [33]. There also appears
to be a greater risk of postoperative VTE follow-
ing transsphenoidal resection of pituitary lesions
in ACTH-dependent CS as compared to ACTH-
independent CS patients [33].
JVT can also be associated with endocrine neo-
plasia. Numerous cases of thyroid carcinoma
directly invading the IJV have been reported in the
literature [34–36]. In such cases, increased vascu-
larity within the thrombus may be demonstrated on
color or power Doppler US [34]. Similarly, intra-
clot enhancement on CT or MRI should raise con-
cern for tumor thrombus in this patient population
rather than bland thrombus. Additionally, 131I,
67
Ga, and 18FDG uptake can be seen within thyroid
cancer tumor thrombus, although 18FDG avidity is
not specific to thyroid carcinoma or a neoplastic
process [35, 36]. Clinically, tumor thrombus in thy-
roid carcinoma patients portends a poor prognosis,
but the actual clinical presentation may vary from
an asymptomatic incidental finding to superior vena
cava syndrome [36]. Lastly, benign thyroid masses,
N.A. Koontz et al.
such as a goiter, can lead to JVT due to mass effect
and altered blood flow dynamics.
Idiopathic
In most scenarios with knowledge of the clinical
context, a causative etiology of JVT will be deter-
mined in the majority of affected patients. An
extremely rare entity, spontaneous (idiopathic)
JVT should be viewed as a diagnosis of exclusion.
There are no well-established guidelines or con-
firmed evidence in the literature to suggest what
percentage of patients lack a discrete etiology for
JVT following appropriate clinical work-up.
However, given the high association with under-
lying malignancy, caution is advised when con-
sidering a diagnosis of idiopathic JVT.
JVT in Pediatric Patients
Venous thrombosis (in particular JVT) in children
is uncommon and most frequently due to CVADs
and infectious etiologies. Other etiologies of JVT in
pediatric patients include postsurgical mass effect
from cervical or upper mediastinal masses, hema-
tologic disorders, trauma, sequelae of congenital
heart disease, and post-intravenous immunoglobu-
lin infusion [37, 38] (Figs. 39 and 40). A potential
association between cystic fibrosis and venous
thrombosis has also been reported.
Perhaps even more so in children than in
adults, central venous catheters are the leading
culprit for causing JVT [38] (Fig. 41). There is
wide variation in the reported incidence of
catheter-related venous thrombosis in children,
but concurrent CVAD use has been reported in
up to 60 % of children and up to 90 % of neonates
with venous thrombosis [38]. Studies have found
that nearly 90 % of these DVTs involve the IJV,
roughly paralleling the distribution of central
venous catheter location [37]. The sensitivity
(20 %) and specificity (86 %) of clinical diagnosis
(i.e., patient history and physical exam) of DVT in
these patients is lacking, thus underscoring the
38 Jugular Vein Thrombosis
Fig. 40 Coronal CECT (a) reveals a filling defect com-
patible with an intraluminal thrombus in the right trans-
verse and sigmoid sinuses (white arrows) in this recently
postoperative patient as noted by the extra-axial fluid and
Fig. 41 Digital subtraction angiography in a pediatric
patient shows a filling defect (black open arrow) in the
proximal right IJV as well as multiple venous collaterals
(black arrows) typical of chronic, catheter-associated JVT
importance of radiologic imaging [37]. Due to its
noninvasive technique and lack of ionizing radia-
tion, US remains the modality of choice for imag-
ing pediatric patients in whom there is a concern
for JVT.
As previously described (cf. “Infectious”),
infectious causes of JVT can affect children.
919
air (white asterisk) overlying the right cerebral hemisphere.
Coronal CECT (b) shows extension of the sigmoid sinus
thrombus into the right IJV (black arrowhead)
A serious, potentially fatal cause of JVT is
Lemierre syndrome, starting from pharyngitis in
a previously healthy child (Fig. 42). JVT has also
been reported in children following mastoiditis.
Imaging
CT
Contrast-enhanced CT (CECT) is a reliable means
of evaluating JVT or, as is often the case, identi-
fying incidental JVT. Just as the clinical presenta-
tion of JVT varies with chronicity of thrombus, so
too does the imaging appearance (Table 4). With
acute JVT, CECT may demonstrate an enlarged
IJV with an intraluminal filling defect (Fig. 43).
Clot attenuation is variable depending on the state
of the hemoglobin, but it can be hyperdense in the
acute phase, possibly masking the finding against
the backdrop of intravascular contrast. For this
reason, it is helpful to broaden the windowing on
the PACS monitors, effectively allowing the radi-
ologist to “see through” the contrast, similar to
evaluating a CT angiogram of the chest for pul-
monary emboli (Fig. 44). While bland (i.e., not
associated with tumor) thrombus does not
920 N.A. Koontz et al.

Fig. 42 Axial CECT (a)

demonstrates a peripherally
enhancing, centrally
hypodense collection (black
asterisk) centered in the left
palatine tonsil. Axial CECT
(b) shows obscuration of
the musculofascial planes
between the carotid,
retropharyngeal, and
prevertebral spaces due to
phlegmon (black open
arrows). The left IJV is
partially thrombosed with a
residual crescent of luminal
opacification (white
arrowhead)

Table 4 CT findings of jugular vein thrombosis (JVT)
Acute JVT
NECT
Frequently normal
Occasionally high-density thrombus
Secondary finding of retropharyngeal edema
CECT
Filling defect within an enlarged internal jugular
vein (IJV)
Thin rim of contrast enhancement at periphery
of clot
Rare enhancing thrombus (bland or tumor
thrombus)
Secondary finding of retropharyngeal edema
Chronic JVT
NECT
Prominent bypass collaterals
Return to normal or subnormal IJV caliber
Variable density clot, typically decreases with age
+/ Resolution of associated edema
CECT
Filling defect within subnormal to normal size IJV
+/ Enhancement within thrombus
Recanalization channels
Tumor thrombus
Prominent bypass collaterals
+/ Resolution of associated edema
typically enhance, a thin rim of contrast enhance-
ment can be seen at the periphery of the thrombus,
secondary to opacification of the engorged vasa
vasorum. In the setting of malignancy, the finding
of true thrombus enhancement must be viewed
with suspicion for underlying tumor, although
the finding is not entirely pathognomonic
(Figs. 45 and 46). In the acute phase, there is
commonly perivascular soft tissue edema with
obscuration of the musculofascial planes and sur-
rounding fat stranding. Additionally, edema/fluid
within the retropharyngeal space may be seen.
In the setting of chronic JVT, there is typically
a decrease or resolution of the associated
perivascular edema, as well as a return to normal
or even subnormal IJV luminal diameter (Fig. 47).
Bland thrombus commonly remains
nonenhancing and the burden of thrombus is
often better delineated. Recanalization of the
thrombus with internal channels of contrast
enhancement can be seen in some cases, so one
must exercise caution when definitively reporting
enhancing clot as tumor thrombus. Prominent
venous collaterals may serve as a physiologic
clot bypass and their presence argues for chronic-
ity of thrombus (Fig. 48).
Due to the timing of the contrast bolus on
conventional CECT, mixing artifact within the
IJV can occur, which may mimic thrombus
(Fig. 49). In such cases, secondary findings may
provide a useful discriminator to distinguish
between artifact and true pathology. CT angiogra-
phy/venography can be considered if findings on
conventional CECT remain equivocal.
JVT can also be diagnosed prospectively on non-
contrast-enhanced CT (NECT), although the
38 Jugular Vein Thrombosis
Fig. 43 Axial CECT (a) and coronal reconstruction from
CECT (b) of the neck in a 45-year-old man with right-sided
neck pain and swelling for a few days demonstrate an acute
JVT with complete occlusion of an enlarged right IJV,
Fig. 44 Coronal chest CECT (a, b) in a pediatric patient
with a thrombogenic right IJV catheter. The narrowing of
the window-level settings on PACS masks the underlying
sensitivity is suboptimal. As above, thrombus may
be hyperdense in the acute phase, but clot attenua-
tion often decreases as it ages. For this reason, true
thrombus visualization on NECT is often limited.
Instead, radiologists must rely more heavily on
secondary signs to make the diagnosis of JVT on
NECT, such as perivascular and retropharyngeal
edema in the acute phase and bypass collaterals in
the chronic phase.
921
including a central filling defect (white arrows) and
marked enhancement of the vasa vasorum (white open
arrows)
right JVT. Broadening of the window-level settings on
PACS (b) reveals the underlying right JVT (white arrow),
a subtle “corner finding” on this examination
US
Ultrasound (US) is an excellent means of evalu-
ating for JVT (Table 5). As it is noninvasive,
sensitive, quick, (relatively) inexpensive, porta-
ble, and lacking ionizing radiation, it can be
argued that duplex ultrasonography should be
the first-line modality for evaluating JVT, partic-
ularly in children, who are more radiation
922
Fig. 45 Coronal T1-WI –C (a) and T1-WI +C (b) in an
81-year-old woman with extensive left IJV tumor throm-
bus from a glomus jugulare paraganglioma. Coronal
T1-WI –C (a) demonstrates expansion of the entire left
IJV (partially out of plane of the image) with tumor (black
arrows) that is slightly hyperintense to skeletal muscle.
Fig. 46 Axial T1-WI –C (a) and T2-WI FS (b) in the same
patient as Fig. 45. Axial T1-WI –C image (a) at the level of
the oropharynx demonstrates expansion of the left IJV with
tumor (white arrow) that is slightly hyperintense to skeletal
muscle. Focal areas of high intrinsic T1 signal (black
arrowheads) correspond with subacute hemorrhage,
manifesting as the “salt” of this “salt and pepper” lesion,
sensitive than adults are. In the acute phase of
JVT, grayscale US shows an enlarged, incom-
pletely compressible IJV filled with hypoechoic
thrombus (Fig. 50). Color or power Doppler US
demonstrates absent flow within bland thrombus,
but intravascular flow can be seen at the margins
N.A. Koontz et al.
Tubular low signal intensity within the tumor corresponds
with high-velocity flow voids (white open arrow), typical
of a paraganglioma. Coronal T1-WI +C (b) demonstrates
avidly enhancing paraganglioma tumor (white arrows)
with high-velocity flow voids (open white arrows) filling
the left IJV
whereas tubular hypointense structures represent high-
velocity flow voids (white open arrow), which comprise
the “pepper.” (b) Axial T2-WI at the level of the orophar-
ynx demonstrates the mass (white arrow) to be
hyperintense to skeletal muscle with multiple internal
high-velocity flow voids (white open arrows)
of the clot in cases of nonocclusive thrombus.
Absence of the normal biphasic wave form on
spectral Doppler analysis is a useful secondary
imaging feature.
In the chronic phase of JVT, decreased luminal
caliber and increased venous collaterals can be
38 Jugular Vein Thrombosis
Fig. 47 Axial CECT of the neck in a 57-year-old woman
with chronic right JVT demonstrates occlusive thrombus
(white arrow) in a narrow caliber IJV as well as enhance-
ment of the vasa vasorum (white open arrow) and a prom-
inent right external jugular vein (white arrowhead) serving
as a bypass collateral
seen on US, just as is seen with CT. Color and
power Doppler modes can be helpful for demon-
strating venous collaterals. The imaging charac-
teristics of the clot also change with time, with
chronic thrombus typically demonstrating
increased echogenicity on grayscale US.
MRI
Less commonly performed for primary detection
of thrombus, MRI is a useful adjunct modality for
assessment of JVT (Table 6). Signal intensity of
the thrombus is dependent upon its molecular
contents, particularly the state of hemoglobin.
On non-contrast T1-weighted imaging (T1-WI),
JVT demonstrates isointensity in the acute phase
and increases in signal intensity in the subacute
phase due to increasing methemoglobin, which is
first intracellular and then extracellular (Figs. 13
and 15). Adding fat saturation to the T1-WI will
increase the conspicuity of the intraluminal
thrombus by nulling the signal intensity from the
perivascular fat. On post-contrast T1-WI,
non-enhancing low signal intensity thrombus
923
Fig. 48 Coronal reformat of CECT of the neck in the same
patient as Fig. 47 demonstrates occlusive thrombus (white
arrow) in a narrow caliber IJV as well as enhancement of
the vasa vasorum (white open arrow) and a prominent right
external jugular vein (white arrowhead) serving as a
bypass collateral. The tip of a potentially thrombogenic
central venous catheter can be seen inferiorly as a “corner
finding” (black arrow)
within an enlarged IJV is typical of acute throm-
bus (Figs. 9, 14, 29, 32, and 33). As with CECT, a
thin rim of peripheral contrast enhancement can
be seen at the margin of the IJV lumen in acute
JVT secondary to engorged vasa vasorum. In
chronic JVT, post-contrast T1-WI reveals a nor-
mal or subnormal-sized IJV with non-enhancing
filling defect in most cases of bland thrombus.
Recanalization channels within the clot may be
seen. Just as with CT, the presence of enhancing
clot in the setting of tumor-related JVT should
raise the concern for vascular invasion and
tumor thrombus. Also just like CT, the presence
of prominent venous collaterals is a useful dis-
criminator for determining the chronicity of a clot.
On T2-weighted imaging (T2-WI), thrombus can
924
Fig. 49 Mixing artifact mimicking JVT. CECT of the
neck in a 51-year-old woman demonstrates an apparent
filling defect in the left IJV (white arrow) that looks like
a nonocclusive thrombus on axial image (a). However,
careful review of the coronal images (b) demonstrates the
“filling defect” to have a more feathery appearance, and
Table 5 Ultrasound findings of jugular vein thrombosis
(JVT)
Acute JVT
Enlarged, incompletely compressive internal jugular
vein (IJV)
Hypoechoic luminal content – thrombus
Absent flow within thrombus on color or power
Doppler
May maintain flow at periphery if nonocclusive
thrombus
Absent biphasic wave form on spectral Doppler
Chronic JVT
Decreased luminal caliber
Prominent bypass collaterals
Increased echogenicity of thrombus
Internal flow within thrombus secondary to
recanalization/organization
be bright in the hyperacute phase, becoming
hypointense in the acute phase due to the
deoxyhemoglobin (Fig. 14). The T2 signal inten-
sity becomes more hypointense as the clot ages.
Adding fat saturation to conventional fast spin
echo, T2-WI is particularly useful for assessing
secondary findings of JVT, including edema
within the perivascular soft tissues and the
retropharyngeal space. Marked low signal inten-
sity susceptibility artifact (“blooming” artifact)
can be seen within the thrombus on T2*
gradient-recalled echo (GRE) imaging.
N.A. Koontz et al.
there is incomplete opacification of the more proximal IJV
(c) in keeping with mixing of contrast-opacified and non-
contrast-opacified blood. In equivocal cases, delayed-
phase imaging or second-look ultrasound are useful for
characterization
MR venography (MRV) can be utilized for
evaluating the extent of thrombus. Traditionally,
this has been performed utilizing two-dimensional
time-of-flight (TOF) imaging which does not
require administration of intravenous contrast.
On TOF MRV, a saturation pulse is applied to
diminish signal from the stationary soft tissues
and, importantly, from the inflowing arterial blood.
The lack of flow-related enhancement at the level of
the JVT manifests as segmental signal absence
within the region of the IJV (Figs. 10 and 11).
A potential pitfall of TOF MRV is subacute
thrombus, which contains methemoglobin and
consequently demonstrates intrinsic high signal.
Dynamic contrast-enhanced MRV (e.g., 3D spoiled
gradient echo (SPGR) technique) has become more
widely used and is less susceptible to this sort of
diagnostic misinterpretation.
Nuclear Medicine
Prior to the widespread use of cross-sectional (CT,
MR, or US) imaging for assessment of upper or
lower extremity venous thrombosis, radionuclide
venography was a relatively common practice.
With this technique, a radioactive blood pool
(i.e., 99mTc-labeled red blood cells) or venous
blood flow (i.e., 99mTcO4 , 99mTc-MAA, or
99m
Tc-DTPA) agent is used to evaluate for
38 Jugular Vein Thrombosis 925

Fig. 50 Grayscale ultrasound (a) of the right IJV in the thrombus. Color Doppler ultrasound (b) of the right IJV in
longitudinal plane demonstrates an eccentric, echogenic, the longitudinal plane illustrates flow in the vessel around
intraluminal filling defect (white arrow) in an incompletely the small intraluminal thrombus
compressible vessel. This is concerning for a nonocclusive

Table 6 MRI findings of jugular vein thrombosis (JVT) patency of an affected venous channel. Reduced
radiotracer activity within a vein relative to the
T1-WI –C FS
Acute – isointense thrombus
contralateral side or relative to an ipsilateral con-
Early subacute – hyperintense due to intracellular tiguous venous segment is suggestive of venous
methemoglobin thrombosis, as is delayed transit time, venous
Late subacute – hyperintense due to extracellular reflux, and increased activity within collateral
methemoglobin channels [39]. While this technique is utilized
T1-WI +C primarily for detection of DVT involving the
Acute extremities, it theoretically could be applied for
Hypo-/isointense thrombus in enlarged internal the evaluation of the deep veins of the neck and
jugular vein (IJV)
might be considered in cases when cross-sectional
+/ Thin rim of peripheral enhancement
imaging cannot be performed or is nondiagnostic.
Chronic
Filling defect within subnormal to normal caliber
Currently relegated to a historical footnote in
IJV DVT imaging, nuclear medicine studies may be
+/ Thrombus enhancement making somewhat of a comeback. While cross-
Organized clot sectional imaging modalities are the first-line
Tumor thrombus modality for making the diagnosis of JVT, new
T2-WI molecular imaging techniques may prove valu-
Hyperacute – hyperintense thrombus able for determining the acuity of a thrombus.
Acute – hypointense thrombus The basic concept is simple: find and exploit a
Early subacute – hypointense thrombus molecular marker that is present in acute throm-
Late subacute – hyperintense due to extracellular bus, but which is not found in organized clot.
methemoglobin
Several radiopharmaceuticals have been targeted,
Chronic – diminishing T2 signal as clot ages
including radiolabeled fibrinogen, platelets,
GRE
Marked low signal intensity
antiplatelets, antifibrin, fibrinolytic enzymes
Often “blooming” artifact (streptokinase and urokinase), soluble fibrin, frag-
MRV ment E1, plasmin, tissue plasminogen activator,
Acute – filling defect in IJV heparin, and multiple synthetic peptides [40]. Cur-
Chronic – filling defect with bypass collaterals rently, one such agent, 99mTc-apcitide (AcuTect,
Diatide Inc, Londonderry, NH), a GPIIb/IIIa
926
Fig. 51 A forty-six-year-old with history of pancreatic
adenocarcinoma and catheter-associated thrombus in the
right IJV. Coronal reconstruction of CECT (a) demon-
strates occlusive filling defect in the right IJV (white
arrow) with enhancement of the vasa vasorum (white
open arrow). Coronal fused FDG-PET/CT image (b) dem-
onstrates a rim of increased FDG avidity at the periphery of
the clot (white arrowhead), corresponding to the engorged
vasa vasorum and surrounding perivascular inflammation.
Coronal FDG-PET image (c) demonstrates that the periph-
eral FDG avidity (white arrowheads) is similar in activity
receptor antagonist, is approved by the Food and
Drug Administration for clinical use, and other
similar agents remain under investigation
[41]. While the efficacy of such molecular agents
has not been reported specific to JVT (solely
reported for extremity DVT), it stands to reason
that the mechanism of action (e.g., binding of
activated platelets in acute thrombosis) should
allow for detection of acute thrombus in the IJV.
More commonly, JVT is encountered as an inci-
dental finding on nuclear medicine studies
performed for other reasons. In such cases, radio-
tracer accumulation along the lateral neck localizes
to the thrombus within the IJV. This is probably
encountered with greatest frequency on FDG-PET/
CT, but one must keep in mind that FDG accumu-
lation is nonspecific to etiology, occurring in areas
with neoplasm, infection, and inflammation [42, 43]
(Figs. 51 and 52). On the other hand, a significant
portion of JVT are not metabolically active on
FDG-PET, reflecting bland thrombus of
N.A. Koontz et al.
to background hepatic activity but slightly increased com-
pared to blood pool activity. Lack of FDG uptake within
the clot itself argues against tumor thrombus. Incidental
note is made of a radiolabeled pulmonary embolus in the
left lung (black arrowhead on b and c), the so-called “hot
clot” sign. This is due to accidental tagging of catheter-
associated thrombus at time of intravenous 18F-FDG
administration, which subsequently embolized to the lung
(Image courtesy of Steven Westphal, M.D., Indiana Uni-
versity School of Medicine, Department of Radiology and
Imaging Sciences)
heterogeneous etiologies: postsurgical, catheter-
associated, posttraumatic, ART- or pregnancy-
associated, APLA-associated, or even malignancy-
associated hypercoagulability [43]. Given lack
of specificity of FDG-PET for both metabolically
active and non-metabolically active thrombi,
JVT encountered on FDG-PET studies requires
diligent review of associated anatomic imaging
and correlation with clinical history.
Detection of IJV tumor thrombus by 131I scin-
tigraphy for papillary thyroid carcinoma and 67Ga
scintigraphy for anaplastic thyroid carcinoma has
been reported [35, 36]. Radioiodine-avid throm-
bus has a high specificity for well-differentiated
(papillary, follicular, or mixed origin) thyroid neo-
plasm but not for anaplastic or medullary thyroid
carcinoma due to lack of iodine concentration in
these tumors. While the finding of 67Ga accumu-
lation in the IJV should raise concern for JVT, its
specificity for anaplastic thyroid carcinoma as a
primary tumor of origin is questionable.
38 Jugular Vein Thrombosis 927

Fig. 52 A forty-six-year-old with history of pancreatic
adenocarcinoma and catheter-associated thrombus in the
right IJV. Axial CECT image (a) demonstrates occlusive
filling defect in the right IJV (white arrow) with enhance-
ment of the vasa vasorum (white open arrow). Axial
FDG-PET image (b) demonstrates a rim of increased FDG
avidity at the periphery of the clot (white arrowhead),
corresponding to the engorged vasa vasorum and surround-
ing perivascular inflammation. Lack of FDG uptake within
the clot itself argues against tumor thrombus (Image cour-
tesy of Steven Westphal, M.D., Indiana University School of
Medicine, Department of Radiology and Imaging Sciences)

Angiography

Retrograde jugular venography was originally

described by Swedish authors in the early 1960s
and pioneered throughout the remainder of the
decade. Prior to the introduction of cross-sectional
imaging, it was utilized primarily for evaluation of
carotid space and base of skull tumors but is also
efficacious for detection of JVT (Fig. 53). A sen-
sitive, but invasive procedure typically employing
common femoral vein puncture and placement of
an angiographic catheter in the internal jugular
vein, conventional retrograde venography is
rarely employed for primary detection of JVT in
current clinical practice. Additionally, retrograde
venography runs the risk of dislodging the throm-
bus with subsequent downstream embolization if
not performed carefully.
Fig. 53 Digital subtraction angiography (DSA) of the left
Presently, JVT is identified by interventional upper extremity in a 45-year-old woman with “uphill
radiologists during the exchange of thrombosis” of the left IJV demonstrates filling defects
malfunctioning central venous catheters due to (black arrows) within the left subclavian and
brachiocephalic veins with no retrograde filling of the
the high frequency of catheter-associated throm-
completely occluded left IJV. Numerous IJV bypass col-
bus (Fig. 41). The actual imaging finding of JVT laterals (black open arrows) are also seen (Image courtesy
on conventional venography parallels that of of Perry P.Ng, MBBS, University of Utah Health Sciences,
cross-sectional modalities with a discrete filling Department of Radiology)
defect found within the IJV, which may be occlu-
sive or nonocclusive. As observed on CT, MRI, Although successful endovascular interven-
and/or US, acute clot tends to enlarge the vessel, tional treatments for JVT have been reported,
whereas chronic JVT typically results in a smaller medical management with anticoagulation is
caliber vessel with prominent venous bypass favored (Fig. 54). In a small study of 12 patients
collaterals. with symptomatic, occlusive JVT, catheter-
928
Fig. 54 (a) Digital subtraction angiography (DSA) just
prior to left IJV suction thrombectomy demonstrates filling
defects (black arrows) within the left IJV as well as numer-
ous bypass collaterals (black open arrows). The suction
thrombectomy catheter is noted (white open arrow in both
a and b). DSA following suction thrombectomy (b)
directed thrombolysis utilizing recombinant tissue
plasminogen activator was shown to be safe and
effective for treatment of JVT regardless of phys-
ical length of clot [44]. Thrombolysis was more
effective in patients with acute thrombus (less
than 2 weeks in age), and 50 % of treated patients
demonstrated a patent IJV at 2-month follow-up.
However, IJV patency was maintained in only
25 % of patients with persistent CVAD use
[44]. Successful mechanical thrombectomy of
JVT has also been reported in small patient
populations with several devices, including the
use of the Oasis mechanical thrombectomy sys-
tem (Medi-Tech/Boston Scientific, Watertown,
MA), rotatable pigtail catheters, and thromboas-
piration techniques [45]. Further investigation
into the safety, efficacy, and durable results of
endovascular interventional therapy of JVT will
be necessary to determine its place in the manage-
ment of this disorder.
Diagnostic Pitfalls
While identifying the finding of JVT is often
straightforward (especially when one remembers
to look for it), there are several pitfalls that may
result in the misdiagnosis of JVT. As with other
N.A. Koontz et al.
demonstrates marked improvement in opacification of the
left IJV with minimal residual filling defects (black arrow)
as well as reduction in venous shunting through numerous
bypass collaterals (black open arrows) (Image courtesy of
Perry P.Ng, MBBS, University of Utah Health Sciences,
Department of Radiology)
facets of medical imaging, a few broad categories
of errors may account for misdiagnosis: failures in
systems-based practice, errors of perception,
errors of judgment, and technical limitations/arti-
facts. For instance, a simple clinical history, such
as “neck pain and swelling – concern for JVT,”
would be extremely helpful for directing the
interpreting radiologist to search for JVT. Unfor-
tunately, the presentation of JVT is variable and
the diagnosis may not be considered by the refer-
ring clinician. Furthermore, even if the clinician is
suspicious of JVT, it is certainly not uncommon
for pertinent history to become truncated along
the workflow pathway from the ordering clinician
to the interpreting radiologist (a failure in
systems-based practice). Misinterpretation of
imaging studies may be due to lack of specific
attention to the appearance of the IJV (error of
perception), which lies toward the edge of the
images rather than in the center of the reader’s
focus. Furthermore, the proximal and distal por-
tions of the IJV are at the beginning and end of an
imaging series, respectively, and may be inadver-
tently overlooked. To avoid misdiagnosis of JVT,
it is important to include venous structures in the
search pattern when reviewing studies. Some-
times the radiologist recognizes the findings of
JVT, such as increased luminal signal on MRI,
38 Jugular Vein Thrombosis
but interprets the pathologic findings as some-
thing benign, such as slow flow (error in judg-
ment). Lastly, all imaging modalities have
intrinsic technical limitations and artifacts that
may obfuscate the diagnosis of JVT, and it is
important to be aware of these to improve diag-
nostic accuracy. Although most of the literature
description of missed diagnoses (i.e., false-
negative studies) and misdiagnoses (i.e., false-
positive studies) of venous thrombosis in neuro-
imaging involves the dural venous sinuses and
cerebral veins, the technical reasons such misin-
terpretations occur are also applicable to JVT
imaging.
False-Negative Misinterpretation
On unenhanced CT, it may be challenging to
delineate the tubular lumen of the IJV from the
surrounding soft tissue structures, particularly
lymph nodes. Additionally, evolution of
intraluminal blood products may cause the throm-
bus to appear isodense to adjacent soft tissues
rather than hyperdense. Occasionally on CECT,
a high-density thrombus may not be visible
against the background of contrast opacification,
wall enhancement (vasa vasorum), or partial
recanalization of the IJV [46]. There may also be
a false-negative study if the timing of scanning
after contrast administration is inappropriate. For
example, if the CECT is performed too soon after
contrast administration, there will be decreased
contrast resolution between the thrombus and the
opacified blood.
On MRI, the challenge of diagnosing JVT lies
in the fact that the signal intensity of thrombus can
simulate normal blood flow. For example, acute
thrombus may appear hypointense on T2-WI,
simulating a flow void of a patent vessel. As the
blood products evolve into methemoglobin, the
thrombus may appear hyperintense on T1-WI and
TOF MRV. As a word of caution, TOF MRV
should always be read in conjunction with the
other MRI sequences to avoid misinterpreting
the hyperintense signal in the thrombus as flow-
related enhancement. Lastly, on contrast-
enhanced T1-WI, one typically expects JVT to
929
present as a filling defect within the IJV. While a
filling defect is a common appearance of throm-
bus, a chronically thrombosed IJV can demon-
strate homogeneous enhancement because of
capillary channels within the clot.
False-Positive Misinterpretation
Although very difficult to call prospectively on
NECT, especially in the absence of secondary
findings such as edema in the surrounding soft
tissues or bypass collaterals, hyperdense throm-
bus may occasionally be identified. However, one
must keep in mind that hyperdense clot within the
IJV may be masked if a patient has an elevated
hematocrit, which may be due to dehydration or a
hypercoagulable status. Hemoconcentration has
been shown to correlate with CT attenuation
[47]. Absolute Hounsfield unit values greater
than 65, Hounsfield unit to hematocrit ratios
greater than 1.7, and venoarterial Hounsfield unit
value differences greater than 15 are useful
adjuncts to NECT of the head for the diagnosis
of dural venous thrombosis [48]. Although correl-
ative values have not been reported with regard to
JVT, it stands to reason that similar findings
should be seen. Lastly, in pediatric patients, one
must keep in mind their physiologically normal,
higher hematocrit, which could mimic thrombus.
When in doubt, one should seek other clues that
the hyperdensity is related to hemoconcentration
rather than true thrombosis, such as homogeneity
of the hyperattenuation and symmetry of involve-
ment of all other venous channels.
On CECT, one of the most common reasons for
a false-positive finding of JVT is the so-called
mixing artifact, which is caused by mixing of
contrasted and non-contrasted blood and results
in the appearance of a filling defect [49]. Mixing
artifact typically demonstrates a swirling, irregu-
lar, or discontinuous configuration, while true
thrombus typically is well defined, linear, and
contiguous (Figs. 49 and 55). Although this usu-
ally does not result in a major diagnostic dilemma,
delayed-phase imaging or second-look US are
useful options to resolve true thrombus from
such a pseudolesion in difficult cases. Other
930
Fig. 55 Coronal reconstruction of CECT of the neck
demonstrates the typical appearance of “mixing” artifact.
The appearance of feathery, swirling low attenuation
within the bilateral IJVs (white open arrows) is dependent
on the timing of the contrast bolus and results from mixing
of contrast-opacified and non-contrast-opacified blood.
This should not be mistaken for thrombus, and in equivocal
cases, delayed-phase imaging or second-look ultrasound
can be considered for further characterization
common reasons for the false-positive diagnosis
of JVT include vascular compression and surgical
ligation. In the neck, the IJV is a thin-walled,
pliable structure, which may be extrinsically com-
pressed by surgical or radiation treatment-related
scar or by a cervical soft tissue mass, including
lymphadenopathy, which can give the appearance
of occlusion. Alternatively, it may be ligated in
certain types of surgical dissections. One must
keep in mind that the absence of contrast
opacification of the IJV does not equate to throm-
bosis, and it is important to follow the IJV through
its course to identify the location of external com-
pression and return to normal luminal caliber.
On MRI, false-positive cases can be seen with
increased IJV signal intensity due to slow flow,
turbulent flow, in-plane flow, and entry-flow phe-
nomenon on spin echo sequences, most com-
monly (Fig. 56). In-plane flow phenomenon
refers to flow parallel to the plane of acquisition,
which results in loss of signal due to saturation of
N.A. Koontz et al.
Fig. 56 Axial T1-WI +C at the level of the right superior
jugular bulb exhibits a focal intraluminal hypointensity,
which is suggestive of a filling defect (white arrowhead).
However, this artifactual pseudolesion is caused by turbu-
lent flow and can easily be misdiagnosed as JVT. In equiv-
ocal cases, SPGR +C or CTV can be performed for
clarification
signal from inflowing protons. In the coronal
plane, this can be seen in the IJV and distal sig-
moid sinus. Phase-contrast venography can be
helpful to differentiate between true occlusion
and slow flow as it demonstrates blood flow char-
acteristics, including directionality, and is not
affected by methemoglobin hyperintensity.
Lastly, there is often asymmetry of the path-
ways of cerebral venous return, including the
dural venous sinuses and internal jugular veins,
which can be seen in up to 25 % of normal sub-
jects and should not be confused for underlying
thrombotic pathology [50] (Fig. 57). Although the
depiction of venous collateral flow has been
described as an indirect diagnostic parameter in
patients with cerebral and jugular venous throm-
bosis, complex variability of cerebrovenous anat-
omy may be confounding and require a
multimodality (CTV, MRV, US, or catheter angi-
ography) approach to reach the appropriate
diagnosis.
38 Jugular Vein Thrombosis
Fig. 57 Coronal maximum intensity projection image
from a dynamic (time-resolved) MRA of the neck demon-
strates anatomic variation of the right IJV, which is focally
narrowed (white arrow). This benign variant is not uncom-
monly seen and should not be confused for JVT
Conclusion
JVT is a not uncommon and serious finding
encountered by imagers. Although not usually a
diagnostic dilemma, the sundry etiologies (iatro-
genic, neoplastic, infectious, traumatic, hemato-
logic, rheumatologic, gynecologic, endocrine,
and idiopathic), the potentially devastating com-
plications (such as pulmonary embolism, septic
emboli, and dural sinus thrombosis), and the high
association with neoplasm warrant additional
investigation into the etiology and associated
complications on the part of the radiologist.
While CECT remains the first-line modality for
imaging JVT, both MRI/MRV and duplex Dopp-
ler US can add important information and play a
valuable role in imaging evaluation of these
patients.
931
References
1. Valecchi D, Bacci D, Gulisano M et al (2010) Internal
jugular vein valves: an assessment of prevalence, mor-
phology and competence by color Doppler echography
in 240 healthy subjects. Ital J Anat Embryol
115(3):185–189
2. Malcom GE, Raio CC, Poordabbagh AP, Chiricolo GC
(2008) Difficult central line placement due to variant
internal jugular vein anatomy. J Emerg Med
35:189–191
3. Dickson BC (2004) Venous thrombosis: on the
history of Virchow’s triad. Univ Tor Med J 81
(3):166–171
4. Pata YS, Ünal M, G€ ulhan S (2008) Internal jugular
vein thrombosis due to distant malignancies: two case
reports and literature review. J Laryngol Otol
122:318–320
5. Gisolf J, van Lieshout JJ, van Heusden K, Pott F, Stok
WJ, Karemaker JM (2004) Human cerebral venous
outflow pathway depends on posture and central
venous pressure. J Physiol 560(Pt 1):317–327
6. Volturo GA, Repeta RJ (2001) Non-lower extremity
deep vein thrombosis. Emerg Med Clin North Am
19(4):877–893
7. Gbaguidi X, Janvresse A, Benicou J, Cailleux N,
Levesque H, Marie I (2011) Internal jugular vein
thrombosis: outcome and risk factors. Q J Med
104:209–219
8. Yardim H, Erkoc R, Soyoral YU, Begenik H, Avcu S
(2012) Assessment of internal jugular vein thrombosis
due to central venous catheter in hemodialysis patients:
a retrospective and prospective serial evaluation with
ultrasonography. Clin Appl Thromb Hemost 18(6):
662–665
9. Arhi CS, Buchanan MA, Allen SA, Pickles J (2010)
Internal jugular vein thrombosis secondary to a perma-
nent cardiac pacemaker: an unusual case of lateral neck
swelling. J Laryngol Otol 124:916–918
10. Yurtseven T, Erşahin Y, Kitiş Ö, Mutluer S (2005)
Thrombosis and thrombophlebitis of the internal
jugular vein as a very rare complication of ventricu-
loatrial shunt. Clin Neurol Neurosurg 107:144–146
11. Agraharkar M, Isaacson S, Mendelssohn D et al (1995)
Percutaneously inserted silastic jugular hemodialysis
catheters seldom cause jugular vein thrombosis.
ASAIO J 41(2):169–172
12. Harada H, Omura K, Takeuchi Y (2003) Patency and
caliber of the internal jugular vein after neck dissection.
Auris Nasus Larynx 30:269–272
13. Sheikh MA, Topoulos AP, Deitcher SR (2002) Isolated
internal jugular vein thrombosis: risk factors and natu-
ral history. Vasc Med 7:177–179
14. Vormittag R, Simanek R, Ay C et al (2009) High factor
VIII levels independently predict venous thromboem-
bolism in cancer patients. Arterioscler Thromb Vasc
Biol 29:2176–2181
932
15. Kanz R, Vukovich T, Vormittag R et al (2011) Throm-
bosis risk and survival in cancer patients with elevated
C-reactive protein. J Thromb Haemost 9:57–63
16. Cheang PP, Fryer J, Ayoub O, Singh V (2004) Sponta-
neous bilateral internal jugular vein thrombosis: a sign
of metastasis. J Laryngol Otol 118:570–572
17. Chen M-H, Chang PM-H, Chen P-M et al (2010)
Prolonged facial edema is an indicator of poor progno-
sis in patients with head and neck squamous cell car-
cinoma. Support Care Cancer 18:1313–1319
18. Veras E, Srodon M, Neijstrom ES, Ronnett BM (2009)
Metastatic HPV-related cervical adenocarcinomas
presenting with thromboembolic events (Trousseau
syndrome): clinicopathologic characteristics of
2 cases. Int J Gynecol Cancer 28:134–139
19. Lin D, Reeck JB, Murr AH (2004) Internal jugular vein
thrombosis and deep neck infection from intravenous
drug use: management strategy. Laryngoscope
114:56–60
20. Karkos PD, Asrani S, Karkos CD et al (2009)
Lemierre’s syndrome: a systematic review. Laryngo-
scope 119:1552–1559
21. Wright WF, Shiner CN, Ribes JA (2012) Lemierre
syndrome. South Med J 105(5):283–288
22. Chirinos JA, Lichtenstein DM, Garcia J, Tamariz LJ
(2002) The evolution of Lemierre syndrome: report of
2 cases and review of the literature. Medicine
81(6):458–465
23. O’Brien WT, Lattin GE, Thompson AK (2006)
Lemierre syndrome: an all-but-forgotten disease. AJR
Am J Roentgenol 187:W324
24. Duke BJ, Ryu RK, Brega KE, Coldwell DM (1997)
Traumatic bilateral jugular vein thrombosis: case report
and review of the literature. Neurosurgery 41
(3):680–683
25. Caplan JM, Kalpey Z, Gates J (2008) Closed traumatic
head injury: dural sinus and internal jugular vein
thrombosis. Emerg Med J 25(11):777–778
26. Mnejja M, Hammami B, Bougacha L et al (2011) Inter-
nal jugular vein thrombosis by protein C activated
resistance. Eur Ann Otorhinolaryngol Head Neck Dis
128:269–271
27. Provenzale JM, Ortel TL, Allen NB (1998) Systemic
thrombosis in patients with antiphospholipid anti-
bodies: lesion distribution and imaging findings. AJR
Am J Roentgenol 170:285–290
28. Misita CP, Moll S (2005) Antiphospholipid antibodies.
Circulation 112:e39–e44
29. Danaci M, Akpolat T, Koyuncu M, Unal R, Belet U
(2000) The advantages of MRI and MRA for diagnos-
ing Behçet’s disease and internal jugular vein throm-
bosis. Comput Med Imaging Graph 24:121–124
30. Chan WS, Ginsberg JS (2006) A review of upper
extremity deep vein thrombosis in pregnancy:
unmasking the ‘ART’ behind the clot. J Thromb
Haemost 4:1673–1677
31. Fleming T, Sacks G, Nasser J (2012) Internal jugular
vein thrombosis following ovarian hyperstimulation
syndrome. Aust N Z J Obstet Gynaecol 52:87–90
N.A. Koontz et al.
32. Arya R, Shehata HA, Patel RK et al (2001) Internal
jugular vein thrombosis after assisted conception ther-
apy. Br J Haematol 115:153–155
33. Stuijver DJF, van Zaane B, Feelders RA et al (2011)
Incidence of venous thromboembolism in patients with
Cushing’s syndrome: a multicenter cohort study. J Clin
Endocrinol Metab 96:3525–3532
34. Panzironi G, Rainaldi R, Ricci F, Casale A, Macciucca
MDV (2003) Gray-scale and color Doppler findings in
bilateral internal jugular vein thrombosis caused by
anaplastic carcinoma of the thyroid. J Clin Ultrasound
31(2):111–115
35. Yoshimura M, Kawamato A, Kakasone K et al (2003)
Gallium-67 accumulation to the tumor thrombus in
anaplastic thyroid cancer. Ann Nucl Med 17
(8):689–691
36. Mugunthan N, Kumar AR, Thomas P (2010) Tumor
thrombus in the great veins from papillary carcinoma
of the thyroid: 131-I scan findings. J Clin Endocrinol
Metab 95:2527–2528
37. Hanslik A, THom K, Haumer M et al (2008) Incidence
and diagnosis of thrombosis in children with short-term
central venous lines of the upper venous system. Pedi-
atrics 122:1284–1291
38. Male C, Kuhle S, Mitchell L (2003) Diagnosis of
venous thromboembolism in children. Semin Thromb
Hemost 29:377–390
39. Khosa F, Otero HJ, Prevedello LM, Rybicki FJ, Di
Salvo DN (2010) Imaging presentation of venous
thrombosis in patients with cancer. AJR Am J
Roentgenol 194:1099–1108
40. Taillefer R (2001) Radiolabeled peptides in the detec-
tion of deep venous thrombosis. Semin Nucl Med
31(2):102–123
41. Bates SM, Lister-James J, Julian JA, Taillefer R, Moyer
BR, Ginsberg JS (2003) Imaging characteristics of a
novel technetium Tc 99m-labeled platelet glycoprotein
IIb/IIIa receptor antagonist in patients with acute deep
vein thrombosis or a history of deep vein thrombosis.
Arch Intern Med 163:452–456
42. Kikuchi M, Yamamoto E, Shiomi Y, Kikuchi M,
Yamamoto E, Shiomi Y et al (2004) Internal and exter-
nal jugular vein thrombosis with marked accumulation
with FDG. Br J Radiol 77:888–890
43. Sopov V, Bernstine H, Stern D, Yefremov N, Sosna J,
Groshar D (2009) The metabolic spectrum of venous
thrombotic disorders found on PET/CT. AJR Am J
Roentgenol 193:W530–W539
44. Chang R, Horne MK, Mayo DJ, Doppman JL (1996)
Pulse-spray treatment of subclavian and jugular venous
thrombi with recombinant tissue plasminogen activa-
tor. J Vasc Interv Radiol 7(6):845–851
45. Tajima H, Murata S, Kumazaki T, Ichikawa K, Tajiri T,
Yamamoto Y (2004) Successful interventional treat-
ment of acute internal jugular vein thrombosis. AJR
Am J Roentgenol 182:467–469
46. Provenzale JM, Kranz PG (2011) Dural sinus throm-
bosis: sources of error in image interpretation. AJR Am
J Roentgenol 196(1):23–31
38 Jugular Vein Thrombosis
47. Black DF, Rad AE, Gray LA, Campeau NG, Kallmes
DF (2011) Cerebral venous sinus density on
noncontrast CT correlates with hematocrit. AJNR Am
J Neuroradiol 32:1354–1357
48. Besachio DA, Quigley EP 3rd, Shah LM, Salzman KL
(2013) Noncontrast computed tomographic Hounsfield
unit evaluation of cerebral venous thrombosis: a quan-
titative evaluation. Neuroradiology 55(8):941–945
933
49. Takhtani D (2005) CT neuroangiography: a glance at
the common pitfalls and their prevention. AJR Am J
Roentgenol 185:772–783
50. Ayanzen RH, Bird CR, Keller PJ, McCully FJ,
Theobald MR, Heiserman JE (2000) Cerebral MR
venography: normal anatomy and potential
diagnostic pitfalls. AJNR Am J Neuroradiol 21
(1):74–78
 Imaging of Cerebral Venous and Sinus
Thrombosis 39
Jennifer Linn

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 936 Cerebral venous and sinus thrombosis (CVST)
constitutes a rare but important cause of stroke
Subtypes of CVST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 936
with an annual incidence of approximately
Clinical Signs and Symptoms . . . . . . . . . . . . . . . . . . . . . . 938 three to four cases per one million adults. In
Laboratory Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 938 recent decades, the clinical outcome of CVST
has significantly improved, largely due to
Prognosis and Outcome . . . . . . . . . . . . . . . . . . . . . . . . . . . . 938
improved neuroimaging methods. However,
Imaging Features of CVST . . . . . . . . . . . . . . . . . . . . . . . . 939 the average interval from symptom onset to
Direct and Indirect Neuroimaging Signs . . . . . . . . . . . . 939
Digital-Subtraction Angiography (DSA) . . . . . . . . . . . . 939 definite diagnosis is still 7 days. This is mainly
Magnetic Resonance Imaging . . . . . . . . . . . . . . . . . . . . . . . 940 due to its subacute onset and its unspecific
Computed Tomography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 945 clinical presentation. To avoid delayed diagno-
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 950 sis, it is essential to consider CVST as a differ-
ential diagnosis in patients presenting with
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 951
unspecific neurological symptoms, particularly
headache, other signs of elevated intracranial
pressure, and seizures.
Magnetic resonance imaging (MRI) and
computed tomography (CT) with CT angiog-
raphy provide excellent techniques for the
diagnosis of sinus or deep cerebral venous
thrombosis. The diagnosis of cortical venous
thrombosis is particularly challenging and
requires T2*WI MRI sequences. Knowledge
of potential diagnostic pitfalls and artifacts is
essential to avoid misinterpretation of MRI
or CT.

Keywords
Cerebral Venous thrombosis • Sinus thrombo-
J. Linn (*)
sis • Imaging • MRI • CT
Institute of Neuroradiology, University Hospital Carl
Gustav Carus, Dresden, Dresden
e-mail: jennifer.linn@uniklinikum-dresden.de

# Springer Science+Business Media New York 2016 935

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_29
936
Introduction
Cerebral venous and sinus thrombosis (CVST) is
a relatively rare but important cause of stroke with
an estimated annual incidence of three to four
cases per one million in adults and approximately
seven cases per one million in children. Although
young and middle-aged women are most com-
monly affected, CVST can occur in all age groups
and in both sexes [1].
The most important predisposing factors and
causes of CVST include (acquired or congenital)
coagulation disorders, intake of oral contraceptive
drugs, pregnancy, lactation, and ear, nose, and
throat infections. In over 40 % of patients, two
Fig. 1 Three subtypes of cerebral venous and sinus
thrombosis (CVST). Based on the affected venous struc-
tures, the following subtypes of CVST can be distin-
guished: dural sinus thrombosis (= thrombotic occlusion
of one or more dural sinuses, indicated in red in a), deep
cerebral venous thrombosis (= thrombosis of the vein of
J. Linn
or more risk factors are present, while in 15–20 %
of all patients, no predisposing factors can be
identified [2].
Subtypes of CVST
Based on the location of the thrombosis, i.e., the
affected venous structures, three subtypes of
CVST should be distinguished, which differ with
regard to their clinical presentation and prognosis
(Fig. 1). In addition, the value of the different
imaging modalities and MRT sequences differs
considerably depending on the site of the
Galen and its tributaries, indicated in blue in b), and corti-
cal venous thrombosis (= thrombosis of the superficial
cortical veins including the anastomotic vein of Labbé
indicated in yellow in c). The straight sinus (arrow in b)
is included with the deep cerebral veins by most authors.
Arrows in (c) indicate the anastomotic vein of Labbé
39 Imaging of Cerebral Venous and Sinus Thrombosis 937

Fig. 2 Deep cerebral

venous thrombosis. T2*WI
gradient-echo images of a
27-year-old female patient
with thrombosis of the
internal cerebral veins
(normal white arrows in
b and c), the vein of Galen
(normal white arrow in d),
the basal vein of Rosenthal
(black arrow in c), and their
tributaries (arrows in a,
dotted white arrow, and
black arrow in d) as well as
of the straight sinus (dotted
arrows in c and d). Dotted
arrow in (b) indicates the
right septal vein and black
arrow in (b) indicates the
left thalamic vein. The
venous clot is delineated as
linear hypointense signal
within the affected veins on
T2*WI images (a–d). Note
that imaging quality is
impaired due to motion
artifacts in this patient
presenting with impaired
consciousness

thrombosis [3]. Combination of the different sub-
types is relatively common.
The most prevalent type of CVST is dural
sinus thrombosis (or sinus thrombosis, SVT),
which refers to thrombotic occlusion of one or
more dural sinuses. The superior sagittal sinus is
most commonly affected, followed by the trans-
verse sinus [2]. Deep cerebral venous thrombosis
(DVT) affects the internal cerebral veins, vein of
Galen, and/or the basal veins of Rosenthal and
their tributaries. The straight sinus is included
with the deep cerebral veins by most authors.
Involvement of the deep cerebral veins is present
in approximately 10 % of all patients with CVST
and is often accompanied by sinus thrombosis.
Isolated thrombosis of the deep cerebral veins is
much less common [4] (Fig. 2).
The term cortical venous thrombosis (CVT)
indicates thrombosis of the superficial cortical
veins including the anastomotic vein of Labbé
(Fig. 1). CVT most commonly affects the frontal
cortical veins, followed by the parietal veins.
Comparable to isolated deep venous thrombosis,
isolated CVT seems to be rather rare and has only
been described in case reports and small case
series. In the wide majority of cases, concomitant
SVT is present, which typically involves the supe-
rior sagittal sinus. It is hypothesized that in those
combined cases, CVT develops secondary to
superior sagittal sinus thrombosis by retrograde
spread of thrombotic material from the sinus into
the cerebral veins draining into it [5].
Evaluation of the cortical veins is challenging
because the cortical veins, unlike the dural sinuses
938
and deep cerebral veins, show considerable intra-
and interindividual variations regarding their
number, diameter, and anatomic course. This
complicates the interpretation of neuroimaging
studies, particularly the analysis of angiographic
datasets. Within the last decade T2*WI gradient-
echo sequences have been found to be very sen-
sitive for the detection of CVT and currently rep-
resent the diagnostic gold standard for the
diagnosis of this specific subtype of CVST (for
details see below) [5–7]. The application of these
sequences in the diagnostic work-up of CVST
resulted in an increased number of reported CVT
cases in recent years. Today it is widely believed
that cortical vein involvement is much more com-
mon in CVST than previously thought and that
isolated cortical vein thrombosis might not be as
rare as previously thought but was often missed
prior to the wide use of T2*WI sequences in
suspected CVST.
Clinical Signs and Symptoms
In contrast to arterial ischemic stroke, CVST usu-
ally presents with a more subacute clinical onset
with variable, rather unspecific, and heteroge-
neous symptoms. More or less severe headache
constitutes the most common initial symptom of
CVST, which is observed in 75–90 % of all
patients. Besides, patients often present with
other signs of increased intracranial pressure
such as dizziness, nausea, and visual disturbances.
Additional symptoms of CVST depend basically
on the location of the thrombosis, i.e., on the
subtype of CVST [1, 8].
Involvement of the deep cerebral veins often
results in an altered level of consciousness, which
is present in over 70 % of DVT cases [8] (Fig. 2),
while cortical venous thrombosis frequently pre-
sents with focal or generalized seizures or focal
neurological deficits as, e.g., hemiparesis, apha-
sia, or hemianopsia. In cases with isolated throm-
bosis of the deep or cortical veins, symptoms of
increased intracranial pressure may be absent
[9]. Other important factors that determine clinical
presentation are patient age, the presence of
J. Linn
concomitant parenchymal changes, and the inter-
val from symptom onset to diagnosis [10, 11].
Laboratory Findings
Although normal D-dimer levels have a high neg-
ative predictive value for suspected CVST, false-
negative results can be observed, particularly in
patients with isolated thrombosis of the deep cere-
bral veins. It has been shown that D-dimer levels
may be normal in up to 25 % of cases, presumably
due to the relatively small volume of thrombotic
material [12].
Prognosis and Outcome
The outcome of CVST has improved significantly
in recent decades, largely due to improvements in
neuroimaging, which enable promptly diagnosis
and treatment with intravenous or low-molecular-
weight heparin. To date, CVST has a very good
prognosis when diagnosed early, with complete
recovery achieved in 80 % of patients [13]. How-
ever, it must be mentioned that the average time
interval from symptom onset to definitive diagno-
sis is still 7 days, mainly due to the diverse and
unspecific clinical presentation of the disease and
its delayed or subacute clinical onset, especially in
the small subset of patients with normal D-dimer
levels [1].
Delayed diagnosis correlates with a poorer
clinical outcome in CVST, which is observed in
approximately 10–15 % of patients. Patients with
deep cerebral venous thrombosis, males, and
patients who only present with signs of increased
intracranial pressure are particularly prone to
delayed diagnosis of CVST [4, 14]. Other risk
factors that are associated with a poor outcome
include advanced age, secondary intracerebral
hemorrhage due to venous congestion, and
involvement of the deep cerebral veins or right
transverse sinus [4, 14]. In addition, outcome is
worse in patients in whom central nervous system
infection or an intracerebral tumor is found as
predisposing cause of CVST [4]. Although
involvement of the deep cerebral veins is one of
39 Imaging of Cerebral Venous and Sinus Thrombosis
the most important risk factors or a poor or even
fatal course of the disease, it has been shown that
even in this specific subtype the outcome has
considerably improved within the recent decades.
A study on 32 patients with isolated or combined
thrombosis of the deep cerebral veins reports a
good clinical outcome with a modified Rankin
Scale (mRS) of 2 or higher indicating little or no
functional disability in 75 % of patients of DVT
patients after prompt treatment with intravenous
or subcutaneous low-molecular-weight heparin.
Twenty-five percent of patients deteriorated with
progressing coma, even though most of them
received venous endovascular recanalization
therapy [12].
Imaging Features of CVST
Direct and Indirect Neuroimaging Signs
Basically, both direct and indirect signs of CVST
can be identified on neuroimaging regardless of
the site of the thrombosis. In this regard, “direct
signs” indicate imaging abnormalities, which are
caused directly by the thrombotic material within
the affected veins and/or sinus. Direct signs
include both “positive” visualization of the throm-
bus on unenhanced CT or MRI and “negative”
visualization of the thrombotic material as a filling
defect in digital-subtraction angiography (DSA),
CT angiography (CTA), or other contrast-
enhanced images or absence of a flow signal in
flow-sensitive venous MR angiography (MRA).
On the contrary, the term “indirect signs” refers
to brain parenchymal changes developing second-
ary to the thrombotic occlusion, e.g., venous
edema or venous infarction, subarachnoid hemor-
rhage, and parenchymal hemorrhage due to
venous congestion. In contrast to arterial ischemic
stroke, edema or infarction caused by CVST does
not conform to arterial territories but crosses ter-
ritory boundaries. Thrombosis of the dural sinus
and/or cortical veins commonly results in venous
edema or infarction in subcortical white matter or
the cerebral cortex. If the deep cerebral veins are
affected, edema or venous infarction typically
involves the thalamus and basal ganglia [3].
939
Particularly in cases with bilateral rather symmet-
ric parenchymal signal abnormalities in the thal-
ami and/or basal ganglia, suspicion for deep
cerebral venous thrombosis should be high. The
presence of secondary intracerebral hemorrhage
at initial diagnosis correlates with more severe
initial symptoms and with a poorer outcome [4].
Digital-Subtraction Angiography (DSA)
While DSA has once been considered the gold
standard for CVST diagnosis, it no longer plays a
significant role in the diagnosis of this disease to
date but has been largely replaced by noninvasive
imaging modalities, i.e., CTA and MRA. How-
ever, due to its higher spatial and temporal reso-
lution, DSA is still superior to MRA and CTA
with regard to dynamic information and can
yield important additional information particu-
larly in collateral venous drainage. This might
justify its diagnostic use in difficult cases.
Some authors hypothesize that DSA might
yield a higher sensitivity for cortical venous
thrombosis compared to noninvasive imaging
modalities, yet systematic data to support this
hypothesis is not available. In patients with iso-
lated cortical vein thrombosis, DSA may show an
“absent” cortical vein and/or a partially opacified
vein with an abrupt cutoff. However, as men-
tioned above, the value of these direct signs of
CVST is considerably limited by the large
interindividual variations regarding both the num-
ber and anatomic location of the cortical veins.
Therefore, diagnosis of cortical venous thrombo-
sis based on DSA mainly relies on indirect signs.
The latter include dilated cortical veins adjacent to
the site of thrombosis (“corkscrew vessels”),
which indicate collateral venous drainage, evi-
dence of a focal delay in venous drainage, as
well as signs of venous congestion in the brain
parenchyma drained by the occluded vein. The
limitations of DSA for the evaluation of cortical
veins likewise apply to noninvasive angiographic
techniques including CTA and MRA.
Besides its role as an additional diagnostic
measure in difficult cases, DSA also is of potential
value as a therapeutic tool in selected patients,
940
Table 1 Stage- and sequence-dependent variability of clot signal characteristics (Modified from Linn and Br€
[3])
Signal Acute stage (0–5
MRI sequence intensity days)a
T1w Hyperintense 30 %
Isointense 68 %
Hypointense 2%
T2w, PDw, Hyperintense 25 %
FLAIR Isointense 10 %
Hypointense 65 %
Numbers printed in bold letters indicate the most common clot signal intensity in the respective stage
T1w T1-weighted spin-echo sequence
T2w T2-weighted spin-echo sequence
PDw proton-density-weighted spin-echo sequence
FLAIR fluid-attenuated inversion recovery-weighted sequence
a
Indicates days after symptom onset
which are considered for interventional treatment
such as local intra-arterial thrombolysis or
mechanical recanalization [13]. Although these
invasive treatment options have been shown to be
effective in single cases and small case series [15,
16], systematic data on their value in CVST are
lacking, and they are currently widely regarded as
last therapeutic option in otherwise devastating cases.
Magnetic Resonance Imaging
Today MRI including MR angiography is com-
monly regarded the diagnostic gold standard in
suspected CVST although CTA is increasingly
reported to be equivalent to MR angiography
(see below) [1, 13].
Direct Signs of CVST on MRI
On MRI, CVST presents with rather complex
MRI sequence-dependent signal abnormalities,
which also vary significantly between the differ-
ent stages of the disease. Depending on the age of
the thrombotic material, an acute, subacute, and
chronic stage of CVST can be distinguished which
all present with different signal characteristics
which are indicated in Table 1. On the contrary,
the location of the thrombus, i.e., the subtype of
CVST, plays a minor role, as the signal character-
istics of the thrombotic material are similar regard-
less of whether the dural sinuses, the deep cerebral
veins, or the cortical veins are involved [3].
J. Linn
uckmann
Subacute stage (6–15 Chronic stage (>15
days)a days)a
71 % 39 %
29 % 54 %
0% 7%
52 % 43 %
32 % 45 %
16 % 12 %
Awareness of the significant time- and sequence-
dependent variations of thrombus signal intensity
is essential for the correct interpretation of MRI
datasets in patients with suspected CVST, which
requires a great deal of experience in order to
avoid diagnostic and technical pitfalls.
Spin-Echo T1-, T2-, Proton-Density-,
and Fluid-Attenuated Inversion Recovery
(FLAIR)-Weighted Sequences
In the acute phase of CVST (days 0–5 from symp-
tom onset), the thrombotic material typically pre-
sents with low signal intensity on T2-WI and a
relative isointense signal on T1-WI spin-echo
sequences (Fig. 3). As these signal characteristics
very much resemble those of regular flow voids in
patent venous structures, thrombus detection early
in the course of the disease is hampered on these
sequences. This fact significantly limits the sensi-
tivity of spin-echo MR sequences, especially dur-
ing the initial days after symptom onset.
During the subacute phase (days 6–15), the
venous clot becomes increasingly hyperintense
in both T2- and T1-WI spin-echo sequences [17]
(Fig. 4).
Although these signal patterns are found in the
majority of cases, thrombotic signal intensity
might vary considerably in the individual case
(Table 1). On FLAIR and proton-density
sequences, signal characteristics of the thrombotic
material are very similar to those found on T2-WI
spin-echo sequences [5] (Figs. 3 and 4).
39 Imaging of Cerebral Venous and Sinus Thrombosis 941

Fig. 3 MRI findings in the acute phase of sinus thrombo-
sis. A 35-year-old female patient with thrombosis of the
right transverse sinus. 3-T MRI was performed 3 days after
symptom onset. The venous clot (arrows) is partially iso-
and partially hyperintense on T1-WI imaging (a), iso- to
slightly hypointense on FLAIR sequence (b), and
profoundly hypointense on T2*WI imaging (c). It shows
no hyperintense signal on diffusion-weighted imaging (d)
and delineates as lack of regular flow signal on maximum
intensity projection (MIP) reconstruction (e) and source
images (f) of venous time-of-flight MR angiography

Fig. 4 MRI findings in the

subacute phase of sinus
thrombosis. A 76-year-old
female patient with
thrombosis of the left
transverse and sigmoid
sinuses. 1.5-T MRI was
performed 2 weeks after
symptom onset. The venous
clot (arrows) is
hyperintense on T1-WI (a),
T2-WI (b), and FLAIR
imaging (d) and delineates
as lack of regular flow signal
on the source image of
venous MR angiography (c)
942 J. Linn

Fig. 5 Hyperintense signal

of venous clot on diffusion-
weighted imaging (DWI). A
27-year-old female patient
with deep cerebral venous
thrombosis. DWI images
(a, b) performed within the
acute phase of the disease
show a hyperintense signal
of the venous clot (arrows
in a and b). On
susceptibility-weighted
imaging (SWI, c), the
thrombotic material is
profoundly hypointense
(normal white arrows in c).
Note also the hypointense
normal flow signal in
regular veins due to the
deoxyhemoglobin fraction
of venous blood (e.g.,
dotted white arrows in c). In
contrast, regular arteries
with oxyhemoglobin are
delineated as hyperintense
signal on this sequence
(e.g., black arrows in c)

One has to be aware of several potential
sources of artifacts and pitfalls on MRI. As an
example, a hyperintense signal from normal
venous flow might mimic a thrombus on T1-WI
spin-echo sequences, if those are acquired without
adequate flow compensation. Its proneness to arti-
facts constitutes a slight disadvantage of MRI
compared to CT with CTA and requires expert
knowledge to avoid misinterpretation in
suspected CVST [3, 17].
Diffusion-Weighted Imaging (DWI)
Venous clot might present with a hyperintense
signal on B1000 images of diffusion-weighted
MR sequences and with a correspondingly
decreased apparent diffusion coefficient (ADC)
on ADC maps, which indicates restricted diffu-
sion within the thrombus itself (Fig. 5).
The sensitivity of this sign ranges from approx-
imately 5 % to 40 % in different studies [5, 18].
It has predominately been observed in the sub-
acute stage of CVST. In this stage, the thrombus is
typically also hyperintense on T2-, T1-, and
FLAIR-weighted images and thus can usually be
easily identified without additional information
provided by DWI. However, there is initial evi-
dence that thrombus signal characteristics on DWI
might serve as a prognostic factor in CVST. In a
pilot study by Favrole et al., the presence of
restricted diffusion in the clot was associated
with low rates of recanalization [19]. The value
of this observation should be investigated in fur-
ther studies.
T2*WI Gradient-Echo Sequences
and Susceptibility-Weighed Imaging
On T2*WI gradient-echo MRI sequences, para-
magnetic substances as, e.g., thrombotic material
induce considerable susceptibility artifacts, which
results in a profound hypointense signal. Recent
39 Imaging of Cerebral Venous and Sinus Thrombosis
Fig. 6 T2*WI imaging in cortical venous thrombosis. A
73-year-old patient with isolated left frontoparietal cortical
venous thrombosis. T2*WI gradient sequences (a–c) show
a profound hypointense signal in the affected cortical veins
(white arrows in a–c). Secondary parenchymal changes,
i.e., indirect signs of cerebral venous thrombosis, are also
present. Axial T2 (d, e) and coronal FLAIR (f) images are
studies have shown that T2*WI sequences are of
considerable additional value in the diagnosis of
CVST, particularly in the acute phase of the dis-
ease, when the thrombus is still isointense on
T1-WI sequences [20–22]. T2*WI gradient-echo
imaging delineates the acute thrombus as a homo-
geneous, markedly hypointense tubular structure,
which allows its detection early after symptom
onset (Figs. 2 and 3).
With regard to the thrombus signal characteris-
tics later in the course of the disease, findings are
somewhat less consistent. Some authors report a
persistence of the hypointense signal of the venous
clot for several months in a considerable percent-
age of cases [21]. In contrast, others observed a
relatively strong time dependence of the
hypointense signal, which they observed in 90 %
of cases during the first week but in only 9–32 % of
cases during the subacute or chronic phases [22].
943
ideal for the depiction of venous edema (arrows in d–f),
while the T2*WI sequence is best suited to detect subse-
quence hemorrhagic changes (black arrows in b and c).
Note that extensive preexisting microangiopathic white
matter lesions were present in this patient (black asterisks
in f)
T2*WI imaging is of particular diagnostic
yield in patients with suspected cortical vein
thrombosis and has been shown to be significantly
superior to all other MRI sequences and CTA with
regard to the evaluation of the cortical veins
(Fig. 6). To date, T2*WI sequences are considered
the gold standard for the diagnosis of isolated or
combined cortical venous thrombosis. Thus, it is
strongly recommended to implement this
sequence in the standard MRI protocol in
suspected CVST [5–7].
However, one has to keep in mind that T2*WI
sequences might also yield false-positive or
(rarely) false-negative results. Chronic convexity
subarachnoid hemorrhage or cortical superficial
siderosis might mimic cortical venous thrombosis
on T2*WI images. Cortical superficial siderosis is
defined as a hypointense linear signal within the
superficial layers of the cerebral cortex, and
944
represents an important MRI biomarker of cere-
bral amyloid angiopathy, but should not be
misinterpreted as a hypointense venous clot [23,
24]. To avoid false-positive diagnoses, a linear
hypointense structure on T2*WI GRE images
should only be regarded as direct signs of venous
clot if it (1) conforms to the anatomic course of a
venous structure, (2) results in a prominent
blooming effect which exceeds the regular diam-
eter of the affected vessel, and (3) shows a round
cross section in slices perpendicular to its course
[23, 24].
The susceptibility effect of paramagnetic sub-
stances on conventional T2*WI sequences can be
further enhanced by either combining both mag-
nitude and phase information of the MRI signal
using adequate post-processing steps or by apply-
ing multiple echo times. These approaches are
referred to as susceptibility-weighted imaging
(SWI) or susceptibility-weighted angiography
(SWAN). Susceptibility-weighted images are
even more sensitive than conventional T2*WI
sequences for detecting subtle abnormalities in
magnetic susceptibility. They allow an excellent
visualization of the normal cerebral venous
anatomy, based on the fact that venous blood is
dark on SWI due to its high deoxyhemoglobin
fraction. Venous clot should appear even more
profoundly hypointense on these sequences. In
single cases SWI has yielded promising results
with regard to diagnosis of CVST [3, 25]
(Fig. 5). Yet, it has to be considered that – as
mentioned before – regularly perfused veins also
appear hypointense on these sequences, which
might hamper their interpretation (Fig. 5). There-
fore, larger studies are needed to determine the
value of susceptibility-weighted imaging in the
diagnosis of CVST.
Venous MR Angiography (MRA)
Different technical approaches are used in imag-
ing the cerebral veins and sinuses on MRI. Under
routine clinical conditions, venous time-of-flight
(TOF-) MRA is most widely used. It is a
noncontrast-enhanced, flow-sensitive technique
in which the thrombus is delineated negatively
as the absence of regular flow in the affected
vascular segment (Fig. 3). Venous TOF-MRA is
J. Linn
considerably prone to artifacts, and one should be
aware of the following potential pitfalls [26]:
Artificial flow gaps on venous TOF-MRA
might result from slow blood flow or from
in-plane flow parallel to the acquisition plane.
Within the non-dominant transverse sinus (typi-
cally the left one), such artificial flow gaps are
observed in over 30 % of patients [26, 27]. In
addition, a hypo- or aplastic transverse sinus,
which is frequently found as a variant, might
easily be misinterpreted as a sinus thrombosis, if
venous MRA is interpreted solely [17, 26]
(Fig. 7).
Finally, if imaging is performed within the
subacute stage of the disease, a T1-hyperintense
venous clot might simulate regular flow on
TOF-MRA [27]. To avoid misinterpretation of
TOF-MRA, it is strongly recommended to care-
fully analyze not only the three-dimensional min-
imum-intensity projection reconstructions but
also the source images.
Besides TOF-MRA, phase-contrast venous
MRA, contrast-enhanced venous MRA, and
contrast-enhanced MPRAGE (magnetization-
prepared rapid acquisition gradient-echo)
sequences are also used to evaluate CVST. The
contrast-enhanced techniques are much less prone
to artifacts than flow-sensitive techniques. New
MRA techniques such as time-resolved MRA
might be of additional value for the diagnosis of
CVST. As an example, the combination of
dynamic and static three-dimensional MRA,
which is known as combined, four-dimensional
venous MRA, has been shown to be superior to
TOF-MRA with regard to the evaluation of the
cerebral venous system [28], but further studies
are required to evaluate its diagnostic yield in
CVST. Regarding the evaluation of the cortical
veins, it should be taken into account that basi-
cally all venous MRA techniques are hampered by
the same limitations as DSA and CT angiography,
namely, by the considerable anatomic variations
of these vessels.
Indirect Signs of CVST in MRI
MRI constitutes the most sensitive modality for
detecting secondary parenchymal changes in
CVST, i.e., indirect signs of venous thrombosis.
39 Imaging of Cerebral Venous and Sinus Thrombosis
Fig. 7 Hypoplastic transverse sinus mimicking sinus
thrombosis on venous MR angiography. Patient with a
hypoplastic proximal left transverse sinus, which might
be misinterpreted as sinus thrombosis on maximum inten-
sity projection (MIP) reconstruction of the venous MR
angiography (a). Distal from the confluence with the left
vein of Labbé, the distal transverse sinus shows a regular
Venous edema and venous infarction are best
delineated on T2-, proton-density-, and/or
FLAIR-weighted sequences, while T2*WI
gradient-echo sequences ideally visualize second-
ary hemorrhagic changes due to venous conges-
tion (Fig. 6).
Diffusion-weighted sequences yield variable
and heterogeneous findings in the brain paren-
chyma in CVST and might indicate cytotoxic or
vasogenic edema or a mixture of both in individ-
ual cases. Clinical studies as well as studies in
animal models provided evidence that cytotoxic
edema precedes vasogenic edema in CVST, which
is reflected by a decreased apparent diffusion
coefficient (ADC value) at early stages of throm-
bosis, followed by a normalization or increase in
ADC value later during the course of the disease
[29, 30] (Fig. 8).
945
lumen. Careful inspection of the source images of the
angiography (b) and of the other MR sequences helps to
exclude the thrombosis. On the source image of the angi-
ography, flow within the hypoplastic part of the sinus can
be depicted (arrows in b). There are no signal abnormali-
ties indicative of venous clot on FLAIR (c) or T2*WI
sequences (d)
Computed Tomography
Unenhanced CT
Due to its wide availability, its cost-effectiveness, and
the short examination time, unenhanced CT still
constitutes the imaging method of first choice in the
emergency setting in most clinical institutions. On
unenhanced CT, an acute thrombus might present as
a homogeneous hyperdense structure. This direct
sign of thrombosis is referred to as cord sign or
dense vein sign [31]. Both terms are used rather
inconsistently in the literature on CVST. It might be
recommended to use the term “cord sign” if throm-
bosis affects the dural sinuses and the term “dense
vein sign” in case of suspected thrombosis of the
cerebral veins. A cord sign or dense vein sign is most
commonly observed in the acute stage of CVST, i.e.,
during the first week after symptom onset (Fig. 9).
946
Fig. 8 Venous edema on T2- and diffusion-weighted
images. A 27-year-old female patient with thrombosis of
the internal cerebral veins (same patient as shown in
Fig. 2). FLAIR image (a) shows symmetrical edema within
the thalami (normal arrows in a) and the caudate nuclei
(dotted arrows in a), bilaterally. On B1000 images of the
During the second week the thrombus first
becomes isodense and later hypodense. Data on
the sensitivity and specificity of the cord sign for
diagnosis of sinus thrombosis varies in recent
studies between 65 % and 73 % for sensitivity
and 97 % and 100 % for specificity [31, 32]. The
sensitivity of the dense vein sign for cortical vein
thrombosis has been found to be very low (25 %)
[5]. In contrast, its diagnostic yield in thrombosis
of the deep cerebral veins seems to be much
higher with a reported sensitivity of 100 % and a
specificity of 99 % compared to MRI as the gold
standard [31] (Fig. 10).
J. Linn
diffusion-weighted sequence (b), there are mixed signal
intensities within the thalami. Parts of the edema show a
hyperintense signal (e.g., arrows in b), with corresponding
low signal on apparent diffusion coefficient (ADC) maps
(c, arrows), indicative of restricted diffusion
On unenhanced CT, several potential pitfalls
may lead to false-positive or false-negative find-
ings on the presence of a cord sign or dense vein
sign. First of all, partial-volume effects, which are
caused by the adjacent skull, may mimic a false-
positive cord sign in the dural sinuses or a dense
vein sign in the cortical veins. Evaluation of the
deep cerebral veins is not hampered considerably
by this effect owing to their deep location. Regular
flow in patent veins might also appear slightly
hyperdense on noncontrast CT and lead to false-
positive results, particularly in patients with ele-
vated hematocrit or high hemoglobin levels, as it
39 Imaging of Cerebral Venous and Sinus Thrombosis
Fig. 9 Cord sign on unenhanced CT. A 34-year-old male
patient with thrombosis of the proximal left transverse
sinus. The venous clot delineated hyperdense on
unenhanced CT (a), a direct sign of thrombosis, which is
referred to as cord sign (arrows in a). The venous clot is
visualized as a filling defect in the affected sinus on CT
has been shown that sinus attenuation correlates
moderately with hematocrit and hemoglobin
levels [33, 34] (Fig. 11).
Quantitative assessment of the sinus density
might be helpful. For patent sinuses, a mean den-
sity of approximately 50 Hounsfield units
(HU) with a standard deviation of 7.5 HU has
been reported [34], while cutoff value of
70 Hounsfield units (HU) seems to be indicative
of a thrombosis [33]. If imaging is not performed
within the acute phase of CVST, i.e., within the
first week, the thrombotic material may appear
isodense again, yielding a false-negative result.
In summary, regarded the limited sensitivity of
unenhanced CT, this method alone does not allow
to reliably detect a CVST but is useful as a screen-
ing examination which should be followed by
CTA or MRI if clinical suspicion is high.
947
angiography (b, arrows) and as a profound hypointense
structure on T2*WI images (c, arrows). Note the sharp
margin of the thrombus (asterisks in b and c). There is
regular flow in the distal part of the sinus (dotted arrows in
b and c)
Contrast-Enhanced Computed
Tomography
To date, contrast-enhanced CT plays only a minor
role in the diagnosis of CVST owing to the wide
availability of CTA. If contrast-enhanced CT is
performed, a so-called “empty delta sign” might
be delineated in CVST patients. This finding indi-
cates a direct sign of sinus thrombosis, which is
typically found in the superior sagittal sinus on
slices perpendicular to the course of the sinus and
refers to a triangular area with an enhancing
periphery and a non-enhancing center [35]. This
imaging appearance of the affected sinus is due to
the fact that the sinus walls show strong contrast
enhancement while the thrombotic material
within the lumen of the sinus does not enhance
significantly after administration of intravenous
contrast agent (Fig. 12).
948
Fig. 10 Dense vein sign on unenhanced CT. A 27-year-
old female patient with thrombosis of the internal cerebral
veins (same patient as shown in Figs. 2 and 8). On
unenhanced CT (a, b) the venous clot delineates as
hyperdense tubular structures within the internal cerebral
veins (normal arrows in a), the vein of Galen (normal
CT Angiography
CT angiography (CTA) performed on modern
multidetector-row CT scanners enables delineation
of the dural sinuses and cerebral veins at relatively
high spatial resolution and allows for three-
dimensional multi-planar reconstructions and/or
maximum intensity projections of the cerebral
venous system. This technology is increasingly
considered equivalent or even superior to MR angi-
ography, particularly with regard to the evaluation
of smaller veins, the inferior sagittal sinus, and the
non-dominant transverse sinus. CVST is visualized
J. Linn
arrow in b), and the straight sinus (dotted arrows in
a and b). This direct sign of venous thrombosis on
unenhanced CT is referred to as “dense vein sign.” On
sagittal reconstructions of the CT angiography (c), the
venous clot is visualized as a filling defect in the affected
veins (c, arrows)
as a filling defect in the affected venous structures
and vessels on CTA (Fig. 13). As a major advan-
tage compared to flow-sensitive TOF-MRA, CTA
is almost not susceptible to flow artifacts. This is
particularly important if anatomic variations are
present in a patient, as, e.g., a hypoplastic left
transverse sinus. CTA is of particular value in the
emergency setting due to its short examination
time, its wide availability, and its ability to simul-
taneously delineate the arterial and the venous
cerebral system in one scan if applied with an
appropriate scan delay [36].
39 Imaging of Cerebral Venous and Sinus Thrombosis
Fig. 11 False-positive cord sign on unenhanced
CT. Patient with an elevated hematocrit resulting in a
relative high density in the dural sinuses on unenhanced
CT (a–c, e.g., arrows). There is a normal flow void on the
T2-WI image (d, arrows), and there are no signal
Fig. 12 Empty delta sign on contrast-enhanced CT. A
45-year-old patient with thrombosis of the superior sagittal
sinus. On contrast-enhanced CT, the thrombosis results in a
so-called empty delta sign (arrow). This direct sign of sinus
thrombosis refers to a triangular area with an enhancing
periphery and a non-enhancing center and is due to the fact
that the sinus walls show strong contrast enhancement
while the thrombotic material within the lumen of the
sinus does not enhance significantly
949
abnormalities indicative of venous clot on T2*WI
sequences (e, arrows). Venous MR angiography (f) con-
firms regular venous flow in the transverse and superior
sagittal sinuses (e.g., arrows in f)
However, the value of CTA in CVST seems to
differ considerably depending on the
localization of the thrombus, i.e., of subtype of
CVST. With regard to sinus thrombosis, it
has been demonstrated that multi-slice CTA is
equally sensitive and specific as MRI [37,
38]. On the contrary, CTA showed a very
low sensitivity for the detection of cortical vein
thrombosis due to similar limitations as
mentioned above for DSA. Therefore, CTA can-
not be considered suitable to diagnose or
exclude isolated or combined cortical vein
thrombosis of the cortical [5]. Data on the delin-
eation of the deep cerebral veins on CTA are
promising; however, systematic studies on the
diagnostic value of this technology for the detec-
tion of deep venous thrombosis are currently
lacking (Fig. 10).
Although CTA is less prone to artifacts com-
pared to MRI, some potential pitfalls have to be
considered, as they might result in misinterpreta-
tion (particularly false-negative results) of CTA.
Especially in the acute phase of the disease (days
0–5 after symptom onset), a very hyperdense
thrombus (corresponding to a cord sign or dense
950
Fig. 13 Sinus thrombosis on CT angiography. A 45-year-
old patient with thrombosis of the superior sagittal sinus
(same patient as shown in Fig. 12). On CT angiography, the
vein sign on unenhanced CT) might mimic nor-
mal contrast filling of the affected venous struc-
ture on CTA and yield false-negative results. To
avoid this source of misinterpretation, CTA scans
should always be interpreted together with
unenhanced CT scans, which directly demonstrate
the cord sign or dense vein sign. This might be
particularly important with regard to the evalua-
tion of the deep cerebral veins on CTA, but further
data on this issue are needed.
In contrast to fresh thrombotic material, an
organized, chronic venous clot can show contrast
enhancement on CTA instead of causing a filling
defect, which might also lead to a false-negative
interpretation. Partial-volume effects from the
enhancing walls of small veins or from the adja-
cent skull might obscure a filling defect and rep-
resent another source of false-negative results.
False-positive interpretation of CTA can be
caused by prominent arachnoid granulations
(pacchionian granulations), which might result in
a filling defect similar to the one caused by throm-
botic vessel occlusion. In most cases, however,
they can rather easily be distinguished from
thrombotic material based on their typical round
or oval shape, smooth margins, and cerebrospinal
fluid isodensity [37].
As the effective radiation dose of a
multidetector-row CTA scan at 120 kV is less
than 1 mSv (and thus even less than the mean
J. Linn
venous clot is visualized as a filling defect in the affected
sinus (a–c, arrows)
effective dose of an unenhanced cranial CT
scan), radiation exposure is only regarded as a
minor disadvantage of CTA compared to
MRI [39].
Summary
To date, MRI including venous MR angiography
and a T2*WI gradient-echo sequence is consid-
ered the gold standard for the diagnosis of CVST.
Its correct interpretation requires a detailed
knowledge of the complex time- and sequence-
dependent signal characteristics of the thrombotic
material. Diagnosis of isolated or combined
involvement of the cortical veins in CVST is
particularly challenging and requires T2*WI
sequences which delineate the venous clot as
hypointense tubular structure.
Unenhanced CT together with multidetector-
row CT angiography provides a very fast, widely
available, and cost-effective alternative technique
which is increasingly considered equivalent to
MRI in suspected sinus thrombosis and deep cere-
bral venous thrombosis, but not in thrombosis of
the cortical veins.
Radiologists should be aware of a variety of
potential pitfalls of both MRI and CT techniques,
which might yield false-positive or false-negative
results regarding the presence of a CVST.
39 Imaging of Cerebral Venous and Sinus Thrombosis
References
1. Stam J (2005) Current concepts: thrombosis of the
cerebral veins and sinuses. N Engl J Med
352:1791–1798
2. Ferro JM, Bacelar-Nicolau H, Rodrigues T, ISCVT and
VENOPORT investigators et al (2009) Risk score to
predict the outcome of patients with cerebral vein and
dural sinus thrombosis. Cerebrovasc Dis 28:39–44
3. Linn J, Br€uckmann H (2010) Cerebral venous and
dural sinus thrombosis: state-of-the-art imaging. Clin
Neuroradiol 20:25–37
4. Ferro JM, Canhao P, Stam J, ISCVT Investigators
et al (2004) Prognosis of cerebral vein and dural sinus
thrombosis: results of the International Study on Cere-
bral Vein and Dural Sinus Thrombosis (ISCVT).
Stroke 35:664–670
5. Linn J, Michl S, Bochmann K et al (2010) Cortical vein
thrombosis: the diagnostic value of different imaging
modalities. Neuroradiology 52:899–911
6. Boukobza M, Crassard I, Bousser MG et al (2009) MR
Imaging features of isolated cortical vein thrombosis:
diagnosis and follow-up. AJNR Am J Neuroradiol
30:344–348
7. Ihn YK, Jung WS, Hwang SS (2013) The value of T2*-
weighted gradient-echo MRI for the diagnosis of cere-
bral venous sinus thrombosis. Clin Imaging 37:446–450
8. Ameri A, Bousser MG (1992) Cerebral venous throm-
bosis. Neurol Clin 10:87–111
9. Jacobs K, Moulin T, Bogousslavsky J et al (1996) The
stroke syndrome of cortical vein thrombosis. Neurol-
ogy 47:376–382
10. Ferro JM, Canhão P, Bousser MG, Stam J,
Barinagarrementeria F, ISCVT Investigators (2005)
Cerebral vein and dural sinus thrombosis in elderly
patients. Stroke 36:1927–1932
11. Ferro JM, Lopes MG, Rosas MJ, Fontes J,
VENOPORT Investigators (2005) Delay in hospital
admission of patients with cerebral vein and dural
sinus thrombosis. Cerebrovasc Dis 19:152–156
12. Pfefferkorn T, Crassard I, Linn J et al (2009) Clinical
features, course and outcome in deep cerebral venous
system thrombosis: an analysis of 32 cases. J Neurol
256:1839–1845
13. Diener HC, Putzki N (eds) (2008) Diagnostic and ther-
apeutic guidelines in neurology, 4th edn. Thieme,
Stuttgart
14. Girot M, Ferro JM, Canhão P, ISCVT Investigators
et al (2007) Predictors of outcome in patients with
cerebral venous thrombosis and intracerebral hemor-
rhage. Stroke 38:337–342
15. Guo XB, Song LJ, Guan S (2014) Endovascular treat-
ment of chronic, recurrent headache secondary to
chronic cerebral venous sinus thrombosis. J Stroke
Cerebrovasc Dis 23:560–563
16. Guo XB, Fu Z, Song LJ, Guan S (2013) Local throm-
bolysis for patients of severe cerebral venous sinus
thrombosis during puerperium. Eur J Radiol
82:165–168
951
17. Teasdale E (2000) Cerebral venous thrombosis: mak-
ing the most of imaging. J R Soc Med 93:234–237
18. Lövblad KO, Bassetti C, Schneider J et al (2001)
Diffusion-weighted MRI in cerebral venous thrombo-
sis. Cerebrovasc Dis 11:169–176
19. Favrole P, Guichard JP, Crassard I et al (2004)
Diffusion-weighted imaging of intravascular clots in
cerebral venous thrombosis. Stroke 35:99–103
20. Selim M, Fink J, Linfante I et al (2002) Diagnosis of
cerebral venous thrombosis with echo-planar T2*-
weighted magnetic resonance imaging. Arch Neurol
59:1021–1026
21. Idbaih A, Boukobza M, Crassard I et al (2006) MRI of
clot in cerebral venous thrombosis. High diagnostic
value of susceptibility-weighted images. Stroke
37:991–995
22. Leach JL, Strub WM, Gaskill-Shipley MF (2007)
Cerebral venous thrombus signal intensity and suscep-
tibility effects on gradient recalled-echo MR imaging.
AJNR Am J Neuroradiol 28:940–945
23. Linn J, Herms J, Dichgans M et al (2008) Subarachnoid
hemosiderosis and superficial cortical hemosiderosis in
cerebral amyloid angiopathy. AJNR Am J Neuroradiol
29:184–186
24. Linn J, Halpin A, Demaerel P et al (2010) Prevalence
of superficial siderosis in patients with cerebral amy-
loid angiopathy. Neurology 74:1346–1350
25. Kawabori M, Kuroda S, Kudo K et al (2009)
Susceptibility-weighted magnetic resonance imaging
detects impaired cerebral hemodynamics in the supe-
rior sagittal sinus thrombosis – case report. Neurol Med
Chir (Tokyo) 49:248–251
26. Ayanzen RH, Bird CR, Keller PJ et al (2000) Cerebral
MR venography: normal anatomy and potential diag-
nostic pitfalls. AJNR Am J Neuroradiol 21:74–78
27. Lafitte F, Boukobza M, Guichard JP et al (1997) MRI
and MRA for diagnosis and follow-up of cerebral
venous thrombosis (CVT). Clin Radiol 52:672–679
28. Meckel S, Reisinger C, Bremerich J et al (2010) Cere-
bral venous thrombosis: diagnostic accuracy of com-
bined, dynamic and static, contrast-enhanced 4D MR
venography. AJNR Am J Neuroradiol 31:527–535
29. Ducreux D, Oppenheim C, Vandamme X et al (2001)
Diffusion-weighted imaging patterns of brain damage
associated with cerebral venous thrombosis. AJNR Am
J Neuroradiol 22:261–268
30. Srivastava AK, Kalita J, Haris M et al (2009) Radio-
logical and histological changes following cerebral
venous sinus thrombosis in a rat model. Neurosci Res
65:343–346
31. Linn J, Pfefferkorn T, Ivanicova K et al (2009)
Noncontrast CT in deep cerebral venous thrombosis
and sinus thrombosis: comparison of its diagnostic
value for both entities. AJNR Am J Neuroradiol
30:728–735
32. Roland T, Jacobs J, Rappaport A et al (2010)
Unenhanced brain CT is useful to decide on further
imaging in suspected venous sinus thrombosis. Clin
Radiol 65:34–39
952
33. Black DF, Rad AE, Gray LA et al (2011) Cerebral
venous sinus density on noncontrast CT correlates
with hematocrit. AJNR Am J Neuroradiol
32:1354–1357
34. Fanous R, Leung A, Karlik S (2010) Quantitative
assessment of the superior sagittal sinus on unenhanced
computed tomography. Eur J Radiol 75:336–342
35. Virapongse C, Cazenave C, Quisling R et al (1987) The
empty delta sign: frequency and significance in
76 cases of dural sinus thrombosis. Radiology
162:779–785
36. Klingebiel R, Zimmer C, Rogalla P et al (2001) Assess-
ment of the arteriovenous cerebrovascular system by
J. Linn
multislice CT. A single-bolus, monophasic protocol.
Acta Radiol 42:560–562
37. Linn J, Ertl-Wagner B, Seelos KC et al (2007) Diag-
nostic value of multidetector-row CT angiography in
the evaluation of thrombosis of the cerebral venous
sinuses. AJNR Am J Neuroradiol 28:946–952
38. Gaikwad AB, Mudalgi BA, Patankar KB et al (2008)
Diagnostic role of 64-slice multidetector row CT scan
and CT venogram in cases of cerebral venous throm-
bosis. Emerg Radiol 15:325–333
39. Ertl-Wagner B, Hoffmann RT, Bruening R et al (2004)
Multi-slice CT angiography of the brain at various
kilovoltage settings. Radiology 231:528–535
 The Role of Diagnostic Imaging
Techniques for Detection 40
of Extracranial Venous System
Abnormalities Associated with Central
Nervous System Disorders

Kresimir Dolic and Robert Zivadinov

Contents Invasive Imaging Modalities . . . . . . . . . . . . . . . . . . . . . . . . 971

Multimodal Imaging Approach . . . . . . . . . . . . . . . . . . . . . . 974
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 954
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 974
Anatomy and Function of the Cerebral
Venous System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 956 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 975
Extracranial Cerebral Venous Drainage Pathway:
Neck Veins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 957
IJV Valves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 957
Other Neck Veins Serving as Collaterals
for Cerebral Venous Drainage . . . . . . . . . . . . . . . . . . . . . . . 957
Extracranial Cerebral Venous Drainage Pathway:
Abdominal and Thoracic Veins . . . . . . . . . . . . . . . . . . . . . . 958
Extracranial Venous Abnormalities Versus
Developmental Variants . . . . . . . . . . . . . . . . . . . . . . . . . . . . 960
Diagnostic Imaging and the Diagnosis
of Extracranial Venous Abnormalities . . . . . . . . . . . . 960
Noninvasive Imaging Modalities . . . . . . . . . . . . . . . . . . . . 960

K. Dolic (*)
Buffalo Neuroimaging Analysis Center, Department of
Neurology, University at Buffalo, State University of
New York, Buffalo, NY, USA
Clinical Department of Diagnostic and Interventional
Radiology, Clinical Hospital Center Split, University of
Split, Split, Croatia
e-mail: kdolic@bnac.net
R. Zivadinov
Buffalo Neuroimaging Analysis Center, Department of
Neurology, University at Buffalo, State University of
New York, Buffalo, NY, USA
MRI Clinical Translational Research Center, University at
Buffalo, State University of New York, Buffalo, NY, USA
e-mail: rzivadinov@bnac.net; rz4@buffalo.edu

# Springer Science+Business Media New York 2016 953

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_12
954 K. Dolic and R. Zivadinov

insufficiency • Cerebral venous system •

Abstract
Neck veins • Jugular vein reflux • Doppler
The extracranial venous system is complex and
sonography • Magnetic resonance venography
variable. In the last decade, it has been repeat-
• Computed tomography venography • Cathe-
edly shown that the presence and severity of
ter venography • Intravascular ultrasound •
jugular vein reflux is associated with a number
Plethysmography • Multimodal imaging
of central nervous system (CNS) disorders
such as transient global amnesia, transient
monocular blindness, cough headache, pri-
Introduction
mary exertional headache, and more recently
Alzheimer’s disease and aging. A newly pro-
The main function of the venous system is to
posed vascular condition, named chronic cere-
return blood to the heart from the periphery and
brospinal venous insufficiency (CCSVI), has
maintain cardiac output. It represents complex,
triggered recently intense interest in better
low-pressure, freely communicating network of
understanding the role of extracranial venous
vessels, which contains 75 % of the body’s circu-
abnormalities and their developmental vari-
lating blood volume [1, 2]. The extracranial
ants. Their relationship to intracranial CNS
venous system is a complex three-dimensional
pathology, especially in patients with multiple
(3D) structure, not widely examined and only
sclerosis (MS), is poorly understood at this
partially understood, that is often asymmetric
time. So far there is no established invasive or
and represents significantly more variability than
noninvasive diagnostic imaging modality that
intra- and extracranial arterial anatomy [2, 3]. The
can serve as a “gold standard” for the detection
role of arterial supply abnormalities in relation to
of these venous abnormalities/developmental
the pathology of central nervous system (CNS)
variants. The use of noninvasive diagnostic
disorders is well defined in comparison with
imaging techniques such as Doppler sonogra-
extracranial venous abnormalities which in the
phy (DS) remains controversial; however, con-
past were mostly acknowledged as
sensus guidelines and standardized protocols
non-pathological findings due to the lack of inves-
are emerging. The use of magnetic resonance
tigations aimed to define their nature and clinical
venography (MRV) and phase-contrast imag-
significance [4, 5]. More recently, it has been
ing is gaining an increasing interest as an alter-
hypothesized that these venous abnormalities are
native noninvasive diagnostic approach.
likely to be truncal venous malformations [6]
Further, catheter venography (CV) and intra-
characterized by intraluminal defects (such as
vascular ultrasound (IVUS) are becoming
flaps, webs, septums, membranes, and malformed
important diagnostic tools for confirming the
valves) or by extraluminal anomalies represented
presence and severity of extracranial venous
by stenoses of the venous wall ([2, 5, 7]; Table 1).
pathology. Most likely, a multimodal imaging
Recently several CNS disorders such as tran-
approach will ultimately be the most compre-
sient global amnesia, transient monocular blind-
hensive means for screening, diagnostic, as
ness, leukoaraiosis, dementia, normal-pressure
well as monitoring purposes. Further research
hydrocephalus, cough headache, and primary
is needed to determine the spectrum and prev-
exertional headache have been linked to the pres-
alence of these extracranial venous abnormal-
ence and severity of uni- or bilateral jugular
ities/developmental variants and to compare
venous reflux [2, 8–12]. Further, more recently
the imaging findings with pathological
the introduction of a composite criteria-based vas-
examinations.
cular condition named chronic cerebrospinal
venous insufficiency (CCSVI) generated an
Keywords intense interest in better understanding the role
Multiple sclerosis • Extracranial venous abnor- of the extracranial venous system in the patho-
malities • Chronic cerebrospinal venous physiology of CNS disorders, especially in
40 The Role of Diagnostic Imaging Techniques for Detection of Extracranial Venous System. . . 955

Table 1 Proposed classification of internal jugular vein diagnosis is based mainly on color Doppler
(IJV) abnormalities on Doppler sonography (Adapted from sonography (DS) findings in the extracranial
Zivadinov and Chung [5])
(neck) and intracranial veins by assessing five
Doppler sonography venous hemodynamic (VH) criteria (with cutoff
classification Types/examples
of 2 positive criteria used for diagnosis of
Intraluminal Web: multiple septae and/or
structural venous flaps located in a cluster CCSVI) [17, 18]. From the time it was first men-
abnormality: This is Flap: thin linear echogenic tioned, the CCSVI theory has not ceased to pro-
an echogenic structure structure extending from voke controversy and attention in the scientific
extending from the endothelial lining of vein wall community and the media, mostly because the
endothelial lining of Septum: thin linear echogenic
the vein wall with/ sensitivity and specificity results originally
structure extending from
without the presence of reported [14, 17] were not reproduced by other
endothelial lining of vein wall
functional and attached to it at both ends. groups [19–21]. Also, the reliability of using DS
abnormality. These Septum may extend across a in the diagnosis of CCSVI is questionable without
abnormalities include vein to attach on opposing
web, flap, septum, proper training and has been the focal point of
sides or attach on same side
membrane, and recent statements from various societies [19, 22,
Membrane: membranous
malformed valve
structure almost occluding the
23]. Further, the CCSVI concept did not take into
entire diameter of the vein account the physiological variations of the IJV in
Malformed valve: diameter [24].
dysdynamic or fibrous valve One of the main criticisms of the CCSVI con-
Extraluminal Stenosis: cross-sectional area cept arose from the use of endovascular proce-
structural venous (CSA) measurement of 3
dures to unblock potentially stenotic IJV and
abnormality: This is a mm2
restriction of the azygos veins in an open-label fashion without
Annulus: circumferential
venous wall or thickened vein wall that is previously establishing: (a) diagnostic imaging
stenosis. These restricting the vein from fully modalities and protocols that will serve as a
abnormalities include expanding with respiratory or “gold standard” for the detection of extracranial
stenosis and annulus positional changes
venous anomalies and (b) safety as well as effi-
Functional venous Reflux/bidirectional flow:
abnormality: This is present in the IJV for more
cacy of the endovascular procedures in properly
an abnormal cerebral than 0.88 s with the head at designed randomized, double-blinded, sham-con-
venous outflow in the 90 and 0 trolled studies, as well as the potential usefulness
presence of a structural Paradox: vein wall not of endovascular treatment of vein pathology
venous anomaly. reacting to respiratory phase;
These abnormalities
related to CCSVI [5, 25–27]. So far there have
noncompliant
include reflux/ been several case reports concerning patients who
No flow: no color flow noted
bidirectional flow, in vein, despite deep breaths had serious side effects after angioplasty for
paradox, and no flow CCSVI like IJV stent thrombosis requiring open
thrombectomy, stent migration, aneurismal vein
dilatation, cranial nerves neuropathy, as well as
multiple sclerosis (MS) [13, 14]. This returned some reports of lethal cases [2, 28, 29].
possible vascular etiology of MS lesions to the Furthermore, classification, existence, and
forefront of MS research. interpretation of venous abnormalities are ques-
CCSVI is characterized by impaired brain and tionable, given the fact that the same can be found
spinal cord venous drainage due to outflow among healthy populations and patients with
obstruction in the extracranial venous system, other CNS disorders [11, 12, 30, 31, 32, 33]. At
mostly related to anomalies in the internal jugular this time, it remains unclear whether extracranial
(IJV) and azygos veins [14–16]. It was hypothe- venous abnormalities represent an acquired
sized that these venous anomalies may cause alter- pathology or developmental variants.
ations to blood flow that eventually result in iron It is pretty clear that there is an urgent need to
deposition, degeneration of neurons, and charac- define and validate the spectrum of extracranial
teristic brain injury patterns. The CCSVI venous abnormalities and their relationship to
956
neurodegenerative diseases. Since there is still no
established noninvasive or invasive diagnostic
imaging modality that can serve as a “gold stan-
dard/benchmark” for the detection of extracranial
venous anomalies, it is most likely that only a
multimodal imaging approach will represent the
most comprehensive means for the screening,
diagnosis, as well as monitoring for these venous
abnormalities ([2, 9, 23, 34]).
Anatomy and Function of the Cerebral
Venous System
It is important to first appreciate the anatomy and
function of the cerebral venous drainage system in
order to understand the potential role of the extra-
cranial venous system in diseases of the CNS and
aging. The venous system contains approximately
70 % of the circulating blood volume, with
approximately three-quarters of it within small
veins and venules [35, 36]. Cerebral venous sys-
tem drainage is composed of two systems: the
superficial and the deep venous system (Fig. 1).
The superficial system drains blood from the cor-
tex and superficial white matter (WM) by cortical
veins, collected by dural sinuses. There are two
important dural sinuses: the superior sagittal sinus
(SSS) draining dorsolaterally and the cavernous
sinus draining antero-ventrally. The transverse
sinus then drains the SSS equally on both sides
in only 20 % of cases and asymmetrically in more
Fig. 1 Intracranial venous
system anatomy of dural
sinuses, cortical veins, deep
intracerebral veins, and
cavernous sinus (The figure
was reproduced with
permission from the
Radiology Assistant
website: (http://www.
radiologyassistant.nl/en/
p4befacb3e4691/cerebral-
venousthrombosis.html))
K. Dolic and R. Zivadinov
than 50 % of cases, depending on the configura-
tion of torcular Herophili. In 20 % of cases, one
transverse sinus drains the SSS in total (most often
on the right side), and the other one drains the
straight sinus, which collects blood from the deep
venous system [5, 37–39].
The cavernous sinus extends from the superior
orbital fissure to the petrous apex, which receives
orbital venous and middle cranial fossa drainage.
From the cavernous sinus, blood drains
posterolaterally along the superior petrosal sinus
into the transverse sinus and inferior-laterally
along the inferior petrosal sinus into the sigmoid
sinus [5, 40]. The deep cerebral venous system
drains the deep WM and the regions surrounding
the lateral and third ventricles or the basal cistern.
Three veins unite just behind the interventricular
foramen of the Monro to form the internal cerebral
vein (s). These include the choroid vein, septal
vein, and thalamostriate vein. The vein of Galen is
a short (1–2 cm long), thick vein that passes
posterosuperiorly behind the splenium of corpus
callosum in the quadrigeminal cistern. The vein of
Galen receives the internal cerebral vein, the basal
veins of Rosenthal, and the posterior fossa veins
and then drains to the anterior end of the straight
sinus where this unites with the inferior sagittal
sinus. The main collecting vein for the deep
venous system is the straight sinus, which
receives the venous blood from the vein of
Galen and flows into the transverse sinus (most
often into the left side). The basal vein of
40 The Role of Diagnostic Imaging Techniques for Detection of Extracranial Venous System. . .
Rosenthal is an important collateral pathway for
the internal cerebral veins and the vein of Galen.
Venous blood flow can bypass the straight sinus
by connecting with the superficial Sylvian vein
via the deep Sylvian vein [38, 39, 41]. Venous
drainage of the posterior fossa mainly depends on
the galenic system and the petrosal system and, to
a lesser extent, the tentorial veins and the trans-
verse sinuses.
Extracranial Cerebral Venous Drainage
Pathway: Neck Veins
The cerebral venous drainage pathway continues
via neck veins, mainly the IJV, vertebral venous
system, and deep cervical veins (veins in cervical
soft tissue) (Fig. 2) [5, 35, 38].
The IJVs are the largest veins in the neck and
are generally considered to be the most important
cerebral venous outflow pathways. Venous drain-
age of the superficial and deep cerebral venous
system is via the transverse sinuses to the sigmoid
sinuses, which then drain into the IJV [5, 32,
36]. The inferior petrosal sinus, a major drainage
route collecting blood from the cavernous sinus,
communicates with the basilar plexus, anterior
and lateral condylar veins, anterior condylar con-
fluence, and vertebral venous plexus before
draining into the IJVs. The IJVs then join with
the subclavian veins to form the brachiocephalic
vein (BV). The confluence of the bilateral BV is
the superior vena cava, which ultimately drains
venous blood into the heart. Several tributaries in
the neck also drain into the IJVs. These bilateral
IJV branches will interconnect with each other at
the midline to form anastomosing plexi that can
serve as collateral channels to maintain adequate
venous drainage when the principal pathways are
obstructed [5].
The vertebral venous system consists of two
components; one is the vertebral venous plexus
and the other is the vertebral vein (VV). The
vertebral venous plexus can be subdivided as
internal (posterior and anterior internal vertebral
plexus) and external (posterior and anterior exter-
nal vertebral plexus). The IJVs can also exhibit
anastomosis with the other extracranial venous
957
drainage system within the craniocervical junc-
tion region, which includes the anterior condylar
confluent (ACC) and its tributes. Numerous anas-
tomoses of the ACC make it a crossroad between
the cavernous sinus, dural sinuses of the posterior
fossa, IJVs, and posterior cervical outflow tract
(vertebral venous system and deep cervical veins)
[5, 36, 42, 43].
IJV Valves
Although there are anatomical variations, the
valves are generally located about 0.5 cm above
the union of the subclavian vein and IJVs at the
lower limit of the jugular bulb, which are shown in
96.8 % of the general population. They make IJVa
buffer zone between large central veins and the
cerebral venous system. The IJV valves are gen-
erally thought to prevent the backflow of venous
blood and backward venous pressure into the
cerebral venous system during conditions where
the central venous pressure or intrathoracic pres-
sure is increased, such as chest compression
during external cardiopulmonary resuscitation
and severe or repetitive cough and straining
[5, 44–46]. The pressure gradient across compe-
tent IJV valves can be as high as 100 mmHg.
Without competent IJV valves, a sustained or
prolonged retrograde-transmitted venous pressure
via IJVs might impair cerebral venous drainage
and lead to neurological deficits. For example, IJV
valve incompetence has been associated with
encephalopathy after cardiopulmonary resuscita-
tion [5, 47].
Other Neck Veins Serving as Collaterals
for Cerebral Venous Drainage
When there is narrowing of the principal path-
ways of the extracranial venous system, collateral
veins probably represent physiological variations
of the venous system that may play a compensa-
tory role [2, 13]. The extra-jugular venous drain-
age system for cerebral venous drainage mainly
consists of the vertebral venous system and deep
cervical veins [5, 36, 40, 48, 49]. The external
958
Fig. 2 Illustration depicting the predominant veins and
sinuses involved in the craniocervical venous outflow.
Venous narrowing is depicted at locations of interest in
chronic cerebrospinal venous insufficiency. (The figure
jugular vein (EJV) and anterior jugular vein
(AJV), compared with the IJV, are located super-
ficially in the neck. They serve as collaterals and
become prominent (enlarged lumen) when the
main cerebral venous drainage pathways (IJV
and VV) are compromised. EJV is formed by the
confluence of the posterior branch of the posterior
facial vein and the posterior auricular vein [5, 50].
It usually terminates into the confluence of the
subclavian vein and IJV. The AJV receives blood
from superficial veins, such as EJVs, facial veins,
or IJVs. They usually end in the subclavian vein
or EJV. Bilateral AJVs may communicate via the
jugular venous arch (JVA), which is located just
above the sternum. The JVA receives tributaries
from the thyroid gland via inferior thyroid veins.
In summary, venous collaterals in the neck include
K. Dolic and R. Zivadinov
was reproduced with permission from Lazzaro MA, Zaidat
OO, Mueller-Kronast N, Taqi MA, Woo D Endovascular
therapy for chronic cerebrospinal venous insufficiency in
multiple sclerosis. Front Neurol 2011, 2:44)
the anterior (jugular venous system) and the
posterior (vertebral and other deep neck venous
system), and different patterns of collateral estab-
lishment may reflect the location and severity of
venous outflow obstruction [5, 51, 52].
Extracranial Cerebral Venous Drainage
Pathway: Abdominal and Thoracic
Veins
There is communication of vertebral venous
plexus system with the deep thoracic and lumbar
veins, intercostal veins, as well as the hemiazygos
and azygos veins. Abnormalities in these abdom-
inal and thoracic veins may impair venous drain-
age from the vertebral venous system, which
40 The Role of Diagnostic Imaging Techniques for Detection of Extracranial Venous System. . . 959

Anastomosis ant.
Ant.
cardinal veins
cardinal
vein

Common Supracardinal
cardinal vein vein

Post.
cardinal vein Hepatic
segment
int. vena cava

Subcardinal
vein

Int. Jugular V.
(Distal segment of Ext. Jugular V.
Precardinal Vein)
Subclavian V.
Innominate (right) V. Innominate (left) V.
(Brachiocephalic Vein)

Sup. Vena Cava

(Proximal segment of
Coronary sinus
Precardinal/Anterior
Cardinal Vein)

Azygos V. Accessory Hemiazygos V.

(The most proximal segment of
Right Postcardinal Vein with
distal extension by Right
Supercardinal Vein)

Hemiazygos V.

Fig. 3 Paired anterior cardinal veins form common cardi-
nal veins with paired posterior cardinal veins, draining
centrally into the sinus venosus (sinus horns) as depicted
(top). Paired anterior cardinals soon form an anastomosis
between them; the connection grows from the left to the
right anterior cardinal vein to form the left brachiocephalic
(innominate) vein (bottom). The left anterior cardinal vein
distal (cranial) to the anastomosis becomes the “left inter-
nal jugular vein,” while the left anterior cardinal vein
proximal to the brachiocephalic anastomosis regresses/
atrophies to become the base of the “coronary sinus” of
the heart as displayed. The right anterior cardinal
(precardinal) vein proximal to the right brachiocephalic
vein forms the superior vena cava (SVC) with the common
cardinal and terminal/proximal segment of the posterior
cardinal (postcardinal) vein (The figure was reproduced
with permission from Lee BB: Venous embryology: the
key to understanding anomalous venous conditions.
Phlebolymphology 2012, 4:170–181)

serves as an important collateral for cerebral
venous drainage. The hemiazygos arch is
connected with the left renal vein that represents
a major outflow route for shunting blood into the
inferior vena cava. Ultimately, the azygos vein
serves as the final venous blood collector and
drains into the superior vena cava. The anatomy
and developmental stages of the abdominal/tho-
racic blood vessels can be quite variable (Fig. 3).
For example, in some rare variations, the azygos
vein also drains thoracic veins, bronchial veins,
and even gonadal veins. The vein is so named
because it has no symmetrically equivalent vein
on the left side of the body [5, 6].
960
Extracranial Venous Abnormalities
Versus Developmental Variants
The venous system development through stages
may be associated with a number of developmen-
tal variants that do not necessarily represent path-
ological findings. This is the reason why there is
still disagreement about the physiological range
of hemodynamic measurements in these veins
including determination of normal or abnormal
function. It has been hypothesized that these
abnormalities may be related to embryologic
developmental arrest, aging, or other
comorbidities [2, 6, 32]. Probably, even more
important is to establish what constitutes a signif-
icant narrowing of the extracranial venous system
with hemodynamic consequences for the intracra-
nial venous drainage.
So far, there have been reported a huge range of
possible abnormalities in the extracranial veins
draining cerebral venous blood flow. These can be
classified as structural/morphological, hemody-
namic/functional, and those determined only by
the composite criteria and use of multimodal imag-
ing (Table 1) [2, 9]. For example, structural/mor-
phological venous abnormalities can be divided
into those creating narrowing or occlusion and
those causing abnormal distensibility. On the
other hand, hemodynamic functional venous
abnormalities represent an abnormal cerebral
venous outflow in the presence or absence of a
structural venous anomaly in the extracranial
veins. Finally, because it is almost impossible to
determine the prevalence of a single structural/mor-
phologic or hemodynamic/functional venous
abnormality, regardless of the imaging modality
or methodology utilized, the need for use of com-
posite criteria by uni- or multimodal imaging
modalities of the extracranial venous system is
emerging [23].
Diagnostic Imaging and the Diagnosis
of Extracranial Venous Abnormalities
Unlike the carotid artery, the vascular wall of the
IJVs is much more flexible with a variable lumen
diameter which can be influenced by postural
K. Dolic and R. Zivadinov
change, respiration, cardiac function,
hypovolemia, and hydration status even by the
pulsation of nearby arteries [2, 50, 53, 54]. The
current definition (narrowing of >50 %) with
respect to the proximal adjacent vein segment is
mainly derived from observations in the arterial
system [5, 55, 56]. Even less is known about the
main drainage routes of the spine, namely, the
azygos venous system and its pathophysiology.
When performing imaging of the extracranial
venous system, it is almost impossible to take all
of the above factors into account, regardless of the
imaging modality utilized. While there is common
knowledge among radiologists about how to per-
form diagnostic imaging and intervention on the
extracranial arteries, no established standardized
guidelines for detection of extracranial venous
abnormalities currently exist. This chapter sum-
marizes current knowledge regarding the advan-
tages and disadvantages of both noninvasive and
invasive imaging modalities for the detection of
those extracranial venous pathologies (Tables 2
and 3).
Noninvasive Imaging Modalities
Doppler Sonography
Description of Imaging Techniques
DS is clinically the most useful technique for
detecting, localizing, and evaluating peripheral
venous obstruction and venous valvular incompe-
tence [57, 58]. The sensitivity and specificity of
venous DS for symptomatic proximal deep vein
thrombosis exceeds 90 % [59, 60]. Spectral anal-
ysis of the DS signal is used to confirm the pres-
ence or absence of flow and indicates its direction
and the patterns. Spectral analysis of DS signal
and color DS are used to confirm the presence of
reflux. Several DS studies have shown a relation-
ship between IJV drainage abnormalities, charac-
terized by JVR and specific neurological diseases
of undetermined etiology along with a higher
prevalence of brain WM hyperintensities in elderly
people ([8, 12, 32, 61]). An increased prevalence of
JVR, dilated vessel lumen, and slowed flow veloc-
ity in the left IJV, as well as decreased
40 The Role of Diagnostic Imaging Techniques for Detection of Extracranial Venous System. . . 961

Table 2 Advantages and disadvantages of noninvasive diagnostic methods for diagnosis of extracranial venous
anomalies (Adapted from Dolic et al. [2])
Noninvasive
diagnostic
methods Advantages Disadvantages
Doppler Noninvasive No standardized guidelines
sonography Without ionizating radiation Operator dependent
Less expensive Time-consuming (60–120 min)
High resolution Blinding procedures are challenging
Real-time information Cannot perform global view of the veins (limited
Sensitive to detect flow changes, intra- and window)
extraluminal abnormalities Misidentification of the veins
Ability to measure velocity Influenced by hydration status
Possible control of respiratory phases
Magnetic Noninvasive No real-time informations
resonance Without ionizating radiation Cannot detect intraluminal abnormalities
venography Well-established method Low specificity of conventional MRV techniques
Operator independent Influenced by hydration status
Less time-consuming than DS Azygos vein examination needs technical
Provide global view of intra- and improvements due to important artifacts
extracranial venous system (breathing, heart movements, etc.)
Easy to blind Underestimate the vascular caliber
Ability to measure flow and velocity with “Snapshot” nature
advanced technique (phase-contrast MRV)
Global view of collateral veins
Can be performed without contrast
(pregnancy, allergy, etc.)
Plethysmography Noninvasive Time-consuming method
Provides valuable information regarding the Higher false-positive rate due to venous
impact of reflux and obstruction on overall compression arising from incorrect patient
venous function positioning or the action of extrinsic masses
Can monitor the dynamics of venous disease Low resolution
over time and evaluation of treatment
outcomes

Table 3 Advantages and disadvantages of invasive diagnostic methods for diagnosis of extracranial venous abnormal-
ities (Adapted from Dolic et al. [2])
Invasive diagnostic
methods Advantages Disadvantages
Catheter Considered the gold standard Invasive method
venography Real-time information can be obtained by Ionizating radiation
using contrast Cannot be performed without contrast (allergy,
Ability to measure pressure toxicity, etc.)
Provide “road map” for planning Operator dependent
endovascular procedures Time-consuming (>45 min)
Can be complemented by use of more Cannot detect intraluminal abnormalities
sophisticated criteria (time to empty contrast No global view of veins and collaterals
from vein or wasting of the balloon) No standardized definition of significant
Vein stenoses
Intravascular Offers 360 view of the vessel’s wall from Invasive method
ultrasound the inside Lack of experience – no standardize protocols
Can detect intraluminal abnormalities Ring-down artifacts
Easily access all parts of IJVs in comparison Geometric distortion – from imaging in an
with DS oblique plane
Provides more accurate assessment of vein Size of IVUS probe – limitation in the imaging
stenosis and wall thickness than CV and DS of severe stenosis
962
Fig. 4 Examples of chronic cerebrospinal venous insuffi-
ciency venous hemodynamic criteria on Doppler sonogra-
phy. (a) Flap anomalies noted in internal jugular vein (IJV)
lumen; (b) annulus in the left IJV: circumferential thick-
ened vein wall that is restricting the vein from fully
expanding with respiratory or positional changes; (c)
thrombus noted in IJV; (d) severe stenosis of left IJV:
time-averaged mean velocity of bilateral IJV was
found in those over 70 years of age ([12, 32]).
Because of the advantages of the DS in
detecting intraluminal venous pathology, it was
initially promoted as a method of choice for
screening of extracranial venous abnormalities
and developmental variants, indicative of CCSVI
[17, 62]. The diagnosis of CCSVI is both based on
hemodynamic and imaging findings that utilize
DS to study the deep cerebral veins, the IJVs,
and the VVs in both erect and supine positions.
Recent findings suggest that the majority of
CCSVI pathology is confined to the intraluminal
portion of extracranial veins which can be
visualised with high resolution B-mode imaging
while visible “stenoses” (Fig. 4) or extraluminal
venous abnormalities most likely develop more
frequently with the progression of the disease or
age [7, 56, 63].
The prevalence of CCSVI and JVR, as well as
their relationship to clinical findings in CNS
K. Dolic and R. Zivadinov
CSA measurement of 3 mm2; (e, f) reflux/bidirectional
flow directed toward the brain for a duration of >0.88 s in
the right IJV in the supine position (E demonstrates reflux
using color flow, while F demonstrates reflux using spec-
tral analysis – waveform noted above baseline for greater
than 0.88 s) (The figure was reproduced with permission
from Dolic et al. [2])
disorders, emphasizes the need for more quantita-
tive and reproducible measures for the integration
of morphological and functional anomalies. These
include blood flow as well as velocity and blood
volume measurements that could be potentially
more reliable in assessing the degree of venous
outflow obstruction in the IJVs and azygos vein
(Fig. 5). Monti et al. performed quantitative evalu-
ation of cerebral venous blood outflow in the
supine and sitting positions and found significant
alterations in cerebral venous outflow in MS
patients in comparison with healthy controls and
patients with other neurological diseases ([64, 65]).
While the value of these CCSVI VH criteria in
detecting venous abnormalities or developmental
variants is uncertain and operator dependent, no
other validated criteria have been proposed at this
time [23]. Recently, the International Society for
Neurovascular Disease (ISNVD) developed a
more comprehensive consensus document that
included the participation of more than
40 The Role of Diagnostic Imaging Techniques for Detection of Extracranial Venous System. . .
Fig. 5 Example of velocity (a) and volume (b) measurement over 4 s phase in internal jugular vein (IJV) (The figure was
reproduced with permission from Dolic et al. [2])
60 international experts in DS imaging [23]. DS
was proposed as a standardized screening tool for
determining CCSVI status. The protocol proposes
also the use of quantitative measures for the def-
inition of functional anomalies such as blood flow
velocity and volume (Fig. 5) that could be poten-
tially more reliable in assessing the degree of
venous outflow obstruction in the IJVs. It also
refines originally proposed VH criteria. A novel
technology called Quality Doppler Processing
(QDP) [62, 66] designed to insonate the veins at
the base of the skull may become useful in the near
future, because it is difficult to visualize the supe-
rior petrosal sinus, the inferior petrosal sinus, and
the contralateral inferior and superior petrosal
sinuses. Because insonation of the petrosal sinuses
requires a Doppler angle close to 90 , detection of
reflux in the form of bidirectional flow can be made
only using a multi-angle Doppler system. The
QDP technology enables the operator to determine
the direction of blood flow within the examined
cerebral veins and make an adjustment of the pulse
repetition frequency (PRF) to clearly visualize the
963
direction without background Doppler noise. But
because of its limited availability, QDP technology
can only be considered as an additional criterion
for diagnosis of CCSVI.
Also recently, the European Society of
Neurosonology and Cerebral Hemodynamics
(ESNCH) expressed considerable concerns regard-
ing the accuracy of the proposed criteria for CCSVI
in MS [19] and proposed the central blinded DS
reading as part of a recent multicenter Italian
CoSMo study investigating the prevalence of
CCSVI in MS patients, healthy controls, and
patients with other neurologic diseases (OND) [67].
Dynamic contrast-enhanced exams can poten-
tially increase the value of DS by overcoming the
operator-dependent analysis of the ultrasonic sig-
nal [68]. Unlike the diffusible agents commonly
used in MR imaging and computed tomography
(CT), DS contrast agents are true intravascular
agents that remain entirely within the vascular
space and possess intravascular characteristics
similar to those of red blood cells. This technique
allows registration of changes in ultrasonic signal
964
intensity during and after intravenous injection of
a contrast agent. Time dependence of enhance-
ment can be shown in a time-intensity curve
(TIC) where the average signal from a region of
interest (ROI) is plotted as a function of time.
Tissue and vascular uptake, transit times, and
washout of the contrast agent can also be obtained
from a TIC [69, 70]. Mancini et al. used contrast-
enhanced ultrasonography of the internal jugular
vein with TIC analysis and revealed significantly
reduction of washout rate in IJVs in MS patients
compared to healthy controls [68].
Advantages
DS has the advantage among other diagnostic
techniques of being noninvasive, providing
high-resolution images with real-time dynamic
information such as flow and velocity, showing
intraluminal as well as extraluminal abnormali-
ties and developmental variants (Fig. 4), and
being considerably less expensive than other
noninvasive imaging techniques. DS imaging
can also be readily applied in the follow-up
period of subjects undergoing endovascular
treatment because it can recognize the associated
complications (residual stenosis, restenosis, or
venous thrombosis) (Fig. 4) [2, 5, 13, 22, 71].
DS can also assess the hemodynamic conse-
quences of outflow derangement, while B-mode
ultrasound detects structural venous intraluminal
anomalies (Fig. 4) [22, 71].
Disadvantages
The main criticism of the recommended DS pro-
tocol is that its reproducibility depends on the
training level and skills of the operator and it is
not easy to be blinded and standardized in either a
research or clinical setting [7, 19, 72, 73]. More-
over, the value of the CCSVI VH criteria is
controversial because they combine functional
and structural intra- and extracranial venous
abnormalities/developmental variants in a single
binary composite. Further, the assessment of the
second CCSVI criterion (reflux in deep cerebral
veins) (Fig. 6) is particularly controversial
because the direction of the blood flow in veins
connecting cortical with deep veins may vary
considerably as a consequence of the physiologic
K. Dolic and R. Zivadinov
interindividual variation of the cerebral venous
anatomy [72].
DS also has limits regarding extracranial vein
characterization, since findings can be influenced
by hydration status. It is a very time-consuming
method, and visualization of the central veins,
particularly in the thorax and abdomen, is often
limited and cannot give the global view of vein
anatomy. Although it can detect extracranial col-
lateral veins which are probably associated with
CCSVI, it is not technically feasible to follow the
complete course of the collateral veins which can
be more easily visualized with the use of magnetic
resonance venography (MRV), computed tomog-
raphy venography (CTV), or CV. Other pitfalls in
DS imaging include the misidentification of veins.
Additionally, overlying bone and muscle may
prevent continuous imaging (cannot visualize
suitably the confluence of the IJVs and the sub-
clavian vein because the clavicle commonly
blocks direct visualization). Similarly, the cervical
part of IJV and the jugular bulb cannot be visual-
ized by DS because of the limited acoustic win-
dow resulting from the spine, mandible, and skull.
So far, none of the recently published DS studies
have reproduced the originally reported CCSVI
prevalence, regardless of diagnostic DS method
utilized [5, 48, 71, 74, 75].
Magnetic Resonance Venography
Description of Imaging Techniques
During the past decade, catheter-based digital
subtraction angiography, as the preferred method
for imaging of the intracranial venous anatomy,
has been increasingly supplanted by MRV, an
often overlooked and underappreciated noninva-
sive and safe method for the evaluation of head
and neck veins [76, 77]. In the absence of better
noninvasive techniques for the imaging of the
dural venous sinuses, well-known and
documented pitfalls associated with flow-
sensitive MR techniques have been tolerated [34,
56, 78–81]. Furthermore, simple protocols that
incorporate two-dimensional time-of-flight
(2D-TOF) pulse sequence have already improved
their accuracy for the diagnosis of deep venous
thrombosis involving the femoral, popliteal, or
40 The Role of Diagnostic Imaging Techniques for Detection of Extracranial Venous System. . . 965

Fig. 6 Example of reflux in the deep cerebral veins using Quality Doppler Profile (QDP). Doppler profile on opposite
sides of baseline (The figure was reproduced with permission from Dolic et al. [2])

iliac veins [82]; however, experience with MRV
techniques in the cervical veins remains still
limited.
The CCSVI concept triggered several MRV
studies that emphasized the value of MRV as a
comprehensive, noninvasive, and relatively
operator-independent technique which provided
a 3D structural assessment of the intra- and extra-
cranial vasculature for the potential identification
of stenosis and quantification of blood flow
through major veins [7, 48, 56, 78, 79, 81,
83–90]. But so far almost none of the MRV stud-
ies found significant differences in extracranial
vein morphology between MS patients and con-
trols. It could be hypothesized that those so-called
abnormalities in venous outflow likely reflect nor-
mal developmental anatomic variants. Further, the
findings of these studies suggest also that MRV
morphologic information by itself may be insuffi-
cient to allow conclusions to be drawn about the
presence of venous abnormalities and their rela-
tionship to CCSVI in MS.
Time of Flight
TOF venography has the advantage of simplicity
because no special pulse sequences are required,
and this technique is available on nearly every
MRI system. TOF pulse sequences are spoiled
gradient-echo or gradient-echo acquisitions
performed sequentially, i.e., all “phase-encode”
steps are played out in a single slice before mov-
ing on to the next slice, which results in much
greater suppression of stationary tissue [34, 77,
82]. It also has the advantage of avoiding the need
for the use of contrast agents, and it remains the
technique of choice in the evaluation of the pregnant
patient with suspected dural sinus thrombosis [34,
78]. Furthermore, the accompanying conventional
MR study is more sensitive in terms of the detection
of cortical venous infarction than a CT [91].
966
Fig. 7 Example of normal and abnormal flow in internal
jugular vein on magnetic resonance venography. Normal
flow in both internal jugular veins (a) and abnormal flow in
The axial orientation of the acquisition allows for
high in-plane resolution, which is ideal for cross-
sectional area (CSA) measurements of the veins
(Fig. 7). However, the TOF sequence is easily
affected by motion artifacts, especially from the
patient’s breathing, swallowing, snoring, or head
motion [87, 92]. Relative insensitivity to in-plane
flow is another limitation of the TOF technique.
Regarding the direction of flow, the optimal acqui-
sition plane lies orthogonal, which is inefficient
from the standpoint of acquisition time and not
always achievable. Although it has a higher spatial
resolution, 2D TOF may overestimate stenosis in
the setting of turbulent or slow flow [78]. All in all,
standard conventional MRV techniques are more
prone to artifacts than phase-contrast MRV and
multidirectional TOF angiography [34, 56]. These
techniques can potentially alleviate some of the
usual MRV artifacts and provide more detailed
flow information. One obvious improvement is to
image at higher field strength, such as 3 T, because
this increases signal-to-noise ratio and better char-
acterizes slow flow.
Phase-Contrast Imaging
In contrast to TOF techniques, which rely mainly
on flow-related enhancement for producing vas-
cular images, phase-contrast MR angiography
(PC-MRA) uses velocity-induced phase shifts
K. Dolic and R. Zivadinov
left internal jugular vein on axial 2D time of flight (b). The
figure was reproduced with permission from Dolic et al. [2]
imparted upon the moving spins to distinguish
flowing blood from the surrounding stationary
tissue, thus providing information regarding
both anatomy and flow (Fig. 8). The major
advantage of PC-MRI angiography is excellent
background suppression as well as quantitative
determination of blood velocities [93, 94]. How-
ever, it requires long imaging times and a prior
estimate of blood flow velocity. Furthermore, it
may also be more sensitive to signal loss due to
turbulence or intravoxel dephasing. So far there
are only a few studies that used PC-MRI to quan-
tify venous flow in MS patients compared to
healthy controls [92, 95, 96]. More studies are
needed to validate the venous flow at the upper
neck level on an adequate number of age- and
gender-matched healthy controls with heteroge-
neous age groups.
Contrast-Enhanced Techniques
Contrast-enhanced (CE) MRV, 3D time-resolved
imaging of contrast kinetics (TRICKS) angiogra-
phy, is a noninvasive and safe method for the
evaluation of head and neck veins, without the
attendant risks of conventional angiography. It is
preferred over TOF angiography because contrast
medium reduces the T1 relaxation time of blood
and virtually eliminates the effect of saturation
([2, 97, 98]; Fig. 9).
40 The Role of Diagnostic Imaging Techniques for Detection of Extracranial Venous System. . . 967

Fig. 8 Example of internal

jugular vein pathology on a c
cine phase-contrast MRI
study. The regions of
interest (ROIs) outlined are
the internal jugular veins.
These ROIs were used to
measure the flow through
these vessels. An example
showing the flow
quantification magnitude
image in stenotic (a) and b d
normal IJV (b) and the flow
quantification on phase
images of the same IJVs (c,
d). Graph showing the
differences in velocity
between stenotic and
non-stenotic IJV (e) (The
figure was reproduced with
permission from Dolic
et al. [2])

Instantaneous Velocity Over Cardiac Cycle

e in the Right Internal Jugular
0
0 5 10 15 20 25 30 35

−5
Instantaneous Velocity (cm/s)

−10

−15

Representative_Stenotic_Velocity
−20
Representative_Nonstenotic_Velocity

−25
Instantaneous Phase of Cardiac Cycle

CE MRV is probably the most widely used
technique and is essentially identical to 3D CE
MR angiography, employing a 3D spoiled
gradient-echo sequence in conjunction with a
bolus of gadolinium-based contrast. Vascular con-
trast results from the T1-shortening effects of gad-
olinium on adjacent water protons and has
relatively little dependence on inflow effects.
In contrast to MRA, the limitation of CE MRV is
that maximal contrast enhancement achieved in
veins is typically lower than arteries because the
contrast bolus is more dilute by the time it reaches
the venous system [99]. To improve background
suppression and emphasize vascular signal, fat
saturation can be added to a 3D spoiled gradient-
echo sequence with a small increase in acquisition
968
Fig. 9 Example of normal and abnormal flow morphol-
ogy in internal jugular vein on magnetic resonance venog-
raphy. Normal (a, b) flow morphology in both and
abnormal (c, d) flow in the left internal jugular vein on
time. 3D reconstruction of CE MRV data is some-
what less straightforward than MR angiography
reconstruction since the vein/background contrast
is lower and there is usually arterial as well as
venous enhancement [2, 91, 100].
Veins can have variable MR imaging signal
intensity due to entry slice phenomenon,
in-plane flow, and flow turbulence effects and
can have variable enhancement. The maximum
intensity projection (MIP) volumetric reconstruc-
tions of these sequences often underestimate the
vascular caliber, especially when there are seg-
ments with decreased flow (velocity or volume)
[2, 91].
Disadvantages of CE MRV include the
expense of the contrast agent as well as contrast
toxicity and patient discomfort in obtaining
antecubital venous access. In case of dural sinus
thrombosis however, confident early diagnosis of
this common and treatable disease can dramati-
cally reduce patient morbidity.
4D Flow Imaging
Another promising MR technique is cine velocity-
encoded phase-contrast 4D flow that may permit
K. Dolic and R. Zivadinov
enhanced 3D time-resolved imaging of contrast kinetics
(TRICKS) (The figure was reproduced with permission
from Dolic et al. [2])
evaluation of not only anatomic stenoses but also
their impact on venous waveforms. It is based on
the principle that moving protons change phase in
proportion to their velocity. By enabling a quali-
tative assessment of the presence and direction of
collateral circulation, velocity-encoded cine MR
imaging provides information about the presence
and severity of obstruction. The technique has
been most extensively used for the evaluation of
patterns of blood flow in the thoracic aorta,
including the characterization of abnormal flow
patterns associated with pathologic disorders such
as ascending aortic aneurysm and dissection
[101]. Recent studies have explored the use of
4D flow imaging for other areas of vascular anat-
omy and pathology, including intracranial arterial
and venous blood flow [79]. With its detailed
characterization of complex, dynamic blood-
flow patterns and its ability to quantify flow, the
technique could supplement both current nonin-
vasive and invasive imaging of intra- and extra-
cranial vascular pathologic disorders. The
diagnostic and monitoring value of 4D flow imag-
ing of venous flow anomalies, indicative of
CCSVI, is currently lacking.
40 The Role of Diagnostic Imaging Techniques for Detection of Extracranial Venous System. . .
Advantages
In relation to DS, the advantages are driven by
MRV being a noninvasive technique, less time-
consuming, and less operator dependent. MRV
can also depict, easily and globally, the anatomy
and morphology of the intra and extracranial
venous system [2, 34, 102]. MRV is also
extremely useful in detecting collateral veins,
which probably represent physiological variations
of the venous system that may play a compensa-
tory role when there are more venous extracranial
abnormalities present because they bypass
blocked veins and thereby reduce resistance to
drainage [17, 56].
The assessment of the possible prominence or
collateralization of the extracranial veins in the
neck by MRV is an important diagnostic step in
examining the status of the venous system. Fur-
ther, thoracic central veins are largely inaccessible
by DS, and MRV is an excellent technique for the
assessment of axillary, jugular, subclavian, supe-
rior vena cava, and pulmonary veins. Addition-
ally, CTV always requires the use of intravenous
contrast, while many non-contrast methods are
available with MRV, making MRV the preferred
technique in patients who also suffer from renal
insufficiency or contrast allergy. CTV may also
require two or more acquisitions to adequately
capture contrast opacification of the veins, thereby
increasing the radiation dose [2, 34, 56].
Disadvantages
Finally, MRV suffers from its “snapshot” nature.
An accurate depiction of these veins requires mul-
tiple views and maneuvers such as inspiration and
expiration, flexion and extension, as well as rota-
tion imaging of the neck. Its main disadvantages
are the lack of MRV dynamism in real time and
lower resolution compared with DS and CV (can-
not evaluate intraluminal pathology such as the
immobile valves, webs, septations, membranes,
and duplications) and it is affected by the nature
of the veins themselves, which are prone to col-
lapse under frequently encountered conditions, as
opposed to arteries. MRV often detects spurious
stenoses that are not confirmed by CV, especially
in the lower parts of IJVs. These stenoses may
represent transient phasic narrowings (functional)
969
or may result from diminished flow above true
stenoses commonly located at the confluent
region of the veins. Additionally, it cannot satis-
factorily evaluate the azygos and hemiazygos
veins [2, 48, 78, 85, 103].
Unlike DS, with most MR scanners, data can
only be collected in the supine position, although
some scanners can do an upright scanning as well.
In a study that used positional MR imaging to
describe the influence of positional changes on
the cerebral venous outflow, it has been found
that IJV strictures are a common finding in healthy
controls in the supine position without relevance
in the erect position, which questions the validity
of the DS VH criterion 5 (lack of collapse of the
IJV in upright posture) for the diagnosis of
CCSVI. It is obvious that this criterion (to study
the change in flow in the IJVs from supine to
sitting position) cannot be studied with the con-
ventional MR system [104].
CTV is another multi-planar imaging tech-
nique complementary to MRV in global venous
system evaluation but with better spatial resolu-
tion and possibility of direct assessment of the
azygos vein (morphology, caliber, course, and
possible calcifications). Although it is more avail-
able and less expensive as well as time-consuming
method, there are no case-control CTV studies in
MS or other neurological patients [2, 50, 56]. This
is most likely due to the potential for radiation
exposure to controls and need for use of contrast
agent. Buffalo group gained preliminary experi-
ence by using CTV as part of a multimodal diag-
nostic approach in a currently ongoing
“Prospective Randomized Endovascular Therapy
in MS (PREMiSe)” study (Fig. 10; [9, 105]).
Plethysmography
Description of Imaging Techniques
Plethysmography is the only existing practical
noninvasive modality for global physiologic eval-
uation of extremity veins. Moreover, the use of
strain-gauge or air plethysmography to diagnose
venous thrombosis in the lower extremities has
been well documented [89, 106–108]. By inflat-
ing a cuff on the thigh, the constriction of veins
causes the venous volume to rise. When the cuff is
970
Fig. 10 Example of multimodality imaging of extracra-
nial neck veins in the Prospective Randomized
Endovascular Therapy in MS (PREMiSe) study. Axial
2D time of flight (a), enhanced 3D time-resolved imaging
of contrast kinetics (b, c), Doppler sonography (d), catheter
released, the sensor detects rapid venous runoff
and a return to the resting blood volume. If throm-
bosis is present, the plethysmography will detect a
delay in the emptying process.
The identification and assessment of venous
obstruction by plethysmography is based on an
estimation of these two parameters: venous capac-
itance and venous resistance. Quite recently,
plethysmography has been used to measure endo-
thelial function as well as the vascular response to
K. Dolic and R. Zivadinov
venography (e), intravascular sonography (f), and axial
computed tomography venography (g) all showing venous
abnormality of the left internal jugular vein (narrowing)
(The figure was reproduced with permission from Dolic
et al. [2])
vasoactive agents [106]. The technique is rarely
used in the cervical region. Despite their value in
the anatomical localization of disease, imaging
modalities such as DS and CV cannot assess the
global severity of reflux or obstruction.
Zamboni et al. recently showed that cervical
plethysmography is much less prone to operator
error compared to DS and has great potential to be
used as an inexpensive diagnostic tool for dem-
onstrating extracranial venous anomalies and
40 The Role of Diagnostic Imaging Techniques for Detection of Extracranial Venous System. . .
development variants [109]. Another study, plac-
ing a strain-gauge collar around CCSVI patients’
necks and tipping them from the upright (90 ) to
supine position (0 ) in a chair showed that hemo-
dynamics of the extracranial venous system are
greatly altered in CCSVI patients [110]. Further,
Viola et al. used transcranial brain photoplethys-
mography to measure the increase in cerebral
blood volume during both IJV compression for
10s in sitting position and concluded that it can be
used as a potential marker for diagnosis of
CCSVI [111].
Further research is needed in identifying cutoff
values and the reproducibility of the test along
with assessing intra- and interobserver variability.
Advantages
The advantage of plethysmography is that it is a
noninvasive technique that provides valuable
information regarding the impact of reflux and
obstruction on overall venous function and can
provide a measure of calf muscle pump function
(strain-gauge plethysmography) [102, 104]. It is
complementary modality to DS and is also valu-
able for monitoring the dynamics of venous dis-
ease over time and for the evaluation of treatment
outcomes [109, 110].
Disadvantages
Unfortunately, as with their invasive counterparts,
most of the noninvasive tests display the funda-
mental dichotomy of providing either anatomic or
hemodynamic information. Plethysmography can
be prone to a higher false-positive rate due to
venous compression arising from incorrect patient
positioning or the action of extrinsic masses
[108]. It is also a time-consuming method.
Invasive Imaging Modalities
Catheter Venography
Description of Imaging Techniques
CV is usually considered to be the “gold standard”
for defining the degree of stenosis in the blood
vessels associated with altered blood flow [13, 15,
85, 112, 113]. However, it has been found to be
971
less sensitive in revealing the exact nature of
narrowed extracranial vein segments. The first
finding to consider when evaluating a patient for
CCSVI is the degree of narrowing within the vein
as seen on CV and the decision as to what consti-
tutes a significant stenosis. The concept of a sig-
nificant obstruction when the vessel has been
reduced to 50 % of its diameter (which corre-
sponds to a 75 % reduction in CSA) is derived
mainly from observations in the arterial system
[85, 114, 115]. However, these criteria may not be
applicable in the venous system because there are
some fundamental differences. One potential
issue is that the IJV can vary significantly in
both size and symmetry with various factors,
including hydration status, cardiac output, respi-
ratory excursions, as well as head position that can
account for some of the noted variability [2, 33,
114]. Thus, determination of the diameter of the
vein by CV is often arbitrary, and therefore it
under- or overestimates the proper size of the
balloon for the angioplasty. In light of the high
pressures necessary to dilate the stenosis, proper
sizing is crucial to avoid injury to the vein by
over-dilatation or early recurrent stenosis by
under-dilatation (Fig. 11).
The recent results from Prospective Random-
ized Endovascular Treatment in MS (PREMiSe)
study also showed that CV may not be sensitive
enough to reveal the exact nature of narrowed vein
segments [5, 9, 116]. Veroux et al. in a large CV
study found venous angioplasty to be effective in
only a minority of MS patients with severe
impaired IJV flow [117]. Further, in a recent
published multicenter CV study, Traboulsee
et al. found that extracranial venous narrowing
of greater than 50 % is a frequent finding in
patients with multiple sclerosis, unaffected sib-
lings, and unrelated controls, which also calls in
question the existence of CCSVI as a clinical
entity [55].
Uncertainties related to the use of CV for
detection of extracranial venous abnormalities
can be grouped into several domains: the pre-
ferred vascular access for CV, whether IVUS
should accompany a routine CV evaluation for
CCSVI, whether a phasic stenosis (i.e., a
narrowing which is transient or dynamic) should
972
Fig. 11 Catheter venography of azygos and internal jug-
ular veins. Example of normal patent lumen of the azygos
vein (a) and left internal jugular vein (IJV) (b). Significant
be regarded as a normal variant or pathological,
and how to classify venographically obvious
CCSVI lesions [2, 9, 23, 116].
Advantages
CV has several important advantages, including
the ability to perform pressure gradient measure-
ments as well as to provide a helpful “road map”
for planning endovascular procedures [2, 17, 115].
It is a commonly performed low-risk procedure
with a safety record established over decades of
experience [27, 112, 118].
Disadvantages
Its invasiveness, use of contrast agents, and radi-
ation exposure make it suboptimal as a routine
screening tool in a clinical setting. It is also oper-
ator dependent, only AP projection views are rou-
tinely obtained, and stenosis assessment may
depend on the precise locations and rates of con-
trast injection; CV can only show the collaterals
that drain the specific vein being injected without
the possibility of showing global extracranial
venous system at once, i.e., as with MRV or
CTV [2, 34, 116]. The display of extracranial
venous structures can be improved with additional
injected contrast medium, more selective cathe-
terization, and additional projections [2, 116].
Although CV is a luminogram, it brings little or
no data regarding the vessel’s intraluminal struc-
tures, because of dense opacification of the lumen
with contrast, which obliterates subtle
intraluminal structures like valve malformations,
K. Dolic and R. Zivadinov
stenosis of the proximal right IJV (c) (The figure was
reproduced with permission from Dolic et al. [2])
septa, and flaps that represent majority of CCSVI
pathology. Further, there is the concern that an
intraluminal anomaly such as septae may easily
be displaced out of the way by an inflated balloon
but upon deflation fall right back in its original
position and continue to functionally obstruct
flow [115, 116]. It is possible to use very dilute
contrast and cone-downed images at high rate of
acquisition to pick up some of these intraluminal
features, but they are generally harder to detect on
CV using conventional acquisition parameters
and contrast strengths. On the other hand, these
abnormalities can be seen with high-resolution
color DS or intravascular ultrasound (IVUS)
B-mode imaging [116]. In addition, malformed
and/or reversed valve cusps can be crossed by
the catheter and kept open artificially, thereby
preventing the documentation of stenosis.
Intravascular Sonography
Description of Imaging Techniques
IVUS is an endoluminal CV-based DS technique
that offers a tomographic, 360 view of the ves-
sel’s wall from the inside [119]. The most com-
mon indications for IVUS have been in the
evaluation and treatment of arterial disease
[120–122]. Its excellent resolution compared
with angiography has contributed to the under-
standing of the pathophysiology and enhanced
diagnosis of coronary artery disease achieving
new milestones in interventional cardiology.
IVUS has been shown to provide a more accurate
40 The Role of Diagnostic Imaging Techniques for Detection of Extracranial Venous System. . .
Fig. 12 Example of intravascular ultrasound in the internal jugular vein. Normal patent lumen (a) and stenotic lumen (b)
with fibrotic wall (The figure was reproduced with permission from Dolic et al. [2])
assessment of vessel circumference and CSA and
thus is useful in detecting critical stenoses [123,
124, 125]. While values for stenosis definition
used for CV (50 %) rely on a ratio between the
stenotic segment diameter and a pre(non)-stenotic
vein which is more variable, the IVUS definition
is more strict (a lumen that embraces the IVUS
probe for a critical stenosis) and does not refer to a
non-stenotic segment. It remains unclear at what
level and with what criteria is there a significant
hemodynamic effect of stenosis by either modal-
ity [116]. Venous stenosis is currently measured
using arterial criteria, which are clearly not opti-
mal. The hemodynamics of venous flow remain a
major area of investigation, and better understand-
ing will likely lead to a revision of stenosis
criteria.
Although diagnostic experience is growing
with the use of IVUS for investigation of both
intra- and extracranial arteries, there is limited
literature regarding its use for the exploration of
venous vasculature in general, as well as specifi-
cally in relation to the investigation of venous
abnormalities and developmental variants indica-
tive of CCSVI (Fig. 12) [9, 116, 126, 127]. Recent
studies showed that the IVUS assessment of IJVs
and azygos vein can detect higher rates of venous
abnormalities than CV and that provides a diag-
nostic advantage over the “gold-standard” CV in
detecting extracranial venous abnormalities and
developmental variants indicative of CCSVI
[116, 126]. Also, PREMiSe study showed that
IVUS is more accurate in the detection of
973
intraluminal venous anomalies in IJVs and azygos
vein and more accurate in measurement of steno-
sis and wall thickness and allows for the explora-
tion of pulsatility in the veins [5].
Advantages
The exploration of IJV valves is particularly well
seen on IVUS. Additionally, thrombus and dissec-
tions are readily seen on IVUS. It can also show
the degrees of echogenicity, both of the vessel
wall and of the intraluminal thrombi, which may
indicate varying degrees of wall thickness and
may correlate with the age of the thrombosis, an
important aspect of the vessel pathology that is not
possible to be determined with CV [80, 119].
Additionally, it provides an image with a
greater resolution of both lumen and wall (with
additional 3D features), providing better vessel
wall information. IVUS imaging may reflect
truly the size of stenotic lesions because it allows
more complete and accurate assessment than is
possible with the use of CV examination
[116]. Further, analysis of the vessel dimensions
allows a more accurate selection of balloon size,
thus reducing the risk of injury and providing a
more effective angioplasty. It provides cross-
sectional, in vivo visualization and the demonstra-
tion of the motility of small intraluminal struc-
tures, like bulging cusps as well as septum and
webs, which cannot be optimally revealed by
traditional diagnostic methods [9, 116, 126]. It
has been shown that such venous pathology in
the iliac vein is unrecognized by CV and is well
974
visualized by IVUS [79]. Moreover, CV is inca-
pable of monitoring respiratory pulsatility which
involves periods with reduced vessel diameter
that can be investigated with IVUS [80].
The advantages of IVUS compared with DS,
among others, include the sonographic penetra-
tion from within the vessel by excluding extravas-
cular soft tissues. It also assesses blood vessels not
easily accessible by conventional DS, such as the
lower part of the IJV (behind the clavicle), upper
part of the IJV, intracranial sinuses, and azygos
vein [9, 126].
Disadvantages
Ring-down artifacts produced by acoustic oscilla-
tions in the piezoelectric transducer that obscures
the near field result in an acoustic catheter size
larger than its physical size and may adversely
affect IVUS images. Geometric distortion can
result from imaging in an oblique plane (not per-
pendicular to the long axis of the vessel) [2,
81]. Furthermore, visible distortion of the image
can be due to another important artifact,
“nonuniform rotational distortion,” which arises
from uneven drag on the drive cable of the mechan-
ical style catheters, resulting in cyclical oscillations
in rotational speed. The physical size of IVUS
catheters (currently  1.0 mm) constitutes an
important limitation in the imaging of severe ste-
noses. Further depending on the probe, there is a
finite limit to IVUS resolution which rapidly
degrades beyond this particular radius typically
10–12 mm [81]. In summary, the frequency of the
transducer, gain settings, depth of penetration, and
focal depth are some of the factors that affect the
sensitivity of the IVUS imaging.
Multimodal Imaging Approach
The dramatic difference in prevalent findings
between different studies using noninvasive and
invasive imaging techniques emphasizes the
urgent need for the use of a multimodal imaging
approach for better understanding of the extracra-
nial venous abnormalities and developmental
variants indicative of CCSVI. In a number of
recent studies, noninvasive and invasive imaging
K. Dolic and R. Zivadinov
techniques were applied and compared [7, 17, 48,
62, 78, 83, 85, 100, 111, 116, 117, 118, 126, 128,
129, 130]. The findings of these studies are
extremely important to understand the true preva-
lence of CCSVI, and the comparison of invasive
vs. noninvasive imaging findings is especially
important in this endeavor. It is emerging that
the prevalence of venous abnormalities and devel-
opmental variants, indicative of CCSVI, is even
higher when investigated with sophisticated inva-
sive imaging techniques [63, 116, 126, 127].
Based on these recent findings, a multimodal
approach is recommended to determine whether
CCSVI exists as a clinical entity and not as an
anatomic variant and to what extent it is present in
various healthy and disease groups as well as MS
subtypes (Fig. 10; [23]). The introduction of more
quantitative criteria to describe extracranial
venous structural and hemodynamic functional
impairment in future multimodal approach studies
will be a significant improvement compared to the
current binary CCSVI diagnosis.
Summary
There is an urgent need to define and validate the
spectrum of extracranial venous abnormalities
and to establish reliable, diagnostic gold-standard
test(s). Each noninvasive and invasive imaging
modality has its own inherent advantages and
disadvantages [2, 9]. Most likely, only multi-
modal imaging will eventually become the reli-
able screening, diagnostic, and monitoring tool
for the assessment of the extracranial venous sys-
tem [23]. The classification of the presence and
severity of extracranial venous abnormalities/
developmental variants by imaging and pathology
findings should be the first step in the determina-
tion of their role in the pathology of CNS disor-
ders and aging.
The use of noninvasive methods to confirm
presence of extracranial venous abnormalities
will be an essential step toward better understand-
ing of its importance in general population and
disease states. Regarding noninvasive techniques
there is still no consensus on DS protocols or use
of MRV as alternative approach which could
40 The Role of Diagnostic Imaging Techniques for Detection of Extracranial Venous System. . .
ensure appropriate quality control for the determi-
nation of extracranial venous abnormalities
[23]. Thus, at present, the true prevalence of extra-
cranial venous abnormalities in MS patients and
other CNS disorders or healthy population has not
been adequately assessed.
Nevertheless, the concept of CCSVI has trig-
gered an intense and rapid accumulation of
knowledge over the past 4 years and has stimu-
lated the need for further research to better under-
stand the function and potential role of the
extracranial venous system. Only properly
designed, safe, and ethical studies using noninva-
sive and invasive imaging techniques IVUS
should be encouraged in collecting this longitudi-
nal observational information.
References
1. Eberhardt RT, Raffetto JD (2005) Chronic venous
insufficiency. Circulation 111(18):2398–2409
2. Dolic K, Siddiqui AH, Karmon Y, Marr K, Zivadinov
R (2013) The role of noninvasive and invasive diag-
nostic imaging techniques for detection of extra-
cranial venous system anomalies and developmental
variants. BMC Med 11:155
3. Burrows PE, Konez O, Bisdorff A (2003) Venous
variations of the brain and cranial vault. Neuroimag-
ing Clin N Am 13(1):13–26
4. Dake MD, Zivadinov R, Haacke EM (2011) Chronic
cerebrospinal venous insufficiency in multiple sclero-
sis: a historical perspective. Funct Neurol 26(4):
181–195
5. Zivadinov R, Chung CP (2013) Potential involve-
ment of the extracranial venous system in central
nervous system disorders and aging. BMC Med
11:260
6. Lee BB, Laredo J, Neville R (2010) Embryological
background of truncular venous malformation in the
extracranial venous pathways as the cause of chronic
cerebro-spinal venous insufficiency. Int Angiol
29:95–108
7. Dolic K, Marr K, Valnarov V et al (2012) Intra- and
extraluminal structural and functional venous anom-
alies in multiple sclerosis, as evidenced by 2 noninva-
sive imaging techniques. AJNR Am J Neuroradiol
33(1):16–23
8. Chung CP, Cheng CY, Zivadinov R et al (2013) Jug-
ular venous reflux and plasma endothelin-1 are asso-
ciated with cough syncope: a case control pilot study.
BMC Neurol 13:9
9. Zivadinov R, Karmon Y, Dolic K et al (2013) Multi-
modal noninvasive and invasive imaging of extracra-
nial venous abnormalities indicative of CCSVI:
975
results of the PREMiSe pilot study. BMC Neurol
13:151
10. Chuang YM, Hu HH (2005) Cough headache and
thoracic inlet valvular competence in uremia. Eur
Neurol 53(2):78–80
11. Chung CP, Hsu HY, Chao AC, Sheng WY, Soong
BW, Hu HH (2007) Transient global amnesia: cere-
bral venous outflow impairment-insight from the
abnormal flow patterns of the internal jugular vein.
Ultrasound Med Biol 33(11):1727–1735
12. Chung CP, Wang PN, Wu YH et al (2011) More
severe white matter changes in the elderly with jugu-
lar venous reflux. Ann Neurol 69(3):553–559
13. Zamboni P, Galeotti R, Menegatti E et al (2009) A
prospective open-label study of endovascular treat-
ment of chronic cerebrospinal venous insufficiency.
J Vasc Surg 50(6):1348–1358, e1-3
14. Zamboni P, Galeotti R, Menegatti E et al (2009)
Chronic cerebrospinal venous insufficiency in
patients with multiple sclerosis. J Neurol Neurosurg
Psychiatry 80(4):392–399
15. Al-Omari MH, Al-Bashir A (2012) Internal jugular
vein valve morphology in the patients with chronic
cerebrospinal venous insufficiency (CCSVI); angio-
graphic findings and schematic demonstrations. Rev
Recent Clin Trials 7(2):83–87
16. Haacke EM (2011) Chronic cerebral spinal venous
insufficiency in multiple sclerosis. Expert Rev
Neurother 11(1):5–9
17. Zamboni P, Menegatti E, Galeotti R et al (2009) The
value of cerebral Doppler venous haemodynamics in
the assessment of multiple sclerosis. J Neurol Sci
282(1–2):21–27
18. Zamboni P, Morovic S, Menegatti E, Viselner G,
Nicolaides AN (2011) Screening for chronic cerebro-
spinal venous insufficiency (CCSVI) using ultrasoun-
d–recommendations for a protocol. Int Angiol
30(6):571–597
19. Baracchini C, Valdueza JM, Del Sette M et al (2012)
CCSVI and MS: a statement from the European Soci-
ety of neurosonology and cerebral hemodynamics.
J Neurol 259(12):2585–2589
20. Mayer CA, Pfeilschifter W, Lorenz MW et al (2011)
The perfect crime? CCSVI not leaving a trace in MS.
J Neurol Neurosurg Psychiatry 82(4):436–440
21. Oger J, Alkhawajah M (2010) CCSVI: hope, hype or
snake oil? Can J Neurol Sci 37(6):716
22. Nicolaides AN, Morovic S, Menegatti E, Viselner G,
Zamboni P (2011) Screening for chronic cerebrospi-
nal venous insufficiency (CCSVI) using ultrasound:
recommendations for a protocol. Funct Neurol
26(4):229–248
23. Zivadinov R, Bastianello S, Dake M, et al (2014)
Recommendations for multimodal noninvasive
and invasive screening for detecting extracranial
venous abnormalities, indicative of chronic cerebro-
spinal venous insufficiency (CCSVI). A position
statement of International Society for Neurovascular
Disease (ISNVD). J Vasc Interv Radiol. Sep 22.
976
pii: S1051–0443(14)00746–5. doi:10.1016/j.
jvir.2014.07.024
24. Van den Berg PJ, Visser LH (2013) The fluctuating
natural course of CCSVI in MS patients and controls,
a prospective follow-up. PLoS One 8(11):e78166
25. Cortes Nino Mdel P, Tampieri D, Melancon D (2010)
Endovascular venous procedures for multiple sclero-
sis? Mult Scler 16(7):771–772
26. Mandato KD, Hegener PF, Siskin GP et al (2012)
Safety of endovascular treatment of chronic cerebro-
spinal venous insufficiency: a report of 240 patients
with multiple sclerosis. J Vasc Interv Radiol 23(1):
55–59
27. Petrov I, Grozdinski L, Kaninski G, Iliev N, Iloska M,
Radev A (2011) Safety profile of endovascular treat-
ment for chronic cerebrospinal venous insufficiency
in patients with multiple sclerosis. J Endovasc Ther
18(3):314–323
28. Burton JM, Alikhani K, Goyal M et al (2011) Compli-
cations in MS patients after CCSVI procedures abroad
(Calgary, AB). Can J Neurol Sci 38(5):741–746
29. Valdueza JM, Doepp F, Schreiber SJ et al (2013)
What went wrong? The flawed concept of cerebrospi-
nal venous insufficiency. J Cereb Blood Flow Metab
33(5):657–668
30. Beggs CB (2013) Venous hemodynamics in neuro-
logical disorders: an analytical review with hydrody-
namic analysis. BMC Med 11:142
31. Beggs C, Chung CP, Bergsland N et al (2013) Jugular
venous reflux and brain parenchyma volumes in
elderly patients with mild cognitive impairment and
Alzheimer’s disease. BMC Neurol 13:157
32. Chung CP, Lin YJ, Chao AC et al (2010) Jugular
venous hemodynamic changes with aging. Ultra-
sound Med Biol 36(11):1776–1782
33. Zivadinov R, Marr K, Cutter G et al (2011) Preva-
lence, sensitivity, and specificity of chronic cerebro-
spinal venous insufficiency in MS. Neurology 77(2):
138–144
34. Zivadinov R, Lopez-Soriano A, Weinstock-Guttman
B et al (2011) Use of MR venography for character-
ization of the extracranial venous system in patients
with multiple sclerosis and healthy control subjects.
Radiology 258(2):562–570
35. Doepp F, Schreiber SJ, von Munster T, Rademacher J,
Klingebiel R, Valdueza JM (2004) How does the
blood leave the brain? A systematic ultrasound anal-
ysis of cerebral venous drainage patterns. Neuroradi-
ology 46(7):565–570
36. San Millan Ruiz D, Gailloud P, Rufenacht DA,
Delavelle J, Henry F, Fasel JH (2002) The
craniocervical venous system in relation to cerebral
venous drainage. AJNR Am J Neuroradiol 23(9):
1500–1508
37. Meder JF, Chiras J, Roland J, Guinet P, Bracard S,
Bargy F (1994) Venous territories of the brain.
J Neuroradiol 21(2):118–133
38. Schaller B (2004) Physiology of cerebral venous
blood flow: from experimental data in animals to
K. Dolic and R. Zivadinov
normal function in humans. Brain Res Brain Res
Rev 46(3):243–260
39. Schmidek HH, Auer LM, Kapp JP (1985) The cere-
bral venous system. Neurosurgery 17(4):663–678
40. Caruso RD, Rosenbaum AE, Chang JK, Joy SE
(1999) Craniocervical junction venous anatomy on
enhanced MR images: the suboccipital cavernous
sinus. AJNR Am J Neuroradiol 20(6):1127–1131
41. Ono M, Rhoton AL Jr, Peace D, Rodriguez RJ (1984)
Microsurgical anatomy of the deep venous system of
the brain. Neurosurgery 15(5):621–657
42. Andeweg J (1996) The anatomy of collateral venous
flow from the brain and its value in aetiological inter-
pretation of intracranial pathology. Neuroradiology
38(7):621–628
43. Epstein HM, Linde HW, Crampton AR, Ciric IS,
Eckenhoff JE (1970) The vertebral venous plexus as
a major cerebral venous outflow tract. Anesthesiology
32(4):332–337
44. Akkawi NM, Agosti C, Borroni B et al (2002) Jugular
valve incompetence: a study using air contrast ultra-
sonography on a general population. J Ultrasound
Med 21(7):747–751
45. Brownlow RL Jr, McKinney WM (1985) Ultrasonic
evaluation of jugular venous valve competence.
J Ultrasound Med 4(4):169–172
46. Silva MA, Deen KI, Fernando DJ, Sheriffdeen AH
(2002) The internal jugular vein valve may have a
significant role in the prevention of venous reflux:
evidence from live and cadaveric human subjects.
Clin Physiol Funct Imaging 22(3):202–205
47. Fisher J, Vaghaiwalla F, Tsitlik J et al (1982) Deter-
minants and clinical significance of jugular venous
valve competence. Circulation 65(1):188–196
48. Dolic K, Marr K, Valnarov V et al (2011) Sensitivity
and specificity for screening of chronic cerebrospinal
venous insufficiency using a multimodal
non-invasive imaging approach in patients with mul-
tiple sclerosis. Funct Neurol 26(4):205–214
49. Zamboni P, Galeotti R (2010) The chronic cerebro-
spinal venous insufficiency syndrome. Phlebology
25(6):269–279
50. Escott EJ, Branstetter BF (2006) It’s not a cervical
lymph node, it’s a vein: CT and MR imaging findings
in the veins of the head and neck. Radiographics
26(5):1501–1515
51. Deslaugiers B, Vaysse P, Combes JM et al (1994)
Contribution to the study of the tributaries and the
termination of the external jugular vein. Surg Radiol
Anat 16(2):173–177
52. Schummer W, Schummer C, Bredle D, Frober R
(2004) The anterior jugular venous system: variability
and clinical impact. Anesth Analg 99(6):1625–1629
53. Beddy P, Geoghegan T, Ramesh N et al (2006)
Valsalva and gravitational variability of the internal
jugular vein and common femoral vein: ultrasound
assessment. Eur J Radiol 58(2):307–309
54. Schreiber SJ, Lurtzing F, Gotze R, Doepp F,
Klingebiel R, Valdueza JM (2003) Extrajugular
40 The Role of Diagnostic Imaging Techniques for Detection of Extracranial Venous System. . .
pathways of human cerebral venous blood drainage
assessed by duplex ultrasound. J Appl Physiol 94(5):
1802–1805
55. Traboulsee AL, Knox KB, Machan L et al (2014)
Prevalence of extracranial venous narrowing on cath-
eter venography in people with multiple sclerosis, their
siblings, and unrelated healthy controls: a blinded,
case–control study. Lancet 383(9912):138–145
56. Zivadinov R, Galeotti R, Hojnacki D et al (2011)
Value of MR venography for detection of internal
jugular vein anomalies in multiple sclerosis: a pilot
longitudinal study. AJNR Am J Neuroradiol 32(5):
938–946
57. Meissner MH, Moneta G, Burnand K et al (2007) The
hemodynamics and diagnosis of venous disease.
J Vasc Surg 46(S):4S–24S
58. Perone N, Bounameaux H, Perrier A (2001) Compar-
ison of four strategies for diagnosing deep vein throm-
bosis: a cost-effectiveness analysis. Am J Med
110(1):33–40
59. Goodacre S, Sampson F, Stevenson M et al (2006)
Measurement of the clinical and cost-effectiveness of
non-invasive diagnostic testing strategies for deep
vein thrombosis. Health Technol Assess 10(15):
1–168, iii-iv
60. Kassai B, Boissel JP, Cucherat M, Sonie S, Shah NR,
Leizorovicz A (2004) A systematic review of the
accuracy of ultrasound in the diagnosis of deep
venous thrombosis in asymptomatic patients. Thromb
Haemost 91(4):655–666
61. Chung CP, Beggs C, Wang PN et al (2014) Jugular
venous reflux and white matter abnormalities in
Alzheimer’s disease: a pilot study. J Alzheimers Dis
39(3):601–609
62. Zamboni P, Consorti G, Galeotti R et al (2009)
Venous collateral circulation of the extracranial cere-
brospinal outflow routes. Curr Neurovasc Res 6(3):
204–212
63. Lugli M, Morelli M, Guerzoni S, Maleti O (2012) The
hypothesis of patho-physiological correlation
between chronic cerebrospinal venous insufficiency
and multiple sclerosis: rationale of treatment. Phle-
bology 27(Suppl 1):178–186
64. Monti L, Menci E, Piu P et al (2014) A sonographic
quantitative cutoff value of cerebral venous outflow in
neurologic diseases: a blinded study of 115 subjects.
AJNR Am J Neuroradiol Jul;35(7):1381–6.
doi:10.3174/ajnr.A3864. Epub 2014 Mar 7
65. Monti L, Menci E, Ulivelli M et al (2011) Quantita-
tive Colour Doppler Sonography evaluation of cere-
bral venous outflow: a comparative study between
patients with multiple sclerosis and controls. PLoS
One 6(9):e25012
66. Ciccone MM, Galeandro AI, Scicchitano P
et al (2012) Multigate quality Doppler profiles and
morphological/hemodynamic alterations in multiple
sclerosis patients. Curr Neurovasc Res 9(2):120–127
67. Comi G, Battaglia MA, Bertolotto A et al (2013) Ital-
ian multicentre observational study of the prevalence
977
of CCSVI in multiple sclerosis (CoSMo study): ratio-
nale, design, and methodology. Neurol Sci 34(8):
1297–1307
68. Mancini M, Morra VB, Di Donato O et al (2012)
Multiple sclerosis: cerebral circulation time. Radiol-
ogy 262(3):947–955
69. Shalhoub J, Owen DR, Gauthier T, Monaco C, Leen
EL, Davies AH (2010) The use of contrast enhanced
ultrasound in carotid arterial disease. Eur J Vasc
Endovasc Surg 39(4):381–387
70. Staub D, Partovi S, Imfeld S et al (2013) Novel
applications of contrast-enhanced ultrasound imaging
in vascular medicine. Vasa 42(1):17–31
71. Thapar A, Lane T, Nicholas R et al (2011) Systematic
review of sonographic chronic cerebrospinal venous
insufficiency findings in multiple sclerosis. Phlebol-
ogy 26(8):319–325
72. Doepp F, Paul F, Valdueza JM, Schmierer K,
Schreiber SJ (2010) No cerebrocervical venous con-
gestion in patients with multiple sclerosis. Ann
Neurol 68(2):173–183
73. Menegatti E, Genova V, Tessari M et al (2010) The
reproducibility of colour Doppler in chronic cerebro-
spinal venous insufficiency associated with multiple
sclerosis. Int Angiol 29(2):121–126
74. McDonald S, Iceton JB (2012) The use of Doppler
ultrasound in the diagnosis of chronic cerebrospinal
venous insufficiency. Tech Vasc Interv Radiol 15(2):
113–120
75. Zaniewski M, Kostecki J, Kuczmik W et al (2013)
Neck duplex Doppler ultrasound evaluation for
assessing chronic cerebrospinal venous insufficiency
in multiple sclerosis patients. Phlebology 28(1):24–31
76. Prince MR, Sostman HD (2003) MR venography:
unsung and underutilized. Radiology 226(3):630–632
77. Vogt FM, Herborn CU, Goyen M (2005) MR venog-
raphy. Magn Reson Imaging Clin N Am 13(1):
113–129, vi
78. Zaharchuk G, Fischbein NJ, Rosenberg J, Herfkens
RJ, Dake MD (2011) Comparison of MR and contrast
venography of the cervical venous system in multiple
sclerosis. AJNR Am J Neuroradiol 32(8):1482–1489
79. Hope MD, Purcell DD, Hope TA et al (2009) Com-
plete intracranial arterial and venous blood flow eval-
uation with 4D flow MR imaging. AJNR Am J
Neuroradiol 30(2):362–366
80. Neglen P, Raju S (2002) Intravascular ultrasound scan
evaluation of the obstructed vein. J Vasc Surg 35(4):
694–700
81. Finet G, Maurincomme E, Tabib A et al (1993) Arti-
facts in intravascular ultrasound imaging: analyses and
implications. Ultrasound Med Biol 19(7):533–547
82. Kanne JP, Lalani TA (2004) Role of computed tomog-
raphy and magnetic resonance imaging for deep
venous thrombosis and pulmonary embolism. Circu-
lation 109(12 Suppl 1):I15–I21
83. Doepp F, Wurfel JT, Pfueller CF et al (2011) Venous
drainage in multiple sclerosis: a combined MRI and
ultrasound study. Neurology 77(19):1745–1751
978
84. Haacke EM, Feng W, Utriainen D et al (2012) Patients
with multiple sclerosis with structural venous abnor-
malities on MR imaging exhibit an abnormal flow
distribution of the internal jugular veins. J Vasc Interv
Radiol 23(1):60–68, e1-3
85. Hojnacki D, Zamboni P, Lopez-Soriano A et al (2010)
Use of neck magnetic resonance venography, Doppler
sonography and selective venography for diagnosis of
chronic cerebrospinal venous insufficiency: a pilot
study in multiple sclerosis patients and healthy con-
trols. Int Angiol 29(2):127–139
86. Utriainen D, Trifan G, Sethi S et al (2012) Magnetic
resonance imaging signatures of vascular pathology
in multiple sclerosis. Neurol Res 34(8):780–792
87. Utriainen D, Feng W, Elias S, Latif Z, Hubbard D,
Haacke EM (2012) Using magnetic resonance imag-
ing as a means to study chronic cerebral spinal venous
insufficiency in multiple sclerosis patients. Tech Vasc
Interv Radiol 15(2):101–112
88. Wattjes MP, van Oosten BW, de Graaf WL
et al (2011) No association of abnormal cranial
venous drainage with multiple sclerosis: a magnetic
resonance venography and flow-quantification study.
J Neurol Neurosurg Psychiatry 82(4):429–435
89. Heijboer H, Buller HR, Lensing AW, Turpie AG,
Colly LP, ten Cate JW (1993) A comparison of real-
time compression ultrasonography with impedance
plethysmography for the diagnosis of deep-vein
thrombosis in symptomatic outpatients. N Engl J
Med 329(19):1365–1369
90. Rosales M, Radeva P, Rodriguez-Leor O, Gil D
(2009) Modelling of image-catheter motion for 3-D
IVUS. Med Image Anal 13(1):91–104
91. Ayanzen RH, Bird CR, Keller PJ, McCully FJ,
Theobald MR, Heiserman JE (2000) Cerebral MR
venography: normal anatomy and potential diagnostic
pitfalls. AJNR Am J Neuroradiol 21(1):74–78
92. Feng W, Utriainen D, Trifan G, Sethi S, Hubbard D,
Haacke EM (2012) Quantitative flow measurements
in the internal jugular veins of multiple sclerosis
patients using magnetic resonance imaging. Rev
Recent Clin Trials 7(2):117–126
93. Alperin N, Hushek SG, Lee SH, Sivaramakrishnan A,
Lichtor T (2005) MRI study of cerebral blood flow
and CSF flow dynamics in an upright posture: the
effect of posture on the intracranial compliance and
pressure. Acta Neurochir Suppl 95:177–181
94. Stoquart-Elsankari S, Lehmann P, Villette A
et al (2009) A phase-contrast MRI study of physio-
logic cerebral venous flow. J Cereb Blood Flow
Metab 29(6):1208–1215
95. Sundstrom P, Wahlin A, Ambarki K, Birgander R,
Eklund A, Malm J (2010) Venous and cerebrospinal
fluid flow in multiple sclerosis: a case–control study.
Ann Neurol 68(2):255–259
96. Macgowan CK, Chan KY, Laughlin S, Marrie RA,
Banwell B (2014) Cerebral arterial and venous
blood flow in adolescent multiple sclerosis patients
and age-matched controls using phase contrast MRI.
K. Dolic and R. Zivadinov
J Magn Reson Imaging Aug;40(2):341–7.
doi:10.1002/jmri.24388.
97. Farb RI, Scott JN, Willinsky RA, Montanera WJ,
Wright GA, terBrugge KG (2003) Intracranial venous
system: gadolinium-enhanced three-dimensional MR
venography with auto-triggered elliptic centric-
ordered sequence–initial experience. Radiology
226(1):203–209
98. Liang L, Korogi Y, Sugahara T et al (2001) Evaluation
of the intracranial dural sinuses with a 3D contrast-
enhanced MP-RAGE sequence: prospective compar-
ison with 2D-TOF MR venography and digital sub-
traction angiography. AJNR Am J Neuroradiol 22(3):
481–492
99. Hingwala DR, Thomas B, Kesavadas C,
Kapilamoorthy TR (2011) Suboptimal contrast
opacification of dynamic head and neck MR angiog-
raphy due to venous stasis and reflux: technical con-
siderations for optimization. AJNR Am J Neuroradiol
32(2):310–314
100. Zivadinov R, Ramanathan M, Dolic K et al (2011)
Chronic cerebrospinal venous insufficiency in multi-
ple sclerosis: diagnostic, pathogenetic, clinical and
treatment perspectives. Expert Rev Neurother 11(9):
1277–1294
101. Hom JJ, Ordovas K, Reddy GP (2008) Velocity-
encoded cine MR imaging in aortic coarctation: func-
tional assessment of hemodynamic events. Radio-
graphics 28(2):407–416
102. Glockner J (2010) Magnetic resonance venography.
Appl Radiol 39(6):36
103. Hartel M, Kluczewska E, Simka M, Ludyga T,
Kostecki J, Zaniewski M (2011) Magnetic resonance
venography of chronic cerebrospinal venous insuffi-
ciency in patients with associated multiple sclerosis.
Pol J Radiol 76(1):59–62
104. Niggemann P, Seifert M, Forg A, Schild HH,
Urbach H, Krings T (2012) Positional venous MR
angiography: an operator-independent tool to evalu-
ate cerebral venous outflow hemodynamics. AJNR
Am J Neuroradiol 33(2):246–251
105. Siddiqui AH, Zivadinov R, Benedict RH et al (2014)
Prospective randomized trial of venous angioplasty in
MS (PREMiSe). Neurology Jul 29;83(5):441–9.
doi:10.1212/WNL.0000000000000638
106. Higashi Y, Yoshizumi M (2003) New methods
to evaluate endothelial function: method for assessing
endothelial function in humans using a strain-gauge
plethysmography: nitric oxide-dependent and -inde-
pendent vasodilation. J Pharmacol Sci 93(4):399–404
107. Janssen MC, Wollersheim H, Haenen JH, van Asten
WN, Thien T (1998) Deep venous thrombosis: a
prospective 3-month follow-up using duplex scan-
ning and strain-gauge plethysmography. Clin Sci
94(6):651–656
108. Locker T, Goodacre S, Sampson F, Webster A, Sutton
AJ (2006) Meta-analysis of plethysmography and
rheography in the diagnosis of deep vein thrombosis.
Emerg Med J 23(8):630–635
40 The Role of Diagnostic Imaging Techniques for Detection of Extracranial Venous System. . .
109. Zamboni P, Menegatti E, Conforti P, Shepherd S,
Tessari M, Beggs C (2012) Assessment of cerebral
venous return by a novel plethysmography method.
J Vasc Surg 56(3):677–685, e1
110. Beggs C, Shepherd S, Zamboni P (2014) Cerebral
venous outflow resistance and interpretation of cervi-
cal plethysmography data with respect to the diagno-
sis of chronic cerebrospinal venous insufficiency.
Phlebology 29(3):191–199
111. Viola S, Viola P, Fiorelli L, Buongarzone M, Litterio P
(2013) Transcranial brain photoplethysmography to
study the venules of cerebral cortex in patients with
multiple sclerosis. Phlebology. Dec 13. [Epub ahead
of print]
112. Ludyga T, Kazibudzki M, Simka M et al (2010)
Endovascular treatment for chronic cerebrospinal
venous insufficiency: is the procedure safe? Phlebol-
ogy 25(6):286–295
113. Simka M (2011) Commentary: Safety of endovascular
treatment for CCSVI and future perspectives.
J Endovasc Ther 18(3):326–327
114. Werner JD, Siskin GP, Mandato K, Englander M, Herr A
(2011) Review of venous anatomy for venographic
interpretation in chronic cerebrospinal venous insuffi-
ciency. J Vasc Interv Radiol 22(12):1681–1690; quiz 91
115. Zivadinov R, Weinstock-Guttman B (2012) Role of
venoplasty for treatment of multiple sclerosis: value
of open-label studies and surrogate treatment out-
comes. J Vasc Interv Radiol 23(10):1308–1310
116. Karmon Y, Zivadinov R, Weinstock-Guttman B
et al (2013) Comparison of intravascular ultrasound
with conventional venography for detection of extra-
cranial venous abnormalities indicative of chronic
cerebrospinal venous insufficiency. J Vasc Interv
Radiol 24(10):1487–1498, e1
117. Veroux P, Giaquinta A, Perricone D et al (2013) Inter-
nal jugular veins out flow in patients with multiple
sclerosis:a catheter venography study. J Vasc Interv
Radiol 24(12):1790–1797
118. Simka M, Latacz P, Ludyga T et al (2011) Prevalence
of extracranial venous abnormalities: results from a
sample of 586 multiple sclerosis patients. Funct
Neurol 26(4):197–203
119. Nissen SE, Yock P (2001) Intravascular ultrasound:
novel pathophysiological insights and current clinical
applications. Circulation 103(4):604–616
979
120. Mintz GS, Popma JJ, Pichard AD et al (1996) Intra-
vascular ultrasound predictors of restenosis after per-
cutaneous transcatheter coronary revascularization.
J Am Coll Cardiol 27(7):1678–1687
121. Miskolczi L, Guterman LR, Flaherty JD, Hopkins LN
(1996) Depiction of carotid plaque ulceration and
other plaque-related disorders by intravascular sonog-
raphy: a flow chamber study. AJNR Am J Neuroradiol
17(10):1881–1890
122. Potkin BN, Bartorelli AL, Gessert JM et al
(1990) Coronary artery imaging with intravascular
high-frequency ultrasound. Circulation 81(5):
1575–1585
123. Reid DB (1996) Intravenous IVUS: room with a view.
J Endovasc Ther 3(4):435
124. Reid DB, Diethrich EB, Marx P, Wrasper R (1996)
Intravascular ultrasound assessment in carotid inter-
ventions. J Endovasc Ther 3(2):203–210
125. Reid DB, Douglas M, Diethrich EB (1995) The clin-
ical value of three-dimensional intravascular ultra-
sound imaging. J Endovasc Ther 2(4):356–364
126. Scalise F, Farina M, Manfredi M, Auguadro C,
Novelli E (2013) Assessment of jugular endovascular
malformations in chronic cerebrospinal venous insuf-
ficiency: colour-doppler scanning and catheter venog-
raphy compared with intravascular ultrasound.
Phlebology 28(8):409–417
127. Sclafani SJ (2012) Intravascular ultrasound in the
diagnosis and treatment of chronic cerebrospinal
venous insufficiency. Tech Vasc Interv Radiol 15(2):
131–143
128. Baracchini C, Perini P, Calabrese M, Causin F,
Rinaldi F, Gallo P (2011) No evidence of chronic
cerebrospinal venous insufficiency at multiple sclero-
sis onset. Ann Neurol 69(1):90–99
129. Baracchini C, Perini P, Causin F, Calabrese M,
Rinaldi F, Gallo P (2011) Progressive multiple scle-
rosis is not associated with chronic cerebrospinal
venous insufficiency. Neurology 77(9):844–850
130. Simka M, Ludyga T, Latacz P, Kazibudzki M (2013)
Diagnostic accuracy of current sonographic criteria
for the detection of outflow abnormalities in the inter-
nal jugular veins. Phlebology 28(6):285–292
 Part VIII
Pediatrics
 Pediatric Vascular Malformations
41
Sachin K. Pandey and Darren B. Orbach

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 984 The appropriate classification and understand-
ing of the pathophysiology of vascular
Imaging Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 984
malformations, both within and outside the
Extracranial Vascular Malformations . . . . . . . . . . . 985 central nervous system, is of the utmost impor-
Venous Malformations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 985 tance in the diagnosis and treatment of these
Capillary Malformations . . . . . . . . . . . . . . . . . . . . . . . . . . . 988
Lymphatic Malformations . . . . . . . . . . . . . . . . . . . . . . . . . . 988 entities. This is critical, both in determining the
Sinus Pericranii . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 994 appropriate types of treatment for given lesions
CNS Vascular Malformations . . . . . . . . . . . . . . . . . . . . 994
and in determining which lesions may require
Cavernous Malformation . . . . . . . . . . . . . . . . . . . . . . . . . . . 994 no treatment at all. Similarly, understanding of
Capillary Telangiectasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . 996 the pathology of a given entity informs the
Arteriovenous Malformations . . . . . . . . . . . . . . . . . . . . . . 997 goals of therapy. As highlighted in this chapter,
Vein of Galen Aneurysmal
Malformation (VGAM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 999
many of these malformations may have pro-
Spinal Vascular Malformations . . . . . . . . . . . . . . . . . . . . . 1003 tean clinical and radiologic presentations. As
such, a multidisciplinary approach to the diag-
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1005
nosis and management of vascular
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1005 malformations is important. Multidisciplinary
medicine is at its core, reliant on precise and
clear communication, and thus precise utiliza-
tion of terminology of greater importance. The
multidisciplinary approach to management, as
well as the rarity of many of these
malformations and operator-dependent risk of
therapy, makes most vascular malformations
ideal candidates for therapy at specialized,
large-volume institutions.
S.K. Pandey (*)
Neurointerventional Radiology, Brigham and Women’s Keywords
Hospital, Boston, MA, USA
Venous malformations • Sclerotherapy • Cap-
e-mail: skpandey@partners.org
illary malformations • Lymphatic
D.B. Orbach
malformations • AVMs • Embolization •
Neurointerventional Radiology, Boston Children’s
Hospital, Harvard Medical School, Boston, MA, USA Sinus pericranii • Cavernous malformations •
e-mail: darren.orbach@childrens.harvard.edu Vein of Galen aneurysmal malformations
# Springer Science+Business Media New York 2016 983
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_31
984 S.K. Pandey and D.B. Orbach

Introduction Table 1 ISSVA classification scheme

Vascular malformations Vascular tumors
The classification of vascular anomalies has tradi- Low-flow malformations Benign
tionally been a source of confusion, in part, due to Lymphatic malformations Infantile
such factors as the rarity of many of these anom- hemangioma
alies and overlapping imaging and clinical mani- Venous malformations Congenital
hemangioma
festations. Additionally, vascular anomalies are
Capillary malformations Tufted angiomas
diagnosed and treated from a multispecialty per-
Spindle-cell
spective, and as such, each group of physicians hemangioma
involved may have its own inherited terminology. Epithelioid
As an example, a recent study noted that 71.3 % of hemangioma
publications from 2009 used the term hemangi- Pyogenic granuloma
oma from that year did so erroneously. High-flow malformations Locally aggressive/
The Mulliken-Glowacki classification scheme borderline
was based on the biological behavior and endo- Arteriovenous Kaposiform
malformations hemangioendothelioma
thelial characteristics of these anomalies [1] and
Arteriovenous fistulae Retiform
was the basis from which the current International hemangioendothelioma
Society for the Study of Vascular Anomalies Papillary
(ISSVA) classification system grew (Table 1). intralymphatic
While a detailed discussion is beyond the scope angioendothelioma
of this chapter, the scheme separates vascular Composite
hemangioendothelioma
anomalies into neoplasms and malformations
Kaposi sarcoma
and with the former category includes infantile
Complex malformations Malignant
and congenital hemangiomas, Kaposiform Two or more vascular Angiosarcoma
hemangioendothelioma, and others. Vascular malformations within the Epithelioid
malformations are distinguished from vascular same lesion hemangioendothelioma
neoplasms by the absence of increased endothelial
cellular mitosis. As such, vascular malformations
tend to grow in proportion to overall childhood extra-CNS vascular malformations. But it is
growth, with some increased growth during apparent from any overview of pediatric vascular
puberty and pregnancy, or as a result of inflam- malformations that the bridges between the two
mation, infection, or hemorrhage. These areas are real and of critical importance, at the
malformations are now separated into three cate- levels of anatomy, genetics, and treatment.
gories: low-flow vascular malformations, high-
flow vascular malformations, and combined or
complex vascular malformations. Imaging Techniques
Although historically, clinicians and investiga-
tors who focused on vascular malformations For extracranial vascular malformations, the diag-
involving the central nervous system (CNS) nosis is often highly suggested by an informed
have worked in “splendid isolation” from those physical exam (Is the lesion warm and pulsatile?
involved with extra-CNS lesions, it has become Is there local cutaneous ischemia, ulceration, or
increasingly evident that complete separation of hemorrhage? Is the lesion purplish and expansile
these areas is artificial and counterproductive. As with recumbency or Valsalva?), and imaging gen-
there are diagnostic, natural history, and manage- erally plays a complementary role. Conversely,
ment issues unique to CNS lesions, this chapter for intracranial malformations, imaging is often
has sections dedicated separately to CNS and the alpha and the omega of diagnosis.
41 Pediatric Vascular Malformations
The imaging tools generally available for the
assessment of these entities are ultrasonography
(US), magnetic resonance imaging (MRI), mag-
netic resonance angiography and venography
(MRA/MRV), computed tomography (CT), CT
angiography (CTA), and digital subtraction angi-
ography (DSA) [2].
For extracranial malformations, US is frequently
the first-line imaging choice, due to lack of ionizing
radiation, wide availability, ability to obtain real-
time flow dynamic information, and lack of need for
sedation. Drawbacks to this modality include
operatory dependence and difficulty in imaging
intracranial intraosseous lesional extension. How-
ever, in addition to its intrinsic diagnostic utility, US
is extremely useful for image-guided therapy [3].
CT and MRI both provide excellent evaluation
of the extent of malformations and their relation-
ship to adjacent anatomic structures. Benefits of
CT include rapid time of acquisition, which can
limit the need for sedation. However, much of the
patient population with vascular malformations is
pediatric, and avoidance of unnecessary ionizing
radiation is vital. As such, every effort should be
made to utilize acceptable imaging alternatives,
and if CT is necessary, scans should be limited to
the area of interest only. On similar grounds,
multiphasic scanning should be kept to a mini-
mum. MRI and MRA provide excellent, poten-
tially time-resolved, tissue and spatial resolution
of vascular malformations, without the need for
ionizing radiation. The most significant drawback
of MR, particularly in the pediatric setting, is the
need for sedation in the majority of patients.
Because of its temporal resolution, unmatched
spatial resolution, and vessel selectivity, DSA has
a critical role in evaluating high-flow vascular
malformations, that is, those with an arterial com-
ponent. DSA allows for precise mapping of
involved anatomy and is an intrinsic part of
catheter-based endovascular therapy. DSA rarely
has a role in the workup or management of vas-
cular malformations that do not harbor an arterial
component and is rarely the first-line diagnostic
modality for any lesion, due to its invasive nature
and patients’ exposure to ionizing radiation.
985
Extracranial Vascular Malformations
Venous Malformations
Venous malformations occur with an overall inci-
dence of approximately 1 per 10,000. They are
generally soft, compressible lesions that are dark
in color. Recent years have seen identification of
genetic loci responsible for both inherited venous
malformations (via germ line mutation) [4] and
sporadic VM (via somatic mutation) [5]. Regard-
less of anatomic location, venous malformations
can cause discomfort, often related to swelling or
intralesional thrombosis. The classic presentation
on clinical exam is that of a fluctuant swelling
with dependent positioning or during crying or
Valsalva maneuver. Depending on the location
of the lesion, other symptoms, such as difficulty
feeding, exophthalmos, or visual blurring, may
also occur.
Various schemata have been proposed for the
classification of venous malformations: types
I–IV, based on the relationship of the lesion to
surrounding venous drainage and the morphology
of that drainage, and truncular versus extra-
truncular types, based on the embryological
stage at which the lesion was presumed to have
developed [6]. Although most venous
malformations are sporadic, inherited multiple
lesions can occur in the context of syndromes
such as blue rubber bleb nevus syndrome, mani-
fest as numerous head, neck, and extremity
venous malformations with the possibility of
new lesions potentially developing
throughout life.
On imaging, venous malformations are com-
pressible lesions of heterogeneous echotexture on
sonography (Fig. 1). Color Doppler interrogation
often reveals slow monophasic flow, if any. If
high-velocity or biphasic flow is encountered, an
alternative diagnosis of high-flow vascular mal-
formation should be entertained. CT can delineate
the extent of involvement of the surrounding tis-
sues and will also highlight phleboliths [3]. MRI
also offers excellent delineation of the extent of
involvement of these malformations. Though VM
986 S.K. Pandey and D.B. Orbach

Fig. 1 A 24-year-old
female visible left cheek
“birthmark,” progressively
expanding since age 8, with
recent skin discoloration.
Enlargement with Valsalva
maneuvers was noted on
examination. Pre- (a) and
post- (b) contrast T1W
images demonstrate
moderate, heterogeneous
enhancement and T2W (c)
image demonstrates a
hyperintense, multiseptated
abnormality. Grayscale
ultrasound (d) demonstrates
a hypoechoic
heterogeneous lesion.
Clinical and imaging
characteristics were
consistent with a venous
malformation

may have varied appearance on conventional
T1-WI and T2-WI images due to the presence or
absence of associated thrombus, they are gener-
ally hyperintense on T2-WI imaging and interme-
diate on T1-WI imaging. VM may demonstrate
variable contrast enhancement (although nearly
all do clearly enhance on delayed imaging) and a
lobulated contour. Contrast enhancement is
important in differentiating VM from
non-enhancing vascular malformations, such as
macrocystic lymphatic malformations. Internal
signal voids on MRI may represent phleboliths
within the VM. Catheter-based angiography is
not helpful in the workup of VM and has almost
no diagnostic role. In cases where angiography
has been done, delayed passage of contrast
through abnormal veins and sinusoidal spaces
has been described, though angiographic appear-
ance is more commonly normal [7].
Therapy for venous malformations had tradi-
tionally been surgical [8], though sclerotherapy is
the current treatment of choice. Sclerotherapy is
generally painful and, as such, is most often
performed under general anesthesia. Though easily
manageable in most locations, post-sclerosis edema
must always be considered in approaching these
lesions. For example, orbital venous malformations
leave little room for post-therapeutic swelling, and
sclerotherapy in this location should be undertaken
with the ophthalmology team available for urgent
decompressive lateral canthotomy, if necessary;
other options include surgical resection with or
without associated endovascular liquid embolic
occlusion. If there is no vision compromise, obser-
vation of these lesions is a reasonable option as
well. Similarly, venous malformations of the head
and neck associated with the airway, particularly
those with baseline airway compromise and the
need for repetitive treatments, may necessitate tra-
cheostomy in conjunction with therapy. As men-
tioned below, bleomycin is associated with less
post-procedural swelling than other sclerosants
and can be an excellent choice in locations such as
the orbit and airway.
41 Pediatric Vascular Malformations
Fig. 2 Lateral projection
from a direct percutaneous
angiogram of the left cheek
venous malformation (a)
demonstrates a large,
disorganized cluster of
venous pouches, ultimately
communicating with the
normal venous drainage
system. Frontal spot
radiograph (b) during
sclerosis with 3 % sodium
tetradecyl sulfate (in a 2:1:1
mixture with lipiodol and
air) demonstrates sclerosant
within the malformation
Sclerotherapy is generally performed under
sonographic and/or fluoroscopic guidance
(Fig. 2) [9]. Results are best with localized venous
malformations, and for very diffuse lesions, ther-
apeutic goals generally focus around the most
symptomatic portions of the malformation. One
important consideration is the identification of
rapid initial venous drainage. In those cases, man-
ual compression of these draining veins for the
initial 2–3 min of sclerosis can be helpful. In some
cases, coil or liquid embolic occlusion of the
outlet from the malformation to the adjacent
draining veins can compartmentalize the lesion,
ensuring that a large amount of sclerosant does not
escape into the venous system. An additional
technique for such situations is the double needle
technique [10], whereby the malformation is
accessed with two needles, using one for
sclerosant injection and the other as a
low-pressure escape valve, to reduce the risk of
nontarget sclerosis.
The choice of sclerosing agent is dependent
both on availability in a given country and on
operator preference. Ethanol is the most potent
of the sclerosing agents but is likely associated
with the most serious potential morbidity; extrav-
asation can result in severe damage or necrosis of
adjacent soft tissue and nerve injury. Doses
exceeding 0.75 mL/kg invariably result in intoxi-
cation [11]. Though there is a risk of cardiovascu-
lar collapse with ethanol, this is rare and can be
avoided with careful, limited use, with the total
987
dose kept below 1 mL/kg (maximum of 60 mL)
and given gradually, in small aliquots. As head
and neck lesions are rarely voluminous enough to
require very large doses of sclerosant
(as compared with the sometimes massive anom-
alies seen in truncal or limb lesions), the cardiac
morbidity and pulmonary hypertension are rarely
seen in this setting. In addition, sclerosants includ-
ing alcohol cause hemolysis, which at a sizable
scale may lead to hemoglobinuria. This is typi-
cally managed by generous hydration, urinary
alkalinization, and close monitoring of urinary
output [9], so other than for treatment of the
smallest lesions, placement of a Foley catheter
for these procedures is of great utility.
Given the potential soft tissue and nerve-
related morbidity of ethanol sclerosis, many oper-
ators prefer other sclerosants, such as detergent
agents. The most commonly used such agent is
sodium tetradecyl sulfate (STS), with others in
this class including polydocanol, sodium
morrhuate, and ethanolamine. These are typically
mixed with contrast agents such as Ethiodol for
procedural visibility. Foaming of the sclerosant
plus Ethiodol plus air is reported to increase effi-
cacy, possibly due to increased surface area of the
microbubbles generated for endothelial contact,
as compared with a pure liquid [12].
Bleomycin, an antibiotic and antitumor agent
originally derived from Streptomyces verticillus in
1966, is also used as a sclerosing agent for venous
malformations. Given that bleomycin induces
988
minimal posttreatment edema in comparison with
other sclerosing agents, it can be of particular
utility for lesions involving the airway. While
bleomycin is associated with possible pulmonary
fibrosis, this morbidity is seen at much larger
systemic doses than are used with percutaneous
sclerosis, with a threshold of >400 units. By
comparison, the doses used for typical pediatric
interventions are 0.5–1 mg/kg, and cases of
bleomycin-associated pulmonary fibrosis have
not been reported in the context of vascular anom-
aly sclerotherapy. Despite this, many groups
obtain baseline pulmonary function tests and
chest radiographs in patients undergoing
bleomycin sclerotherapy and then continue to
monitor pulmonary function after treatment.
Although not helpful as a primary treatment
agent, as mentioned above, coils can be a useful
adjunct, in particular, for malformations with
rapid communication to adjacent veins. Liquid
embolic agents, such as acrylate glue (nBCA) or
ethylene vinyl copolymer (Onyx), although
designed for the treatment of high-flow vascular
malformations, can also be used to isolate venous
malformations from adjacent veins into which
they drain, allowing for safe and effective
sclerotherapy.
Capillary Malformations
Capillary malformations are vascular lesions
involving the superficial dermal layers and occur
in 3 of every 1,000 newborns, darkening with age.
Histopathologically, capillary malformations rep-
resent ectatic blood vessels in relatively dener-
vated dermis. The most important feature of
these lesions, aside from cosmesis, is their associ-
ation with vascular syndromes such as Sturge-
Weber syndrome, Klippel-Trenaunay syndrome,
and Parkes Weber syndrome. In addition, it is
commonly the case that capillary malformations
involve the skin over an underlying fast-flow
lesion, such as AVM, and at times, may involve
the skin over lymphatic or venous malformations
as well.
In the case of Sturge-Weber, which involves
ocular and leptomeningeal abnormalities
S.K. Pandey and D.B. Orbach
ipsilateral to the capillary malformation, the skin
lesion itself generally conforms to a distribution
tracking the trigeminal nerve subsegments. Up to
25 % of all patients with a V1 distribution capil-
lary malformation have additional signs of Sturge-
Weber syndrome [13], while the association is
weaker with V2 and V3 distribution lesions. Neu-
rological involvement in Sturge-Weber syndrome
typically results in seizures, hemiplegia, and
developmental delay.
The imaging evaluation of capillary
malformations is primarily focused on assessment
for associated pathology in underlying tissue.
Doppler ultrasound is of particular utility in
confirming or excluding associated arteriovenous
shunting. Treatment of capillary malformations,
aimed at improving cosmesis, is generally by way
of pulsed-dye laser. Although only 15–20 % of
lesions clear completely. For capillary
malformations with underlying AVM or other
vascular anomalies, laser therapy will not have
real utility, and repeated attempts at treatment
most often lead to skin injury with little benefit.
Lymphatic Malformations
Lymphatic malformations affect between 0.02 %
and 0.05 % of the population, with half identified
at birth and nearly all by 2 years of age [14]; they
occur with equal incidence among females and
males. The pathogenesis of these malformations
is likely related to a defect occurring during
embryonic lymphangiogenesis, whereby buds of
lymphatic tissue become sequestered during
development. Lymphatic malformations are most
often cutaneous in location, though 10 % may
involve the viscera. However, they are particu-
larly trans-spatial in nature, frequently involving
all types of soft tissue and bone within a region.
Lymphatic endothelium may be confirmed patho-
logically with staining for the D2-40 antibody.
Lymphatic malformations (LMs) can be clas-
sified as microcystic, macrocystic, or mixed.
Many authors suggest that cysts <2 cm be classi-
fied as microcysts and those >2 cm as macrocysts,
while others advocate for classification based on
typical appearance: those cysts appearing
41 Pediatric Vascular Malformations
Fig. 3 A 3-year-old female with exodeviation and ptosis,
leading to discovery of an orbital macrocystic lymphatic
malformation. Axial pre-contrast T1W image (a) and cor-
onal T2W image (b) demonstrate a multiseptated,
sonographically anechoic, with clearly delineated
spaces and thin echogenic walls, are categorized
as macrocystic, while those appearing as a near-
solid echogenic mass are classified as microcystic
[9]. As the classification system has most rele-
vance in guiding treatment, this latter schema
has utility in its favor.
Macrocystic lymphatic malformations are most
commonly found in the axilla, chest, and neck and
can frequently be identified on prenatal ultrasound.
In that setting, important syndromic associations to
consider include Down, Turner, or Noonan syn-
drome. On examination, macrocystic
malformations are typically soft and compressible,
although striking changes in size and firmness can
occur with hemorrhage or infection. Mixed
microcystic and macrocystic lymphatic
malformations are common in the head and neck
and can result in facial deformities; when involving
the tongue or pharynx, airway compromise may
occur, particularly in the setting of hemorrhage or
infection. Syndromic associations have also
recently been described with the CLAPO syn-
drome, with the initial description consisting of six
patients with lower lip capillary malformation, face
and neck lymphatic malformation, facial and limb
asymmetry, and partial or generalized overgrowth.
Interestingly, periorbital lymphatic malformations
have been associated with intracranial vascular
anomalies, including developmental venous anom-
alies, cavernous malformations, and others [15].
989
heterogeneous left orbital malformation. Power Doppler
ultrasound image (c) confirms a multiseptated cystic
lesion, with no evidence of internal hypervascularity on
power Doppler interrogation
Imaging of lymphatic malformations is often
focused on delineation of infiltrative components
or to help refine differential possibilities when
with dealing indeterminate cystic lesions. The
specific appearance is dependent upon the
macrocystic, microcystic, or mixed nature of the
lesion. Macrocystic lesions have a multilocular
cystic sonographic appearance without evidence
of internal flow (Fig. 3). Layering debris from
prior hemorrhage or infection may also be seen.
Conversely, microcystic lymphatic malformations
typically appear as an echogenic mass due to the
density of acoustic interfaces within the malfor-
mation. By CT, macrocystic LMs typically appear
as hypodense, septated masses [3]. MR imaging
generally reveals hypointense T1-WI signal and
hyperintense T2-WI signal with or without fluid-
fluid levels. Post-contrast imaging may demon-
strate mural or septal enhancement, but no
enhancement is seen within cystic spaces of the
malformation.
Therapy of LM is guided by the size, area of
involvement, and symptoms related to the malfor-
mation. Treatment modalities for macrocystic
malformations include sclerotherapy, in some
cases laser therapy, and surgical excision
(Fig. 4). Unfortunately, recurrence is not rare
after resection, and scarring and damage to adja-
cent nerves or blood vessels may occur. Sclero-
therapy of LM typically involves cannulation of
each cyst, in which fluid is then aspirated and a
990 S.K. Pandey and D.B. Orbach

Fig. 4 A 3-year-old female

with left orbital macrocystic
lymphatic malformation.
Grayscale sonographic
image (a) demonstrates
percutaneous needle access
of the malformation, with
confirmation of
intralesional position noted
on spot lateral radiography
(b). Axial T2W images
prior to (c) and following
(d) sclerotherapy with
doxycycline 10 mg/mL
diluted in iodinated contrast
demonstrate interval
decrease in size of the
malformation (white
arrows)

similar volume of sclerosing agent is injected.
Some lesions may be so large as to necessitate
placement of a pigtail catheter which can remain
in place for several days, facilitating repeat aspi-
ration and injection.
Sclerosing agents used for lymphatic
malformations include ethanol, doxycycline,
sodium tetradecyl sulfate, bleomycin, alcoholic
solution of zein (Ethibloc), and OK-432
(Picibanil) [9]. Doxycycline is the most com-
monly used of these agents, usually at a concen-
tration of 10 mg/mL, due to its low morbidity and
high efficacy; it is, however, painful on injection,
and requires general anesthesia, even in adults. In
neonates, doses greater than 150 mg in a single
treatment setting are avoided, due to risks of met-
abolic acidosis, hypoglycemia, and hemolytic
anemia. OK-432 (Picibanil), derived from a
low-virulence strain of Streptococcus pyogenes,
is not currently approved for use in the United
States. A large prospective randomized trial of
OK-432 demonstrated that ~94 % of patients
with macrocystic lymphatic malformations had
regression or resolution of their malformations
[16]. Alcoholic solution of zein is a combination
of corn protein (zein), oleum in ethanol, and
sodium diatrizoate, also not available in the
United States but used in Canada and Europe,
with outcomes comparable to other agents [17].
In the natural history of LM, superinfection is
frequent. As such, many practitioners give pro-
phylactic antimicrobial therapy during and fol-
lowing sclerotherapy, with some extending this
for 7–10 days following treatment. Involution of
cysts is typically assessed at 4–6 weeks post-
procedure, with staged treatments generally
occurring at 6–8-week intervals.
Microcystic malformations are technically
more difficult to treat, tend to have poorer
response to therapy, and tend to recur after treat-
ment more frequently than macrocystic LM
[18]. Direct intralesional injection with
41 Pediatric Vascular Malformations
Table 2 Syndromic associations with high-flow vascular
malformations
PHACES
Lumbar
Parkes Weber
CM-AVM
Cobb
Osler-Weber-Rendu
Wyburn-Mason
PTEN hamartoma tumor syndrome
pingyangmycin (a bleomycin variant), as well as
radiofrequency ablation [19], has been described
as showing efficacy.
Extracranial AVM
In comparison to low-flow vascular
malformations, extracranial high-flow vascular
malformations are rare, with a slightly higher
prevalence in females. As is true for most vascular
malformations, AVMs generally grow in propor-
tion to the child, but they have been shown to
grow rapidly during puberty, pregnancy, infec-
tion, and trauma [20]. Some extracranial AVMs
have been associated with mutations of the
RASA1 gene [21]. RASA1 encodes for a p120
Ras GTPase-activating protein which itself is
involved in angiogenesis and cell adhesion, and
children harboring mutations present with a com-
bination of skin capillary stains and a propensity
to develop AVMs throughout the body
(CM-AVM). The PTEN gene on chromosome
10q has also been associated with AVMs. This
gene encodes a tumor suppressor protein involved
in regulating angiogenesis, cell growth, and cell
proliferation, and mutations are seen in multiple
eponymous syndromes, such as Bannayan-Riley-
Ruvalcaba and Cowden syndrome (Table 2). The
preferred designation for syndromes resulting
from such PTEN mutations is PTEN hamartoma
tumor syndrome (PHTS). Interestingly, greater
than half of these patients have high-flow vascular
anomalies of the head and neck. However, these
are unlike typical AVMs, in that they often have
ectopic fat within or adjacent to the malformation,
disproportionate dilation of draining veins, and
multiple non-contiguous sites of involvement
991
[22]. PTEN-associated arteriovenous lesions
tend to respond poorly, if at all, to embolization.
In the head and neck, AVMs most frequently
manifest with symptoms of hemorrhage, pain, or
deformity. Particularly because total cure is so
elusive and can only be entertained for the
smallest of lesions, where complete resection
might be possible, therapeutic goals are typically
focused on symptomatic control, on preservation
of critical functions such as vision, and on
cosmesis. There is no universal treatment algo-
rithm, with different teams not infrequently opting
for different embolic agents, frequency of treat-
ment, modality of treatment, etc. In those few
cases where complete cure is the goal of therapy,
the best results seem to occur with preoperative
embolization followed by surgical resection.
The choice of embolic agent is dependent on
the overall context of therapy for a given patient.
Specifically, preoperative embolization can be
effectively performed with short-term embolic
agents such as PVA particles, Gelfoam, or
Embospheres. While these provide excellent
devascularization for the purposes of reducing
intraoperative blood loss, these are not durable
embolic agents, and thus, if the goal of treatment
is permanent (i.e., anything other than preopera-
tive embolization), liquid embolic agents, such as
ethylene vinyl alcohol copolymer (Onyx), etha-
nol, and n-butyl cyanoacrylate (nBCA) glue, are
preferred. nBCA is a clear, liquid, monomeric
substance which is frequently mixed with tanta-
lum powder and Ethiodol for radiopacity. Expo-
sure to anions in blood causes the glue to
polymerize and form a cast that is adherent to
the endothelium, with the rate of polymerization
of the glue controllable by the glue-to-Ethiodol
ratio. In addition to the acute inflammatory fibro-
sis caused by the glue, there is a delayed foreign-
body giant cell granulomatous reaction
[23]. Onyx is a liquid embolic agent that precipi-
tates at its surface in contact with blood, while
maintaining a liquid core. Thus embolization
technique consists of the formation of a proximal
plug around the microcatheter tip, with further
infusion allowing for controlled delivery of the
liquid embolic, with longer and more extensive
embolization than is possible with nBCA. Onyx
992 S.K. Pandey and D.B. Orbach

Fig. 5 A 21-year-old male

with pulsatile tinnitus and
progressive enlargement of
a soft left auricular “mass.”
Axial fat-saturated proton
density (a) and T1W post-
contrast (b) MR images
demonstrate asymmetric
enlargement of the left
external ear with multiple
small flow voids. AP (c) and
lateral (d) projections from
a left external carotid
angiogram demonstrate
abnormal arteriovenous
shunting, consistent with a
left auricular arteriovenous
malformation

additionally causes less endothelial damage and
incites less inflammatory response than nBCA,
resulting in improved handling of embolized
malformations during surgical resection. Some
advocate ethanol, which results in endothelial
denudation, as the embolic treatment of choice
[24], though with potentially the greatest
morbidity risk.
The sonographic appearance of arteriovenous
malformations is characterized by multiple
hypoechoic vascular channels, without a well-
defined soft tissue mass. Doppler analysis reveals
arterialized draining venous waveforms [3]. CT
angiographic imaging demonstrates increased cal-
iber of feeding arteries and enlarged draining
veins. Similar findings are also noted on MRI,
where the rapid flow may also manifest as abnor-
mal flow voids on spin-echo sequences. On all
imaging modalities, the absence of an enhancing
soft tissue mass is critical in distinguishing these
lesions from others. DSA is virtually never needed
to establish the diagnosis. Dedicated diagnostic
catheter angiography is the routine standard of
care for intracranial lesions, where multiphasic
evaluation of flow dynamics to clearly identify
arteriovenous shunting, precise delineation of
arterial inflow, nidal vessels, and venous drainage
and superselective angiography for identification
of normal parenchymal branches arising from
feeding pedicles is critical [25]. In contradistinc-
tion, for extracranial AVM, diagnostic angiogra-
phy can typically be performed as part of the first
session of embolization.
In terms of location within the head and neck,
midface lesions, supplied by the internal maxil-
lary artery, are disproportionately represented.
The cheek and ear are relatively common sites,
with the latter generally confined to the outer ear
and possible extra-auricular spread; middle and
inner ear involvement virtually never occurs
(Fig. 5) [26]. Because arterial supply to ear
AVMs is almost always from end arteries,
41 Pediatric Vascular Malformations
Fig. 6 A 16-year-old male with purple left perimandibular
discoloration and growth, with complaint of aching left
mandibular pain with prolonged mastication. Coronal CT
image (a) displayed in bone windows demonstrates abnor-
mal irregular lytic lesion in the left body of the mandible
with cortical disruption. Lateral projection from left exter-
nal carotid angiography (b) demonstrates arteriovenous
complete endovascular embolization may
unavoidably induce tissue necrosis and lead to
amputation. As such, ear AVMs are generally
observed until symptoms necessitate intervention.
Although amputation and plastic surgical recon-
struction may be the likely end point, staged pal-
liative embolizations can delay amputation for
several years.
Mandibular and maxillary arteriovenous
malformations pose unique risks for life-
threatening hemorrhage associated with dental
eruptions or dental extractions. Arterial supply to
maxillary malformations is usually from internal
maxillary branches, though ascending pharyngeal
branches, labial branches of the facial artery, and
993
shunting with a primarily intraosseous left mandibular
arteriovenous malformation. Repeat lateral projection
image from left external carotid angiography (c) and fron-
tal spot radiograph (d) obtained following Onyx liquid
embolization demonstrate interval occlusion of the major-
ity of the malformation
posterior ethmoidal artery branches may also con-
tribute (Fig. 6). Given the intraosseous nature of
these malformations, surgical resection is chal-
lenging and necessitates bony excision and recon-
struction with graft. While some authors advocate
for a combined embolization and surgical
approach from the outset for these lesions, others
have reported that most can be controlled with
embolization alone [27]. High-volume vascular
anomaly centers often use embolization as first-
line therapy for newly discovered maxillary or
mandibular AVM and to approach any loose
tooth or required dental extraction with a further
preparatory embolization, with extraction in the
angiography suite.
994
Fig. 7 A 10-year-old boy with enlarging scalp mass, with
expansion during Valsalva. Axial T2W MRI (a) and sagit-
tal CTA (b) demonstrate a focal frontal calvarial defect
with subcutaneous T2 hyperintense, contrast enhancing
malformation appearing to communicate with the
Sinus Pericranii
Sinus pericranii is a link between extracranial and
intracranial vascular anomalies. Patients with this
condition have a transosseous venous bridge
allowing for direct flow between intracranial and
extracranial venous drainage pathways (Fig. 7).
The intracranial contributor flowing into the
transosseous bridge may be a venous sinus (with
the superior sagittal sinus being by far the most
common) or a pial vein or developmental venous
anomaly. Sinus pericranii is rare overall, most
commonly involves small venous channels, and
is infrequently symptomatic, with incidental dis-
covery on imaging being the rule. When the con-
dition does present clinically, the most common
manifestation is as a nonpulsatile mass that
expands with recumbency, crying, or Valsalva.
The extracranial, scalp portion of the sinus
pericranii may have the mass-like, globular con-
figuration of a venous malformation, rather than
the morphology of a tubular venous channel.
These cases are clinically apparent from birth
and are most commonly located frontally, running
from the glabella up toward the hairline,
manifesting as a large, purplish, expansile lesion.
Though in a minority of cases, sinus pericranii
may occur secondarily as a result of trauma, the
vast majority of cases are congenital, and they are
frequently associated with intracranial develop-
mental venous anomalies [28].
S.K. Pandey and D.B. Orbach
subjacent superior sagittal sinus. Lateral projection from
a right internal carotid artery angiogram in the venous
phase (c) confirms sinus pericranii with demonstration of
communication between the palpable malformation and
the underlying sinus
Diagnosis of sinus pericranii can usually be
made on the basis of clinical exam. Confirmation
of the presence of transosseous channels may be
obtained with use of sonography. MR imaging,
particularly sagittal and coronal high-resolution
contrast-enhanced sequences with fat suppres-
sion, is helpful in clearly localizing the bridging
channels and also in providing a more global view
of the overall intracranial venous drainage.
Symptomatic cases of sinus pericranii can be
treated, either with sclerotherapy (taking care that
the sclerosant does not travel intracranially) or
resection. For cases where pial veins or a DVA
serves as the intracranial contributor to the bridg-
ing vessel (rather than a venous sinus), catheter
angiography is necessary in order to ensure that
alternative drainage is available. Cases where the
sinus pericranii serves as an obligate venous
drainage pathway for a brain region cannot be
safely treated without incurring a high risk of
venous infarct.
CNS Vascular Malformations
Cavernous Malformation
Cavernous malformations (CMs) are well-
circumscribed, sinusoidal endothelium-lined
channels. As is true of other vascular
malformations, they are frequently referred to by
41 Pediatric Vascular Malformations
Fig. 8 A 14-year-old male with seizure disorder and left
parietal spikes on EEG. Axial non-contrast CT scan of the
head (a), axial T2W MRI (b), and axial minimum intensity
projection susceptibility-weighted image (c) all
other names such as cavernomas, cavernous hem-
angiomas, and cavernous angiomas. The vascular
spaces in cavernous malformations are immature,
lacking elastin and having minimal smooth mus-
cle [29], with a hallmark being absence of inter-
vening brain parenchyma. CMs are common,
occurring in just under 1 % of the general popula-
tion. The distribution of these malformations follows
the volume of the neuraxis, with 80–90 % being
supratentorial, 8–20 % being infratentorial, and
approximately 3 % being spinal. The most common
clinical presentation of these malformations, in 36 %
of cases, is seizures. Twenty-five percent of patients
present with hemorrhage, and the balance is discov-
ered incidentally on imaging.
While some CMs are congenital or present in
early childhood, de novo cavernous
malformations have been detected following radi-
ation therapy, prior brain biopsy, post-viral infec-
tions, and in certain familial cases. Cavernous
malformations frequently coincide with develop-
mental venous anomalies (DVAs), with the latter
affecting approximately 4 % of the general popu-
lation (Fig. 8). DVAs, as originally described by
Lasjaunias and colleagues [30], represent variants
of normal venous drainage, rather than true
malformations. However, a causal relationship
between DVAs and cavernous malformations has
been proposed, whereby DVAs with collector
995
demonstrate a left parietal cavernous malformation (white
arrows). Also noted is an associated development venous
anomaly (white arrowhead)
vein stenosis develop increased capillary bed
pressure that in turn leads to microhemorrhage.
This hemorrhage may induce recruitment of fibro-
blasts which lay down fragile endothelialized ves-
sels. Although speculative and although clear
venous stenosis is not seen in most cases of
DVA with CM, support for this mechanism can
be derived from the observation of endothelial
development at the site of chronic intracranial
hematomas [31].
The natural history of CM has been controver-
sial. Some of the earlier literature, written when
many believed that all cavernous malformations
were congenital, considered the at-risk period to
extend all the way from birth, thus mistakenly
lowering the calculated annual risk of hemor-
rhage. Further adding to confusion, some patients
may experience clinical deterioration in the
absence of imaging-documented hemorrhage or
other change. Meta-analysis data from prospec-
tive databases suggests an overall rate of hemor-
rhage of 2.4 % per patient year. Interestingly, the
size of cavernous malformations appears to play
little or no role in risk of hemorrhage, while loca-
tion has been shown to be more strongly corre-
lated, with deep brain lesions and infratentorial
lesions having higher hemorrhagic risk
[29]. While some authors suggest that deeper elo-
quent lesions are more likely to manifest with
996
clinical symptoms secondary to hemorrhage than
superficial non-eloquent lesions, the reasons for
the positional dependence remain unclarified.
The imaging tool of choice for the diagnosis
and characterization of cavernous malformations
is MRI. On MRI, CMs appear as well-defined,
lobulated lesions with heterogeneous signal inten-
sity on conventional T1-WI and T2-WI images,
often with a “popcorn”-like appearance. This het-
erogeneity reflects mineralization, hemorrhage,
and fibrosis, all of which frequently coexist within
a cavernous malformation. Gradient echo or
susceptibility-weighted images demonstrate
rim-like magnetic susceptibility, reflecting hemo-
siderin deposition in the surrounding brain paren-
chyma. Zabramski and colleagues developed a
classification scheme for CM, based on the MRI
appearance. CT imaging may demonstrate focal
parenchymal hyperdensities, in keeping with min-
eralization or hemorrhage, though the appearance
is less specific than that of MRI. Traditionally,
DSA has no role in the evaluation of cavernous
malformations as they are angiographically
occult, but a recent report demonstrates visualiza-
tion of a CM using cone-beam CT with intra-
arterial angiographic injection, including delinea-
tion of an adjacent DVA, where none had been
seen on MRI [32]. Despite the well-characterized
appearance on MRI in most cases, a specific diag-
nosis is sometimes not possible, particularly in the
setting of acute hemorrhage, and in those situa-
tions, serial imaging to exclude other hemorrhagic
lesions is often helpful. Of particular note, early
postoperative imaging after resection of cavern-
ous malformations may be particularly challeng-
ing, as the appearance can be very similar to the
preoperative imaging despite total surgical resec-
tion [29]. As such, postoperative imaging should
be interpreted with caution and delayed imaging
may also be helpful.
Surgical resection has been the mainstay of
treatment for cavernous malformations. Given the
more benign course of superficial lesions, they are
frequently observed clinically and radiographi-
cally. However, if a particular lesion can be
shown to be an epileptogenic source, resection of
superficial lesions may be warranted. Deep brain
cavernous malformations, owing to their worse
S.K. Pandey and D.B. Orbach
prognosis, may merit more aggressive surgical
intervention. Conversely, resection of these deep
brain malformations carries increased risk of surgi-
cal morbidity. This combination of considerations
has led to interest in stereotactic radiosurgery for
deep brain cavernous malformations. However,
several studies have suggested that the high inci-
dence of hemorrhage- and radiation-related com-
plications, as much as sevenfold greater than with
arteriovenous malformation therapy, renders ste-
reotactic radiosurgery inappropriate for all but the
most exceptional cases [33] of CM.
Capillary Telangiectasia
Capillary telangiectasia represents a collection of
abnormally dilated capillaries. Similar to cavern-
ous malformations, they demonstrate paucity of
elastin and smooth muscle. Unlike cavernous
malformations, however, these capillaries are
interspersed with normal brain parenchyma.
Capillary telangiectasia occurs with an estimated
overall prevalence of 0.7 % and demonstrates
a predilection for pontine location, accounting
for approximately 78 % of all cases [34]. Less
common locations include the basal ganglia
and cerebral hemispheres. Clinically, the over-
whelming majority of these lesions are asymp-
tomatic and are predominantly incidental
imaging findings.
The imaging modality of choice for character-
ization of capillary telangiectasia is MRI. These
malformations most commonly demonstrate mild
enhancement on post-contrast MR imaging, with
gradient echo and susceptibility-weighted MRI
demonstrating abnormal magnetic susceptibility
(Fig. 9). Since telangiectasias are very rarely asso-
ciated with hemorrhage, the magnetic susceptibil-
ity is postulated to represent sluggish,
deoxygenated blood within the malformation
rather than hemorrhagic blood products. The asso-
ciation of these malformations with developmen-
tal venous anomalies is weaker than that observed
with cavernous malformations, with only 11 %
demonstrating an associated DVA [35]. Similar
to cavernous malformations, capillary telangiec-
tasia is angiographically occult.
41 Pediatric Vascular Malformations 997

Fig. 9 A 17-year-old male

with incidental finding of
pontine capillary
telangiectasia (white
arrows) on MRI done for
headaches. Axial T1W
images pre- (a) and post-
(b) contrast demonstrate
faint “brush-like”
enhancement. The
malformation is isointense
to pontine parenchyma on
axial T2W imaging (c) and
demonstrates classic
susceptibility artifact on
susceptibility-weighted
imaging (d)

These malformations generally require no
interventional therapy, reflecting their benign nat-
ural history, and the primary importance of imag-
ing for telangiectasia is specific diagnosis, so as to
avoid unnecessary surgical intervention. Given
the incidental nature of their discovery in most
cases and the benign natural history, routine
follow-up is generally not indicated [34].
Arteriovenous Malformations
As their name denotes, arteriovenous
malformations (AVMs) are characterized by direct
connections between arteries and veins,
circumventing the capillary bed (Fig. 10).
AVMs, in the great majority of cases, are thought
to be congenital, related to errors in vascular mor-
phogenesis. The abnormal arteriovenous shunting
that is the hallmark of these malformations leads
to the subsequent pathology: the draining veins,
being subjected to abnormally pressurized flow,
may rupture. Additionally, the feeding arteries,
subject to fast flow into a low-pressure outlet,
may develop flow-related aneurysms, also subject
to hemorrhage (Fig. 11). Additionally, the sink
effect of the malformation may cause a steal phe-
nomenon resulting in ischemia to surrounding
brain parenchyma.
Specific features of brain AVMs associated
with hemorrhagic presentation have been studied
in the pediatric population and include smaller
nidal size, exclusively deep venous drainage,
and infratentorial location [36]. Although tradi-
tional estimates place the annual risk of hemor-
rhage at between 2 % and 4 % per year,
historically, the cases that have come to clinical
attention are those that presented with hemor-
rhage, likely artificially skewing the annualized
risk upwards. In an attempt to delineate the natural
history, the recent ARUBA study was conducted,
where patients with unruptured brain AVMs were
randomized to treatment versus observation.
Although the study was halted because of an
998 S.K. Pandey and D.B. Orbach

Fig. 10 A 21-year-old female with headaches. AP (a) and malformation. The nidus (black arrows) receives left supe-
lateral (b) projections from a selective left vertebral angio- rior cerebellar and anterior-inferior cerebellar arterial sup-
gram, as well as 3-D volume-rendered reconstruction (c) of ply. There are two feeding artery aneurysms (arrowheads)
the same, demonstrate a left cerebellar arteriovenous associated with the superior cerebellar arterial pedicle

Fig. 11 Lateral projection

from a post-embolization
selective left vertebral
artery angiogram (a) and
spot lateral projection skull
radiograph post-
embolization (b). Onyx-18
liquid embolic material is
visualized (black arrow)
within the superior
cerebellar arterial pedicle,
with occlusion of the two
feeding artery aneurysms
seen preoperatively

improved outcome in the observational arm [37],
many observers criticized the study design.
Unlike aneurysmal subarachnoid hemorrhage,
intracranial hemorrhage from arteriovenous
malformations is less severe, though still with a
10 % risk of death.
Treatment of brain AVMs is focused on oblit-
eration of the nidus. This may accomplished with
radiotherapy, surgically or endovascularly.
Regardless of whether the approach is surgical
or endovascular, general principles of arteriove-
nous malformation management include avoid-
ance of proximal feeding artery occlusion, as
doing so allows the nidus to continue to recruit
arterial supply, while also blocking access routes
for additional endovascular therapy. Similarly,
subtotal surgical resection also frequently leads
to progression. Surgical risk has been linked to
the Spetzler-Martin grading system for adults
(Table 3). Though this has some relevance for
the pediatric population, some factors within the
system, such as determining cerebral eloquence
and the focality of venous drainage within a
restricted brain volume, likely differ in children
and adults [38].
Stereotactic radiotherapy is also a commonly
used therapeutic modality for brain AVMs, partic-
ularly for cases measuring less than 4 cm in size.
41 Pediatric Vascular Malformations
Table 3 Spetzler-Martin classification scheme for intra-
cranial arteriovenous malformations
Size <3 cm – 1 point
3 cm–6 cm – 2 points
>6 cm – 3 points
Venous drainage Superficial – 0 points
Deep – 1 point
Eloquence Non-eloquent – 0 points
Eloquent – 1 point
Superficial venous drainage includes cortical veins and
convexity sinuses. Eloquent locations include sensorimo-
tor cortex, language areas, visual cortex, hypothalamus,
internal capsule, brainstem, cerebellar peduncles, or deep
cerebellar nuclei
Table 4 Modified Pollock-Flickinger score for predicting
arteriovenous malformation obliteration with radiosurgery
Modified Pollock-Flickinger score
AVM score = 0.1 (volume, mL) + 0.02 (age, year) + 0.5
(location)
Location score = 0 for hemispheres/corpus callosum/cere-
bellum. Location score = 1 for basal ganglia/thalamus/
brainstem
The Spetzler-Martin scale is of less utility in
predicting outcomes for radiotherapy than it is
for surgical prognostication. As such, the
Pollock-Flickinger score was derived and has
been well correlated to outcomes in both adults
and children following radiotherapy (Table 4).
Drawbacks to radiosurgery include the delayed
nature of nidal occlusion, with a median time to
obliteration of approximately 3 years. Other con-
cerns include the risk of radiation necrosis involv-
ing adjacent brain parenchyma, often presenting
with significant delay, and increased risk of tumor
induction from the stochastic effects of radiation
exposure, particularly in children.
Syndromic associations with AVMs are myr-
iad, though rare. Nonhereditary segmental abnor-
malities that involve AVMs are thought to relate to
the fact that neural crest and adjacent mesoderm
originate from a shared metamere. Thus, in situa-
tions of developmental disruption prior to daugh-
ter cell migration, malformations may be
segmental in distribution. This group of
cerebrofacial arteriovenous metameric syndromes
(CAMS, as named by Lasjaunias and colleagues
999
[39]) includes Wyburn-Mason syndrome, or
Bonnet-Dechaume-Blanc syndrome, in which
orbital, periorbital, and retinal extracranial
AVMs track backwards along the optic pathways,
following the optic nerve to the chiasm, and back
to the lateral geniculate region of the
thalamus [40].
Hereditary syndromic associations with AVMs
have also been described. Genetic defects in
endoglin and ALK-1 binding proteins for
transforming growth factor-β result in endothelial
dysfunction and have been associated with AVMs
in Osler-Weber-Rendu syndrome (or hereditary
hemorrhagic telangiectasia) (Fig. 12). In addition
to the abovementioned association of mutations in
the RASA1 gene with extracranial CM-AVM,
RASA1 mutations have also been associated
with brain arteriovenous malformations and
fistulae [21].
Vein of Galen Aneurysmal
Malformation (VGAM)
VGAM represents specific choroidal types of
intracranial arteriovenous lesions, manifest in
young infants. Specifically, they are supplied by
choroidal arteries (anterior and posterior), with
venous drainage to the median prosencephalic
vein of Markowski, an embryonic precursor to
the vein of Galen [41], with no development of a
true deep venous system in most cases (Fig. 13)
[42]. Developmentally, the median prosence-
phalic vein is the primary venous drainage from
6 to 11 weeks of gestational age, after which the
median prosencephalic vein is partially assimi-
lated into the vein of Galen. However, when the
connections between choroidal arteries and the
median prosencephalic vein fail to regress by
approximately 8 weeks gestational age, progres-
sive enlargement of the vein may occur, leading to
VGAM. Several similar appearing entities are
often confused with VGAM, including tectal
AVMs which drain into the vein of Galen, and
vein of Galen varices. As the treatment approach
for these lesions is quite different than for VGAM,
a precise delineation of a given lesion’s anatomy
is critical.
1000 S.K. Pandey and D.B. Orbach

Fig. 12 A 24-year-old
female with seizures,
epistaxis, and remote
history of cerebral abscess.
AP projection (a) from a
right vertebral artery
angiogram in mid-arterial
phase demonstrates three
distinct arteriovenous
malformations (black
arrowheads) with
associated early venous
drainage. Lateral projection
(b) from a right internal
carotid artery angiogram
demonstrates an additional
right parietal arteriovenous
malformation (black
arrowhead). Oblique
sagittal MIP (c) and 3-D
volume-rendered
reconstruction (d) from a
subsequent CT angiogram
of the chest help to confirm
the diagnosis of Osler-
Weber-Rendu syndrome
with identification of
pulmonary arteriovenous
malformation (white
arrowhead)

Fig. 13 A 6-month-old male with abnormal prenatal vein of Galen aneurysmal malformation. There is notable
imaging. Sagittal post-contrast T1W MR image (a), axial dilation of the median prosencephalic vein of Markowski
T2W MR image (b), and lateral projection from a left (white arrows), without communication to the normal deep
internal carotid angiogram (c) all demonstrate abnormal venous drainage system
41 Pediatric Vascular Malformations
Fig. 14 Choroidal (a) and mural (b) architectural sub-
types of vein of Galen aneurysmal malformations. The
choroidal type demonstrates numerous choroidal arterial
feeding arteries converging on a complex fistulous site
(black arrowhead), which in turn drains to the dilated
Nidal architecture of a VGAM can generally be
divided into two primary subtypes: mural and
choroidal (Fig. 14). The mural configuration, gen-
erally better tolerated clinically, consists of direct
arteriovenous fistulas in the wall of the median
prosencephalic vein. Choroidal types generally
have more severe neonatal symptoms and involve
an interposed nidal network between the choroi-
dal arteries and the large venous pouch. The pres-
ence of a high-flow VGAM may interfere with
normal maturation of skull base venous sinuses
and the jugular bulbs, resulting in the persistence
of fetal occipital or marginal sinuses and incom-
plete maturation of the jugular bulbs, and venous
outlet stenosis may portend extremely poor prog-
nosis. In contrast, venous drainage may lead to the
cavernous sinuses and thereafter to the facial and
scalp veins, often predictive of good prognosis
(Fig. 15). In cases where outlet obstruction leads
to venous congestion, MRI may demonstrate
focal white matter lesions or diffuse destruction
of the brain (the so-called melting brain) [42].
Clinical presentation of VGAM generally falls
into two categories: those related to high-output
cardiac failure and neurological symptoms related
to venous congestion and hydrocephalus (Table 5)
[43]. As noted previously, the severity of these
1001
median prosencephalic vein of Markowski (black arrows).
The mural type demonstrates more direct abnormal drain-
age of feeding arteries to the lateral wall of the median
prosencephalic vein of Markowski, without intervening
complex fistulous bed
symptoms is generally related to the angioarch-
itecture of the VGAM. Cardiac manifestations
stem from abnormally elevated return to the
right side of the heart, leading to right atrioven-
tricular dilation and pulmonary hypertension
[44]. High-output left-sided heart failure is com-
monly present as well. Cardiac symptoms can
range from the very severe, including those
detected sonographically in utero and those
requiring intervention in the first few days of
life, to the milder, allowing for observation until
4–5 months of age. Neurological symptoms are
generally the result of venous congestion and
hydrocephalus. As focal neurological symptoms
are difficult to detect in newborns and young
infants, macrocrania may be an early presenting
symptom. Other neurological manifestations may
include neurocognitive delay, which, when
severe, may be clinically irreversible despite
embolization.
Endovascular embolization is the mainstay of
VGAM therapy (Fig. 16). The goals of VGAM
therapy are primarily to reduce or reverse cardiac
symptoms and to prevent neurological manifesta-
tions. As such, complete angiographic oblitera-
tion of the VGAM is not generally the end point
of therapy as this may involve unnecessary
1002 S.K. Pandey and D.B. Orbach

Fig. 15 Lateral projection from an early (a) and late (b)
venous phase of a left internal carotid angiogram in a child
with a vein of Galen aneurysmal malformation. The earlier
phase demonstrates a “pseudophlebitic” appearance to the
congested cortical veins. Note the absence of typical deep
venous drainage and extensive drainage to the cavernous
sinus anteriorly. The later phase demonstrates the thalamic
(black arrowhead) and lateral perimesencephalic (white
arrowhead) veins forming the common “epsilon” pattern
of venous drainage

Table 5 Bicetre neonatal evaluation score for vein of Galen aneurysmal malformations
Cardiac
Score Cerebral function function Respiratory function Hepatic function Renal function
0 Permanent Resistance to Assisted ventilation, with Abnormal Anuria
neurological medical therapy desaturation coagulation,
deficit transaminitis
1 Seizures Ventilation Assisted ventilation, with Moderate or Unstable
necessary normal saturation at FiO2 transient hepatic diuresis with
>25 % insufficiency medical therapy
2 Isolated Failure, Assisted ventilation, with Hepatomegaly Transient anuria
convulsion unstable with normal saturation at FiO2
medical therapy <25 %
3 Nonconvulsive, Failure, stable Tachypnea, unable to Normal Normal
episodic with medical finish a bottle
neurological signs therapy
4 Subclinical EEG Overload, with Tachypnea, able to finish N/A N/A
abnormalities no medical a bottle
therapy
5 Normal Normal Normal N/A N/A
Maximal score = 21

neurological risk. The trans-arterial route is pre-
ferred by most practitioners to transvenous embo-
lization, as it is associated with lower morbidity.
Early heart failure may compel embolization
within the first few days of life, with reduction of
the shunt by approximately 30 %, generally
allowing for clinical abatement of cardiopulmo-
nary symptoms for the first few months of life.
Despite impressive results in this population at
experienced centers [45], neonatal patients with
VGAM have the highest rates of mortality and
neurological morbidity. If early heart failure is
not present, treatment is typically deferred to ~5
months of age and is often staged over several
months or longer. Noncardiac clinical scenarios
necessitating early embolization include worsen-
ing ventriculomegaly or any neurodevelopmental
concerns.
41 Pediatric Vascular Malformations
Fig. 16 Axial T2W MR
images from a young child
with vein of Galen
aneurysmal malformation at
6 months (a) and 15 months
(b) of age. The dilation of
the median prosencephalic
vein of Markowski (white
arrow), the extent of
hydrocephalus, and the
resultant thinning of the
parieto-occipital lobes have
all markedly improved
following interval trans-
arterial Onyx and coil
embolization
Spinal Vascular Malformations
In an effort to understand and approach spinal
vascular malformations, numerous classification
schemes have been developed, particularly for the
high-flow malformations. Early schemes classi-
fied high-flow spinal vascular malformations
into four subtypes. Type I lesions are spinal
dural arteriovenous fistulae; type II lesions are
intramedullary “glomus” AVMs; type III lesions
are metameric AVMs, sometimes involving the
entire spinal canal and paraspinal structures at a
given level; and type IV lesions are perimedullary
arteriovenous fistulae. While this scheme has
gained widespread use, alternatives and revisions
have been proposed, such as a scheme that focuses
on the distinction between arteriovenous fistula
and malformation and then further classifying by
axial location into intramedullary,
extramedullary-intradural, extradural, or trans-
spatial [46].
Intramedullary spinal AVMs may be either
compact or diffuse. The former, referred to as
glomus AVMs, represent approximately 90 % of
the intramedullary cases. Overall, spinal AVMs
occur with approximately one-tenth the frequency
of their intracranial counterparts. They most com-
mon occur at the cervicomedullary junction and
are frequently diagnosed in childhood or young
adulthood (in contradistinction to spinal dural
arteriovenous fistulae, which are much more com-
monly diagnosed later in life). Arterial supply to
intramedullary lesions is typically from multiple
1003
anterior and posterior spinal arterial branches
(Fig. 17). Venous egress is usually diffuse, involv-
ing both dorsal and ventral surfaces of the spinal
cord and with drainage extending cranially and
caudally. As with cranial AVMs, features of
angioarchitectural weakness, such as feeding
artery aneurysms, may occur. Therapy, always
high risk for intramedullary lesions, is generally
focused on complete surgical excision, facilitated
in select cases by preoperative embolization.
Intramedullary lesions may also extend into
extramedullary tissues, corresponding to the type
III or metameric subtype of spinal vascular mal-
formation (also known as Cobb syndrome). These
malformations may involve multiple spinal levels
and extend axially to involve paraspinal tissues.
This can also result in secondary osseous spinal
changes, such as expansion of the spinal canal, as
well as laminar and pedicular erosion [47]. Type
III malformations most commonly manifest in
children as progressive myelopathy and pain.
Given the infiltrative nature of these
malformations, treatment can be particularly dif-
ficult and challenging and often focuses on
debulking for reduction of mass effect on the cord.
Intramedullary lesions usually present with
back or radicular pain, myelopathy, bowel or blad-
der dysfunction, and sensorimotor deterioration.
Symptom onset may be sudden, in cases related to
lesional hemorrhage. Unlike spinal dural arterio-
venous fistulae, this hemorrhage may be
intramedullary as well as in the subarachnoid
space. This latter point is of particular importance
1004
Fig. 17 A 16-year-old
male with worsening
bilateral lower extremity
weakness. Sagittal T2W
MR image (a) demonstrates
expansion of the
mid-thoracic spinal cord
with abnormal
intramedullary flow voids
and prominent
extramedullary flow voids.
AP projection from a left T7
artery angiogram (b)
demonstrates abnormal
arteriovenous shunting
related to an intramedullary
spinal arteriovenous
malformation
when considering that these malformations can
commonly occur at the cervicomedullary junc-
tion, and thus, in cases of intracranial subarach-
noid hemorrhage, with no identifiable intracranial
etiology, a spinal AVM must always be
considered.
The mainstay of the initial imaging evaluation
of spinal malformations is MRI. T2-WI imaging
often demonstrates spinal cord signal
hyperintensity adjacent to the malformation,
suggesting venous hypertension or developing
myelomalacia. Flow-related signal voids that are
part of the lesion itself and additionally dilated
flow voids along the surface of the cord,
representing supply and draining vessels, are fre-
quently present. However, digital subtraction
angiography remains the gold standard in pre-
cisely characterizing spinal vascular
malformations. Additionally, angiography allows
for characterization of normal vessels, such as the
artery of Adamkiewicz, that must be preserved in
the course of treatment (Fig. 18).
It has long been known that the natural history
of intramedullary spinal AVMs is poor, with more
than one-third of young patients developing
severe impairment after 3 years of evolution.
S.K. Pandey and D.B. Orbach
Lesions in the posterior aspect of the cord are
more amenable to safe resections, as are lesions
in the filum terminale [47]. Embolization can have
a role in select cases, either on its own or as a
preoperative measure. Often, the critical determi-
nant of the feasibility of safe embolization is the
anatomic relationship of the malformation to the
anterior spinal arterial axis.
High-flow spinal malformations without an
intervening nidus between arterial supply and
venous egress are categorized as spinal arteriove-
nous fistulae. As with AVMs, these are categorized
by anatomic location, namely, intradural or
extradural. The latter is rare and usually involves
radicular arterial branches communicating with the
epidural venous plexus [48], most commonly in the
cervical spine. If the pressurized epidural plexus
does not transmit arterial pressure in retrograde
fashion to the perimedullary veins of the cord,
neurological symptoms as a result of venous con-
gestion are unlikely. Rather, epidural fistulae usu-
ally become symptomatic as a result of local mass
effect from congested epidural venous plexus
resulting in spinal canal stenosis [49], although
associated perimedullary venous communication
and myelopathy has been reported [50].
41 Pediatric Vascular Malformations 1005

treatment of these entities. This is important both

in determining the appropriate types of treatment
for given lesions and in determining which lesions
may require no treatment at all. Similarly, under-
standing of the pathology of a given entity
informs the goals of therapy. As highlighted
above, many of these malformations may have
protean clinical and radiologic presentations. As
such, a multidisciplinary approach to the diagno-
sis and management of vascular malformations is
important. Multidisciplinary medicine is at its
core, reliant on precise and clear communication
and thus precise utilization of terminology of
greater importance. The multidisciplinary
approach to management, as well as the rarity of
many of these malformations and operator-
dependent risk of therapy, makes most vascular
malformations ideal candidates for therapy at spe-
cialized, large-volume institutions.
Fig. 18 AP image of a selective left T10 angiogram,
demonstrating supply to the artery radiculomedullaris
magna (artery of Adamkiewicz)
References
The more common type of arteriovenous fis-
tula is intradural, representing approximately 1. Mulliken JB, Glowacki J (1982) Hemangiomas and
vascular malformations in infants and children: a clas-
70 % of all spinal vascular malformations. These sification based on endothelial characteristics. Plast
develop in the lateral epidural space and the point Reconstr Surg 69(3):412–422
of fistulization in these cases is at the dural nerve 2. Kollipara R, Odhav A, Rentas KE, Rivard DC, Lowe
root sleeve, with communication between a radic- LH, Dinneen L (2013) Vascular anomalies in pediatric
patients: updated classification, imaging, and therapy.
ular artery and perimedullary veins. There is a Radiol Clin North Am 51(4):659–672
notable male predominance of these fistulae, as 3. Dubois J, Garel L (1999) Imaging and therapeutic
is the case for intracranial lateral epidural arterio- approach of hemangiomas and vascular malformations
venous fistulae. Symptoms related to these fistulae in the pediatric age group. Pediatr Radiol 29(12):
879–893
are generally related to venous hypertension and 4. Boon LM, Mulliken JB, Vikkula M, Watkins H,
vascular steal, with slow, progressive sensorimo- Seidman J, Olsen BR et al (1994) Assignment of a
tor myelopathy being the most common presenta- locus for dominantly inherited venous malformations
tion. This insidious nature may explain the to chromosome 9p. Hum Mol Genet 3(9):1583–1587
5. Uebelhoer M, Natynki M, Kangas J, Mendola A,
paucity of these entities in children, though a Nguyen HL, Soblet J et al (2013) Venous
likelier explanation is that these fistulae are malformation-causative TIE2 mutations mediate an
acquired later in life rather than congenital AKT-dependent decrease in PDGFB. Hum Mol Genet
malformation. 22(17):3438–3448
6. Lee BB, Baumgartner I, Berlien P, Bianchini G,
Burrows P, Gloviczki P et al (2014) Diagnosis and
treatment of venous malformations consensus docu-
Summary ment of the International Union of Phlebology (IUP):
updated 2013. Int Angiol: J Int Union Angiol
123:206–212
The appropriate classification and understanding 7. Konez O, Burrows PE (2002) Magnetic resonance of
of the pathophysiology of vascular malformations vascular anomalies. Magn Reson Imaging Clin N Am
is of the utmost importance in the diagnosis and 10(2):363–388, vii
1006
8. Kang GC, Song C (2008) Forty-one cervicofacial
vascular anomalies and their surgical
treatment–retrospection and review. Ann Acad Med
Singapore 37(3):165–179
9. Choi DJ, Alomari AI, Chaudry G, Orbach DB (2009)
Neurointerventional management of low-flow vascular
malformations of the head and neck. Neuroimaging
Clin N Am 19(2):199–218
10. Puig S, Casati B, Staudenherz A, Paya K (2005) Vas-
cular low-flow malformations in children: current con-
cepts for classification, diagnosis and therapy. Eur J
Radiol 53(1):35–45
11. Burrows PE, Mason KP (2004) Percutaneous treatment
of low flow vascular malformations. J Vasc Interv
Radiol 15(5):431–445
12. Yamaki T, Nozaki M, Sakurai H, Takeuchi M,
Soejima K, Kono T (2008) Prospective randomized
efficacy of ultrasound-guided foam sclerotherapy com-
pared with ultrasound-guided liquid sclerotherapy in
the treatment of symptomatic venous malformations.
J Vasc Surg 47(3):578–584
13. Ch’ng S, Tan ST (2008) Facial port-wine stains –
clinical stratification and risks of neuro-ocular involve-
ment. J Plast Reconstr Aesthet Surg 61(8):889–893
14. Perkins JA, Manning SC, Tempero RM, Cunningham
MJ, Edmonds JL Jr, Hoffer FA et al (2010) Lymphatic
malformations: review of current treatment.
Otolaryngol Head Neck Surg 142(6):795–803, e1
15. Bisdorff A, Mulliken JB, Carrico J, Robertson RL,
Burrows PE (2007) Intracranial vascular anomalies in
patients with periorbital lymphatic and lymphati-
covenous malformations. AJNR Am J Neuroradiol
28(2):335–341
16. Smith MC, Zimmerman MB, Burke DK, Bauman NM,
Sato Y, Smith RJ et al (2009) Efficacy and safety of
OK-432 immunotherapy of lymphatic malformations.
Laryngoscope 119(1):107–115
17. Dubois J, Garel L, Abela A, Laberge L, Yazbeck S
(1997) Lymphangiomas in children: percutaneous
sclerotherapy with an alcoholic solution of zein. Radi-
ology 204(3):651–654
18. Alomari AI, Karian VE, Lord DJ, Padua HM, Burrows
PE (2006) Percutaneous sclerotherapy for lymphatic
malformations: a retrospective analysis of patient-
evaluated improvement. J Vasc Interv Radiol 17(10):
1639–1648
19. Kim SW, Kavanagh K, Orbach DB, Alomari AI, Mul-
liken JB, Rahbar R (2011) Long-term outcome of
radiofrequency ablation for intraoral microcystic lym-
phatic malformation. Arch Otolaryngol Head Neck
Surg 137(12):1247–1250
20. Liu AS, Mulliken JB, Zurakowski D, Fishman SJ,
Greene AK (2010) Extracranial arteriovenous
malformations: natural progression and recurrence
after treatment. Plast Reconstr Surg 125(4):1185–1194
21. Revencu N, Boon LM, Mulliken JB, Enjolras O,
Cordisco MR, Burrows PE et al (2008) Parkes Weber
syndrome, vein of Galen aneurysmal malformation,
S.K. Pandey and D.B. Orbach
and other fast-flow vascular anomalies are caused by
RASA1 mutations. Hum Mutat 29(7):959–965
22. Tan WH, Baris HN, Burrows PE, Robson CD, Alomari
AI, Mulliken JB et al (2007) The spectrum of vascular
anomalies in patients with PTEN mutations: implica-
tions for diagnosis and management. J Med Genet
44(9):594–602
23. Gemmete JJ, Pandey AS, Kasten SJ, Chaudhary N
(2013) Endovascular methods for the treatment of vas-
cular anomalies. Neuroimaging Clin N Am 23(4):
703–728
24. Lee BB, Baumgartner I, Berlien HP, Bianchini G,
Burrows P, Do YS et al (2013) Consensus document
of the International Union of Angiology (IUA)-2013.
Current concept on the management of arterio-venous
management. Int Angiol 32(1):9–36
25. Kalani MY, Albuquerque FC, Fiorella D, McDougall
CG (2013) Endovascular treatment of cerebral arterio-
venous malformations. Neuroimaging Clin N Am
23(4):605–624
26. Wu JK, Bisdorff A, Gelbert F, Enjolras O, Burrows PE,
Mulliken JB (2005) Auricular arteriovenous malforma-
tion: evaluation, management, and outcome. Plast
Reconstr Surg 115(4):985–995
27. Persky MS, Yoo HJ, Berenstein A (2003) Management
of vascular malformations of the mandible and maxilla.
Laryngoscope 113(11):1885–1892
28. San Millan Ruiz D, Gailloud P (2010) Cerebral devel-
opmental venous anomalies. Childs Nerv Syst
26(10):1395–1406
29. Rivera PP, Willinsky RA, Porter PJ (2003) Intracranial
cavernous malformations. Neuroimaging Clin N Am
13(1):27–40
30. Lasjaunias P, Burrows P, Planet C (1986) Developmen-
tal venous anomalies (DVA): the so-called venous
angioma. Neurosurg Rev 9(3):233–242
31. Maeder P, Gudinchet F, Meuli R, de Tribolet N (1998)
Development of a cavernous malformation of the brain.
AJNR Am J Neuroradiol 19(6):1141–1143
32. Radvany MG, Rigamonti D, Gailloud P (2014) Angio-
graphic detection of cerebral cavernous malformations
with C-arm cone beam CT imaging in three patients.
J Neurointerv Surg 6(2):e17
33. Pollock BE, Garces YI, Stafford SL, Foote RL,
Schomberg PJ, Link MJ (2000) Stereotactic radiosur-
gery for cavernous malformations. J Neurosurg
93(6):987–991
34. Gross BA, Puri AS, Popp AJ, Du R (2013) Cerebral
capillary telangiectasias: a meta-analysis and review of
the literature. Neurosurg Rev 36(2):187–193, discus-
sion 94
35. Lee RR, Becher MW, Benson ML, Rigamonti D (1997)
Brain capillary telangiectasia: MR imaging appearance
and clinicohistopathologic findings. Radiology
205(3):797–805
36. Ellis MJ, Armstrong D, Vachhrajani S, Kulkarni AV,
Dirks PB, Drake JM et al (2013) Angioarchitectural
features associated with hemorrhagic presentation in
41 Pediatric Vascular Malformations
pediatric cerebral arteriovenous malformations.
J Neurointerv Surg 5(3):191–195
37. Mohr JP, Parides MK, Stapf C, Moquete E, Moy CS,
Overbey JR et al (2014) Medical management with or
without interventional therapy for unruptured brain
arteriovenous malformations (ARUBA): a multicentre,
non-blinded, randomised trial. Lancet 383(9917):
614–621
38. Krings T, Geibprasert S, Terbrugge K (2010) Classifi-
cation and endovascular management of pediatric cere-
bral vascular malformations. Neurosurg Clin N Am
21(3):463–482
39. Bhattacharya JJ, Luo CB, Suh DC, Alvarez H,
Rodesch G, Lasjaunias P (2001) Wyburn-Mason or
Bonnet-Dechaume-Blanc as Cerebrofacial Arteriove-
nous Metameric Syndromes (CAMS). A new concept
and a new classification. Interv Neuroradiol 7(1):5–17
40. Krings T, Geibprasert S, Luo CB, Bhattacharya JJ,
Alvarez H, Lasjaunias P (2007) Segmental
neurovascular syndromes in children. Neuroimaging
Clin N Am 17(2):245–258
41. Raybaud CA, Strother CM, Hald JK (1989) Aneu-
rysms of the vein of Galen: embryonic considerations
and anatomical features relating to the pathogenesis of
the malformation. Neuroradiology 31(2):109–128
42. Lasjaunias P (2007) Vascular diseases in neonates,
infants and children: interventional neuroradiology
management, 2nd edn
43. Mortazavi MM, Griessenauer CJ, Foreman P, Bavarsad
Shahripour R, Shoja MM, Rozzelle CJ et al (2013)
Vein of Galen aneurysmal malformations: critical
1007
analysis of the literature with proposal of a new classi-
fication system. J Neurosurg Pediatr 12(3):293–306
44. Chevret L, Durand P, Alvarez H, Lambert V,
Caeymax L, Rodesch G et al (2002) Severe cardiac
failure in newborns with VGAM. Prognosis signifi-
cance of hemodynamic parameters in neonates
presenting with severe heart failure owing to vein of
Galen arteriovenous malformation. Intensive Care
Med 28(8):1126–1130
45. Berenstein A, Fifi JT, Niimi Y, Presti S, Ortiz R, Ghatan
S et al (2012) Vein of Galen malformations in neonates:
new management paradigms for improving outcomes.
Neurosurgery 70(5):1207–1213, discussion 13–4
46. Spetzler RF, Detwiler PW, Riina HA, Porter RW
(2002) Modified classification of spinal cord vascular
lesions. J Neurosurg 96(2 Suppl):145–156
47. Patsalides A, Knopman J, Santillan A, Tsiouris AJ,
Riina H, Gobin YP (2011) Endovascular treatment of
spinal arteriovenous lesions: beyond the dural fistula.
AJNR Am J Neuroradiol 32(5):798–808
48. Kim LJ, Spetzler RF (2006) Classification and surgical
management of spinal arteriovenous lesions: arteriove-
nous fistulae and arteriovenous malformations.
Neurosurgery 59(5 Suppl 3):S195–S201, discussion
S3–13
49. Krings T, Geibprasert S (2009) Spinal dural arteriove-
nous fistulas. AJNR Am J Neuroradiol 30(4):639–648
50. Silva N Jr, Januel AC, Tall P, Cognard C (2007) Spinal
epidural arteriovenous fistulas associated with progres-
sive myelopathy. Report of four cases. J Neurosurg
Spine 6(6):552–558
 Pediatric Stroke
42
Daniel Tibussek, Gabrielle deVeber, and Manohar Shroff

Contents Perinatal Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1030

Perinatal Arterial Ischemic Stroke . . . . . . . . . . . . . . . . . 1030
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1010 Clinical Presentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1030
Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1010 Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1030
Clinical Features and Diagnostic Delays . . . . . . . . . . . 1011 Periventricular Venous Infarction . . . . . . . . . . . . . . . . . . 1031
Perinatal Hemorrhagic Stroke . . . . . . . . . . . . . . . . . . . . . . 1031
Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1012
Computed Tomography . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1012 Pediatric Cerebral Sinovenous Thrombosis . . . . 1032
Magnetic Resonance Imaging . . . . . . . . . . . . . . . . . . . . . . 1013 Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1032
Conventional Angiography . . . . . . . . . . . . . . . . . . . . . . . . . 1015 Clinical Presentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1032
Patterns of Childhood Arterial Ischemic Stroke . . . 1015 Risk Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1032
Imaging of CSVT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1032
Risk Factors of Arterial Ischemic Strokes . . . . . . 1019 Magnetic Resonance Imaging and Venography . . . 1033
Treatment of Arterial Ischemic Stroke . . . . . . . . . . . . . 1027 Computed Tomography . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1035
Follow-Up Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1035
Hemorrhagic Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1028
Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1035
Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1029
Outcome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1035
Prognosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1029
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1037
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1037

D. Tibussek (*)
Children’s Stroke Program, Division of Neurology,
The Hospital for Sick Children, University of Toronto,
Toronto, ON, Canada
Department of General Pediatrics, Neonatalogy and
Pediatric Cardiology, University Children’s Hospital,
Heinrich-Heine University, D€usseldorf, Germany
e-mail: daniel.tibussek@sickkids.ca;
daniel.tibussek@gmx.net
G. deVeber
Children’s Stroke Program, Division of Neurology,
The Hospital for Sick Children, University of Toronto,
Toronto, ON, Canada
e-mail: gabrielledev@gmail.com;
gabrielle.deveber@sickkids.ca
M. Shroff
Department of Diagnostic Imaging, The Hospital for Sick
Children, University of Toronto, Toronto, ON, Canada
e-mail: manohar.shroff@sickkids.ca

# Springer Science+Business Media New York 2016 1009

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_32
1010 D. Tibussek et al.

Abstract
Introduction
Over the last two decades, childhood ischemic
Over the last two decades, childhood ischemic
stroke, including arterial ischemic stroke and
stroke, including arterial ischemic stroke and cere-
cerebral sinovenous thrombosis (CSVT), has
bral sinovenous thrombosis (CSVT), has been
been increasingly recognized. Despite the
increasingly recognized. Although childhood
widely held belief that the increased plasticity
arterial ischemic stroke (CAIS) is rare compared
of the young brain protects against the effects
with adults, the estimated incidence ranges from
of injury, early brain injury due to stroke com-
1.6 to 13 per 100,000 children [1] and appears to
monly results in significant long-term
be rising [2]. Children under 1 year of age are
impairment.
particularly at risk with a peak incidence in the
Considerable differences exist between
perinatal period [1].
stroke in childhood and adults. These include
Considerable differences exist between stroke
maturational changes of the vascular, coagula-
in childhood and adults. These include matura-
tion, and nervous systems in neonates, older
tional changes of the vascular, coagulation, and
infants, children, and adults. In addition, etiol-
nervous systems in neonates, older infants, chil-
ogy differs significantly from adulthood.
dren, and adults. In addition, etiology differs sig-
Presenting symptoms and clinical findings
nificantly from adulthood. Presenting symptoms
vary with the age of the child. Finally, imaging
and clinical findings vary with the age of the child
may be more challenging with high rates of
and are frequently nonspecific, particularly in
falsely negative CT scans and a large number
infants. Finally, imaging may be more challenging
of clinical stroke mimics to consider. Together
with high rates of falsely negative CT scans and a
with a general lack of awareness, these factors
large number of clinical stroke mimics to con-
frequently result in significant delays in recog-
sider. Together with a general lack of awareness,
nition and appropriate investigation and treat-
these factors frequently result in significant delays
ment. Due to different underlying mechanisms
in recognition and appropriate investigation and
and important developmental changes
treatment.
throughout childhood, neuroimaging and man-
Despite the widely held belief that the
agement decisions in pediatric stroke cannot be
increased plasticity of the young brain protects
simply extrapolated from adults.
against the effects of injury, early brain injury
MRI imaging plays a crucial role in pediat-
due to stroke commonly results in significant
ric stroke in order to detect and characterize an
long-term impairment. Recent studies suggest
ischemic lesion and exclude common stroke
that early stroke often leads to widespread motor,
mimics. It also provides valuable information
behavioral, and cognitive dysfunction [3, 4].
about the potential etiology and prognosis.
CAIS results in neurologic morbidity in over
This in turn guides further diagnostic steps
two-thirds of survivors with considerable socio-
and assists in treatment decisions.
economic burden. For children with stroke
In the near future, more sophisticated imag-
beyond the perinatal period, recurrence rate is up
ing techniques such as vessel wall imaging,
to 25 % and, without antithrombotic treatment,
DTI, and functional imaging techniques will
even higher [5].
likely emerge into the clinical arena and play
an important role in enhancing our understand-
ing of pediatric stroke.
Epidemiology
Keywords
Stroke • Children • Neonatal • Sinovenous Stroke in childhood has long been thought to be
thrombosis • Ischemia • Haemorrhagic • rare and relatively benign. However, in the past
Haemorrhagic • Arteriopathy • Imaging three decades, epidemiological studies have
42 Pediatric Stroke
Fig. 1 Age distribution of 16
cases with childhood
arterial ischemic stroke (n = 14
96) (From Mallick et al. [1]
with permission) 12
Number of cases
10
0
<1
repeatedly demonstrated that, in childhood, stroke
is as common as brain tumors. A recent
population-based study of arterial ischemic stroke
in children (excluding neonates) found an inci-
dence rate of 1.6 per 100,000 per year [1]. A
significant effect of age on incidence was found,
with children under the age of 1 year being par-
ticularly at risk (4.14 per 100,000 per year). An
incidence above average has also been reported in
adolescence (Fig. 1).
Studies that included neonates have reported
higher incidence rates of 2–3 per 100,000 per
year. Neonates accounted for 8–46 % of total
pediatric AIS cases across different studies, and
perinatal stroke is as common as 1 in 2,500 live
births [6]. It is not completely understood why the
incidence rates of childhood arterial ischemic
stroke vary with age, but different susceptibilities
related to different underlying risk factors with
age are likely relevant. Ethnicity background
plays an important role in stroke risk with African
and Asian children having a significantly higher
incidence of stroke than Caucasian children
[1]. This is partially due to sickle cell disease in
African and moyamoya disease in Asian children.
Clinical Features and Diagnostic Delays
Any acute onset of a focal neurological deficit in a
child is a medical emergency, with ischemic or
hemorrhagic stroke being the most urgent
1011
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Age at stroke onset (completed years)
diagnosis to exclude. Although the majority of
children with CAIS will not be eligible for tPA,
an early diagnosis is still essential to guide imme-
diate neuroprotective and urgent antithrombotic
treatments. However, despite an increasing num-
ber of international publications about CAIS and
increased awareness of the entity over the last
decade, the diagnosis of stroke in children is still
frequently delayed or missed. A diagnostic delay
of 24 h and more from symptom onset is not
uncommon even when the child is brought to
medical attention within the first 6 h after symp-
tom onset [7]. Diagnostic delay is frequently max-
imal in the period following hospital arrival.
Besides a lack of awareness of CAIS among par-
ents and health professionals, a complex differen-
tial diagnosis with a wide spectrum of common
stroke mimics occurring in up to 30 % of apparent
strokes contributes to the diagnostic confusion. In
addition, in children with definite CAIS, a falsely
negative CT scan, still often the first imaging
study performed, has been reported in up to
84 % [7]. This has resulted in a move to urgent
MRI as initial imaging for CAIS.
Acute hemiparesis is the most common
presenting focal deficit. However, a variety of
other focal neurological deficits have been
described in CAIS (Table 1). Unlike in adulthood,
seizures at stroke onset are frequent in children
[1]. Age-related differences in clinical presenta-
tion have to be considered. Especially in younger
children, infants, and neonates where over
1012 D. Tibussek et al.

Table 1 Presenting signs and symptoms of arterial ische- Table 2 Advantages of MR imaging versus computed
mic stroke in different age groups (Adapted from Mallick tomography in childhood
et al. [1])
MR imaging Computed tomography
<1 1–5 6–10 11–15 Less widely available Widespread access
year (%) years (%) years (%) years (%)
Extended acquisition time; Very fast acquisition
Focal 75 89 70 91 sedation may be needed speed
features
No radiation Uses ionizing radiation
Hemiparesis 69 85 60 52
Diffusion-weighted Misses stroke diagnosis
Facial 25 47 40 39 imaging highly sensitive in the majority of
weakness and specific for early stroke children
Speech 13 32 40 48 Gradient-echo imaging Highly sensitive for
disturbances sensitive for detection of detection of acute
Diffuse 63 47 100 74 hemorrhage hemorrhage
features MR angiography (TOF) CT angiography requires
Decreased 60 36 50 39 does not require contrast intravenous contrast
conscious
TOF angiography is prone CT angiography provides
level
to flow artifacts more accurate anatomic
Headache 0 13 50 52 illustration of cerebral
Seizures 75 26 20 9 arterial system
Susceptibility-weighted Imaging of cerebral
images provide valuable venous system requires
information about cerebral contrast
venous system
three-quarters present with only seizures, symp- Mismatch studies (core Mismatch studies (core
tomatology can be very nonspecific and mislead- vs. penumbra) usually not vs. penumbra) usually
ing. CAIS may occur in the context of an acute available in pediatric not available in pediatric
centers centers
illness, such as meningitis or post cardiac surgery,
when clinical evaluation can be particularly
challenging.

posterior fossa. However, immediate access to an

Imaging
Urgent neuroimaging is indicated to establish the
diagnosis in every case of a child presenting with
an acute neurological deficit that is not otherwise
sufficiently explained. Neuroimaging will also help
to rule out complications, such as cerebral edema
or intracranial hemorrhages, and possibly establish
the etiology. Importantly, in young infants with an
open fontanel, cranial ultrasound (CUS) is not reli-
able and, due to wide operator variability and a
peripheral location of many infarcts, frequently
missed the diagnosis of CAIS [8].
Computed Tomography
MRI is superior to CT in detecting strokes in
children (Table 2). This is particularly true for
early and small strokes and those located in the
MRI is not always feasible and additional sedation
or anesthesia may be required compared with CT
leading to delayed imaging. Therefore, CT still
has a place in the acute diagnostic process. A
head CT scan can rule out intracranial hemorrhage
(ICH) and may identify subtle findings suggestive
of stroke. As in adults the main CT findings are
areas of cortical–subcortical hypoattenuating
within a vascular territory, obscuration and loss
of gray matter–white matter differentiation in the
basal ganglia, and demonstration of a
“hyperdense artery sign” (Fig. 2). CT may also
reveal hemorrhagic transformation or malignant
infarct edema. However, sensitivity to early ische-
mia is generally low [7] (Fig. 3). The presence of a
persistent arterial occlusion on CTA can also sug-
gest the diagnosis.
Modern CT techniques such as perfusion CT
are not readily available in pediatric radiology
departments. Furthermore, its use in childhood is
42 Pediatric Stroke
generally discouraged because of an excessive
radiation dosage. Therefore, compared to adults,
the role of an acute head CT to diagnose CAIS in
the pediatric population is limited.
Fig. 2 Hyperdense artery sign
Fig. 3 CT versus MRI of the same patient, 4 and 6 h after presentation with acute hemiparesis, respectively. The CT had
been reported as normal
1013
Magnetic Resonance Imaging
A high-priority MRI is the preferred imaging
technique for diagnosing acute stroke. Besides a
high sensitivity and specificity to identify early
strokes, identification of associated treatable con-
ditions and exclusion of stroke mimics (see
below) pose other important advantages. By com-
paring different MRI sequences, e.g., diffusion
weighted (DWI) and FLAIR, the age of ischemic
lesions from within 7 days to older can be esti-
mated. Other features including mass effect or
atrophy can also aid in estimating lesion age. In
older teenagers, if tPA is a therapeutic option
priority, 1 MRI is indicated (within 1 h) to confirm
CAIS versus other mimics. Otherwise an early
MRI as early as possible but within 4 h after
presentation is still recommended in order to ini-
tiate urgent antithrombotic and neuroprotective
therapy in those with definite CAIS. This is feasi-
ble with most children. A specific hospital stroke
protocol focusing on the most relevant findings of
interest is useful (Definite ischemia? Recent
vs. old lesion, bleeding? Arterial anatomy). A
stroke protocol MRI can usually be completed
within 10 min (Fig. 4).
1014 D. Tibussek et al.

Fig. 4 Simplified stroke 1) Diffusion weighted imaging/Apparent diffusion coefficient

protocol. This approach
allows exclusion of a stroke 2) EPI gradient echo T2*
after a maximum of 10 min 3) Axial FLAIR (>2 yrs) or Axial T2 (<2 yrs)
on the MR table
4) Time of flight MRA of circle of Willis

Study is supervised by neuroradiologist online.

Additional sequences added

according to findings of 1) – 4)

- Fat saturated T1
- Vessel wall imaging
- MR venography

DWI is the gold standard for diagnosis of acute

CAIS and has been proven helpful identifying
ischemic lesions within hours in neonates and
within minutes in older children [9]. Acute ische-
mic lesions demonstrate restricted diffusion,
evidenced by hyperintensity on trace diffusion
images. A matching reduced signal on apparent
diffusion coefficient (ADC) map helps to discrim-
inate recent from old lesions with so-called T2
shine-through. T2-WI and fluid-attenuated
inverted recovery (FLAIR) images may show
acutely infarcted tissue within 12–24 h of injury,
and abnormalities will usually persist.
MRA of the intracranial and neck vessels help
to identify and characterize an underlying
arteriopathy. Vascular imaging is an essential Fig. 5 Twelve-year-old girl with recurrent episodes of
slurred speech and right arm weakness. MR angiography
part of cerebral imaging in all children with stroke
shows proximal M1 stenosis
(Fig. 5). Vascular imaging demonstrates an intra-
or extracranial cerebral arteriopathy in about fifty
percent of children presenting with an AIS [10]. flight but not routinely performed in most pediat-
However, technical limitations require consid- ric centers.
eration. Because MRA relies on flow signal, areas SWI has recently been proven to be a useful
of normal (areas of branching or tortuosity) or MR sequence for CAIS for detecting hemorrhagic
abnormal (areas of stenosis) turbulent flow can components within the region of infarction and
mimic occlusions especially in infancy when identifying and quantifying microbleeds. In adults
arteries are smaller. TOF-MRA has been reported these have been shown to predict the risk of hem-
to overestimate the severity and length of a steno- orrhagic transformation before initiating thrombo-
sis. In addition, MRA has its limitations in reso- lytic treatment. Preliminary data suggest that SWI
lution, interfering with its ability to identify may be helpful for visualization of the penumbra
medium- or small-sized vessel diseases [11]. Gad- in children [12]. Potential pitfalls in SWI interpre-
olinium-enhanced MRA is superior to time of tation have been nicely summarized recently [13].
42 Pediatric Stroke 1015

Conventional Angiography

Despite advances in noninvasive imaging tech-

niques, conventional angiography (CA) remains
the gold standard for detailed cerebral vascular
imaging. Cerebral CA is the most sensitive and
specific method of diagnosing most cerebral
arteriopathies, many of which may only be iden-
tifiable with CA. CA also provides a dynamic
view of the arterial perfusion including areas of
low flow or, for collateral vessels and widened
zones of perfusion into the ischemic territory,
possible artery-to-artery embolism sources and
distal narrowing or occlusion. CA may identify
characteristic features of inflammatory and
non-inflammatory vasculopathies such as dissec-
tion and is recommended when central nervous
system vasculitis is suspected, especially when
MR imaging is not conclusive. In addition, CA
Fig. 6 Three-year-old male patient with acute onset left
including both intracranial and extracranial arm and face weakness. Diffusion-weighted images
carotid arteries is crucial for presurgical planning revealed a right paramedian pontine ischemia
of revascularization, e.g., in cases of moyamoya
disease.
Therefore, CA should be considered in all chil- peripheral wedge-shaped infarct involving the
dren with AIS in whom an etiology for CAIS cerebral cortex and subjacent white matter in ter-
remains elusive after initial MRA or CTA imag- ritorial distributions.
ing. With pediatric expertise provided, the com- The stroke pattern may differ according to the
plication rate for diagnostic CA is low even in underlying etiology. As an example, large, iso-
children below 3 years of age [14]. However, lated lenticulostriate AIS is commonly associated
radiation remains a concern, and carefully with parainfectious or inflammatory intracranial,
weighing up the expected additional information unilateral arteriopathy including transient cerebral
and the risk of unnecessary exposure to radiation arteriopathy or post-varicella arteriopathy [16,
is warranted. 17]. Small lacunar infarcts are rare in childhood
age groups except in cardiac or sickle cell disease.

Patterns of Childhood Arterial Ischemic Posterior Circulation Strokes

Stroke
Typical radiological features of AIS seen in chil-
dren are similar to those of young adults. How-
ever, maturational changes that occur during brain
development need to be taken into account. Imag-
ing patterns of CAIS will depend on the location
and the size of the vessels involved. Most CAIS
involve the territory of the middle cerebral artery
(MCA), most frequently affecting the left hemi-
sphere [15]. The classical imaging appearance of
occlusion of a large cerebral artery is the
The vertebrovascular or posterior circulation is
involved in up to 37 % of all CAIS cases [18],
and boys are more commonly affected. Typical
posterior circulation stroke patterns include occlu-
sions of basilar artery and its branches (Fig. 6),
proximal posterior cerebral artery (PCA) occlu-
sion with large-vessel territory infarction in the
occipital and temporal lobes, and infarcts involv-
ing the perforating artery territories in the thala-
mus and splenium. Multiple posterior circulation
infarcts involving different territories are not
uncommon (Fig. 7). Underlying arterial
1016
Fig. 7 Five-year-old male with sudden onset headache and unsteady gate. Multifocal diffusion restriction in the posterior
circulation
dissection or arteriopathies are frequently identi-
fied [18]. Although basilar artery strokes can
potentially be life threatening, recent pediatric
data suggest a much better prognosis in children
compared with adults [18, 19]. Death rate for
childhood brain-stem infarction is below 10 %,
and survivors frequently have good outcomes. In
contrast, recurrent ischemic events after posterior
circulation strokes have been described in up to
50 % [18].
Watershed (Border Zone) Infarcts
In contrast to the majority of CAIS which is due to
thrombotic vascular occlusion of one artery,
apparently isolated focal ischemic lesions can
also be caused by diffuse hypoperfusion. These
are usually located in typical watershed or border
zones at the anterior, middle, and/or posterior
cerebral artery (PCA) junctions. External (corti-
cal) border zone infarcts can be differentiated
from those affecting internal border zones
(Table 3). In childhood, internal or “deep” border
zone infarcts are frequently observed in children
with congenital heart disease or after
D. Tibussek et al.
Table 3 Classification of border zone infarcts (From
Mangla et al. [20] with permission)
External (cortical) infarcts
Frontal cortex (between the anterior and middle
cerebral arteries)
Occipital cortex (between the middle and posterior
cerebral arteries)
Paramedian white matter (between the anterior and
middle cerebral arteries)
Internal (subcortical) infarcts
Between the lenticulostriate and middle cerebral
arteries
Between the lenticulostriate and anterior cerebral
arteries
Between the Heubner and anterior cerebral arteries
Between the anterior choroidal and middle cerebral
arteries
Between the anterior choroidal and posterior cerebral
arteries
hypoxic–ischemic events (Fig. 8). Differentiating
these from cortical embolic strokes is not always
straightforward. Although watershed infarcts are
typically bilateral, they also can be asymmetric or
unilateral.
42 Pediatric Stroke 1017

Fig. 8 Neonate with congenital heart disease and neonatal seizures. Acute MRI revealed multifocal cortical diffusion
restrictions. A follow-up MRI 5 days later showed cortical laminar necrosis suggesting a hypoperfusion injury

Internal border zone infarcts, involving the
deep white matter, suggest occlusion at the distal
internal carotid or proximal anterior cerebral
artery (ACA) or MCA frequently occurring with
chronic occlusive vasculopathies, including
moyamoya (see below).
Hemorrhagic Transformation
Knowledge about hemorrhagic transformation
(HT) in CAIS is limited. In a retrospective series
of 61 children with AIS, hemorrhagic
transformation occurred in 19 (30 %), of
whom only two were symptomatic. Increased
infarct volume was associated with a higher risk
of HT, whereas the risk was low with
vasculopathies as the stroke mechanism. Use of
anticoagulation and antiplatelet therapy is not a
significant risk factor for HT in children [21, 22].
When HT occurs, it is usually ECASS
Grade 1 (some of which may actually be laminar
necrosis) or Grade 2 and usually
asymptomatic [22].
1018 D. Tibussek et al.

With any episode of neurological deterioration 1.3 % of more than 700 children treated for ische-
early after stroke, intracranial hemorrhage needs mic stroke [24]. However, in another smaller
to be considered and emergent neuroimaging is population-based study, 12.5 % (2 out of 16) of
indicated (Table 4). all pediatric ischemic strokes over two 1-year
periods developed malignant MCA
Malignant MCA Infarction infarctions [25].
Malignant MCA infarctions in the pediatric pop- Without decompressive surgery, death is fre-
ulation may not be as rare as previously thought. quent. Outcome data in children undergoing
In a recent multicenter hospital-based study, decompressive surgery however remains scarce,
malignant MCA infarctions were identified in and specific eligibility criteria for
hemicraniectomy in the pediatric population
Table 4 Temporal changes of hematoma on MR imaging have not yet been defined. However, available
(Adapted from Rasalkar and Chu [23] with permission) data suggest that pediatric survivors of malignant
Stage and Signal Signal MCA infarction and hemicraniectomy may have a
time from Hemoglobin intensity intensity more favorable functional outcome than those
hematoma stage on T1 on T2 without surgery [24]. Therefore, identification of
Hyperacute Oxy-Hb Gray White predictors of a malignant course of the MCA in
(<6 h)
the pediatric population would be of great
Acute Deoxy-Hb Gray Black
(6 h–3 importance.
days) In adults involvement of more than 50 % of the
Early Intracellular White Black MCA territory was found to be a reliable predic-
subacute meth-Hb tor. Preliminary pediatric data suggest that a high
(3–7 days)
pediatric ASPECT score (Fig. 9) and a high pedi-
Late Extracellular White White
subacute meth-Hb
atric NIHSS are predictive of malignant MCA.
(1–4 weeks) Large infarcts (>10 % of total brain volume)
Early Extracellular White White also predict poorer functional outcome in chil-
chronic (>4 meth-Hb with with dren. The modified pediatric ASPECT score cor-
weeks) hemosiderin black relates well with infarct volume (= percent of
rim rim
supratentorial brain volume) in children of differ-
Late Hemosiderin Black Black
chronic ent ages with excellent inter-rater reliability and
(months to validity. In addition, the modified ASPECTS
years) assigns points to small but important brain struc-
tures (e.g., PLIC), not just volumes, thereby

Fig. 9 Scoring areas for the modified pediatric C caudate, L lentiform, I insula, IC internal capsule,
ASPECTS. A1 proximal anterior cerebral artery (ACA) T thalamus, P2 superior portion of PCA territory, A2 distal
territory, P1 inferior portion of PCA territory, M MCA, ACA territory (From Beslow et al. [26] with permission)
42 Pediatric Stroke
Table 5 Stroke mimics in childhood (Adapted from
Shellhaas et al. [28])
Migraine
Conversion disorders
Seizures
Postictal paralysis
ADEM
Cerebellitis
Tumor
PRESa
Intoxication
Intracranial abscess
AVM
Idiopathic intracranial hypertension
Periodic hypertensive episodes
Metabolic stroke
Subdural empyema
Moyamoya
a
PRES posterior reversible encephalopathy syndrome
incorporating location along with volume in the
score. Therefore, this easy to use score seems to be
a useful tool for estimating infarct volume [26].
The permeability of the blood–brain barrier as
assessed with MRI, PET, and SPECT also predicts
a malignant course of MCA infarction. However,
these techniques need further prospective evaluation.
Stroke Mimics
In the majority of children referred to tertiary
centers with an acute focal neurological deficit, a
stroke mimic will be diagnosed including
migraine, seizures, Bell palsy, and conversion dis-
orders [27]. Even when stroke is the suspected
diagnosis prior to imaging, around 20 % will
turn out to be other stroke mimics [28] (Table 5).
It can be assumed that the number is even higher
in non-tertiary centers. Knowledge about these
conditions is therefore of great importance, and
establishing the diagnosis requires expertise in
both child neurology and pediatric neuroimaging.
Careful history taking and examination may help
to define a provisional diagnosis. However, MRI
will often be required to exclude a stroke with
certainty and to establish the diagnosis of AIS in
most cases [23]. Many of the conditions mimick-
ing stroke require a specific treatment or further
workup.
1019
Table 6 Categorization of childhood arterial ischemic
stroke risk factors (International Pediatric Stroke Study)
(Mackay et al. [10])
Arteriopathies
Cardiac disorders
Chronic systemic disorders
Prothrombotic states
Acute systemic disorders
Chronic head and neck disorders
Acute head and neck disorders
Risk factors for artheriosclerosis in adulthood
Risk Factors of Arterial Ischemic
Strokes
Risk factors associated with CAIS are diverse and
differ substantially from adults in whom risk factors
act to promote atherosclerosis of the cerebral (or in
cardiogenic stroke, the coronary) arteries (Table 6).
According to the data from the International Pedi-
atric Stroke Study [10], preexisting risk factors were
present in almost 50 % of children presenting with
acute ischemic stroke beyond the perinatal period.
These consisted mainly of congenital heart disease
and other chronic childhood disorders. Following a
thorough etiologic evaluation, currently up to 90 %
of children will have at least one risk factor identi-
fied, with the majority having two or more identifi-
able factors. Although causality is not always easy
to proof, the identification of an underlying etiology
is important as it may help to determine the risk of
stroke recurrence and lead to a more specific or even
causal therapy [5]. Therefore, a systemic diagnostic
approach is always warranted. The results of neuro-
imaging will guide further management.
The most relevant risk factors of CAIS will be
discussed in more detail:
1. Cerebral arteriopathies
Up to 53 % of children with stroke who
undergo vascular imaging have evidence of
an underlying cerebral arteriopathy [29, 10].
Since arteriopathies have been shown to be a
strong predictor of stroke recurrence, detailed
cerebral and cervical vascular imaging for
detection and classification of an arteriopathy
is crucial in all children with CAIS [5, 16, 17].
1020
Classification of childhood arteriopathies
includes six categories:
(A) Transient cerebral arteriopathy
(B) Moyamoya
(C) Fibromuscular dysplasia
(D) Dissection
(E) Vasculitis
(F) Congenital arteriopathies
Among these, the four most commonly
encountered will be described below, namely,
(A)–(D):
(A) Transient cerebral arteriopathy
Transient cerebral arteriopathy (TCA) is a
nonprogressive, often reversible, unilat-
eral arteriopathy involving the anterior cir-
culation at the circle of Willis. It has been
recognized as a leading cause of CAIS
[29, 17]. Because the diagnosis of TCA
is confirmed by a reversible or stable
course on repeated vascular imaging after
6 months, the less specific term focal cere-
bral arteriopathy (FCA) has been
suggested at baseline, when the further
course is still unpredictable [30].
However, recently in 79 children with
initial imaging showing a unilateral
arteriopathy characteristic for TCA,
94 % evolved to definite TCA indicating
that the diagnosis can be made with a
high degree of confidence on baseline
imaging [17]. In the latter study 6 %
evolved to other conditions, namely,
moyamoya. However, in these children,
nearly all had visible moyamoya vessels
at onset. More recently it was demon-
strated that intracranial dissection can
also mimic TCA [31].
TCA has been presumed to be a result
of inflammatory mechanisms. Therefore,
it has been referred to as nonprogressive
childhood primary angiitis of the central
nervous system in the rheumatology liter-
ature [11]. This forms the rationale for
steroid treatment during the initial months
after stroke at some centers in combina-
tion with antithrombotic and, if post-
varicella, with acyclovir treatment. Arte-
rial wall imaging using high-resolution
D. Tibussek et al.
contrast MRI may indeed reveal gadolin-
ium wall enhancement supporting the
diagnosis of vasculitis. However, differen-
tiation from normal periarterial enhance-
ment is important. If associated with
chicken pox in the preceding 12 months,
the term “post-varicella angiopathy” is
used [16]. There is now strong epidemio-
logical evidence to support a causal link
between a cerebral arteriopathy and vari-
cella zoster infection in the preceding year
[32] (Fig. 10).
Common imaging features of TCA are
[33]:
1. Unilateral arteriopathy of the large ves-
sels of the anterior circulation, typi-
cally affecting the distal ICA and
proximal segments of the MCA
and ACA.
2. Focal or segmental stenosis or occlu-
sion of these vessels on conventional
angiography. “Banding” or “striated”
appearance is often described on initial
CA (Fig. 11). Features of other
arteriopathies, such as dissection or
moyamoya, are absent.
3. Dynamic nature during the first weeks
and months with repeated imaging
often demonstrating worsening of arte-
rial changes.
4. A monophasic course over the long
term, with follow-up imaging after
6 months confirming no further pro-
gression, and then either partial or com-
plete resolution of the arteriopathy.
In addition, infarcts are nearly always
located in perforator territories within the
basal ganglia zone with cortical involve-
ment limited to the overlying insular/tem-
poral cortex. TCA classically affects
previously healthy school-aged children.
Importantly, the risk of stroke recurrence
is at least 18 % and is maximal in the
initial few weeks after index stroke.
Many patients are left with some degree
of residual arterial stenosis. This high-
lights the importance of serial imaging in
these cases.
42 Pediatric Stroke
Fig. 10 Five-year-old boy who suddenly collapsed.
Developed right-sided hemiparesis. MRI shows left basal
ganglia stroke as a typical location of strokes related to
Fig. 11 Five-year-old boy, history of two episodes of
transient neurological symptoms over the past 2–3 days.
Prior MR showed right lentiform infarct. Conventional
angiography showing beading appearance of M1 segment
of right MCA (arrow)
(B) Moyamoya disease and moyamoya
syndrome
Moyamoya is a rare progressive cerebro-
vascular disorder characterized by
1021
vasculopathy. A history of chicken pox 2 months prior
implies a causal relationship
occlusive disease of the terminal internal
carotid arteries or the other arteries around
the circle of Willis with prominent arterial
collateral circulation. The characteristic
appearance of a dilated network of
lenticulostriate perforating vessels leads
to the term “moyamoya” (“puff of
smoke” in Japanese). If moyamoya is
associated with an underlying
condition (e.g., sickle cell disease, Neuro-
fibromatosis Type 1, Down syndrome), it
is referred to as “moyamoya syndrome.”
Moyamoya disease and moyamoya syn-
drome occur in children and adults of all
ages. Rarely it can present in infancy
(Fig. 12). Moyamoya can be bilateral or
unilateral.
Common clinical presentations in
childhood include headaches, transient
ischemic attacks, and AIS. Symptomatic
ischemic episodes can often be triggered
by crying, blowing, coughing, or hyper-
ventilation. Cognitive decline is another
concern. Intracranial hemorrhage, com-
mon in adults, is relatively rare in
childhood.
1022
Fig. 12 Seven-month-old male with sickle cell disease.
Presented with acute focal seizures and eye deviation to the
right. Acute diffusion restriction involving the MCA and
Studies from Asia suggest that
two-thirds of children will have symptom-
atic progression at 5 years after diagnosis.
The annual stroke risk in asymptomatic
pediatric patients was estimated at 3.2 %
[34]. Little is known about the natural
course of moyamoya in non-Asian
children.
After diagnosis it is now widely
accepted that all affected children should
be referred for surgical evaluation. How-
ever, no strict criteria exist to define which
child should undergo revascularization
surgery. Imaging findings, including cere-
brovascular reserve (CVR) studies, clini-
cal symptoms, and ischemic events (TIA,
AIS), as well as neuropsychological
assessment, should all be taken into
consideration.
(a) Imaging of Moyamoya
The diagnosis of moyamoya can be easily
made using MRI and MRA (Fig. 13).
Besides visualization of unilateral or
bilateral stenosis of the terminal inter-
nal carotid arteries (ICAs) and the per-
forator artery collateral network,
conventional MRI sequences may
demonstrate the so-called ivy sign.
This is a sensitive but nonspecific sign
D. Tibussek et al.
ACA territory. MRA revealed non-visualization of the
right ACA and MCA. The patient was later diagnosed
with moyamoya syndrome
for moyamoya describing the presence
of high signal intensity at the outer
edge of the cortical ribbon on FLAIR
sequences, likely reflecting engorged
pial vessels. However, artifact-related
causes of FLAIR hyperintensity have
to be kept in mind, above all in children
receiving supplemental oxygen during
MRI under anesthesia.
In addition to focal infarction presumably
due to thrombotic occlusion in
established arterial territories,
moyamoya also produces a second,
very important mechanism for infarc-
tion, namely, hypoperfusion leading to
internal or deep watershed infarcts that
may be subclinical. The typical imag-
ing pattern of these deep white matter
infarcts is a “string of pearls” appear-
ance (Fig. 14). Impaired cerebrovascu-
lar autoregulation appears to be an
important underlying mechanism for
these infarcts [35]. Functional vascular
imaging can directly or indirectly deter-
mine inadequate perfusion and cere-
brovascular reactivity (CVR). In some
centers these techniques are routinely
used in patients with moyamoya dis-
ease and may assist in treatment deci-
sions and help to predict the success of
42 Pediatric Stroke 1023

Fig. 13 Seven-year-old girl with a recent history of “complicated” migraine, including headache and unilateral
weakness and tingling. MRI revealed an ivy sign. Moyamoya disease was confirmed with conventional angiography

angiographic findings include stenosis

or occlusion at the distal ICA and/or the
origin of the anterior cerebral and middle
cerebral arteries on both sides. In addi-
tion, moyamoya vessels are characteris-
tic reflecting the collateral vascular
networks of the lenticulostriate perfora-
tors. CA also helps to visualize the
dynamics of the blood flow, collatera-
lization, and areas of cerebral hypo-
perfusion. Before revascularization it
also assists in displaying arteries poten-
tially available for bypass surgery
(superficial temporal artery, external
carotid artery).

(b) Sickle Cell Disease (SCD) and Moyamoya

Syndrome
Strokes affect 7–11 % of children with
Fig. 14 Same patient as Fig. 13: typical “string of pearls” sickle cell disease. Ischemic strokes
appearance indicating chronic deep (internal) border zone predominately affect young patients
infarcts between the ages of 2 and 5. Hemor-
rhagic strokes have the highest preva-
lence in patients between the ages of
reperfusion after revascularization 20 and 29 years with SCD. Silent
surgery. infarcts were identified in 21.8 % of
The gold standard to establish the diagnosis children between 6 and 19 years of
of moyamoya is conventional angiogra- age with SCD-SS in one study. Further
phy (Fig. 13). Characteristic studies have shown an estimated
1024
cumulative incidence of silent infarct as
high as 37 % in patients with SCD by
14 years of age [36].
How SCD leads to cerebrovascular disease
is only partly understood and likely mul-
tifactorial. The two favored mechanisms
include a progressive internal carotid
moyamoya-type arteriopathy and an
occlusion of small arteries by sickled
cells and related thrombus resulting in
randomly distributed multifocal small
infarcts.
In about 64 % of children with homozy-
gous SCD, stroke is related to a large-
vessel occlusive vasculopathy, usually
moyamoya syndrome. The
supraclinoid ICA is most affected with
less involvement of the proximal mid-
dle and anterior cerebral arteries. Pro-
gression of this vasculopathy with
formation of moyamoya collaterals
has been reported in up to 40 % of
patients [37]. However, a small per-
centage of children have clinically
apparent strokes without evidence of a
large-vessel vasculopathy.
Up to 37 % of all children with sickle cell
disease will have areas of signal abnor-
mality on MRI without a history of
acute neurological impairment indicat-
ing silent ischemic infarction. Such
children develop progressive cognitive
decline as the lesions accumulate.
These lesions are typically small and
multiple and follow the distribution of
the internal white matter border zones.
In serial MRI with MRA, screening for
presymptomatic cerebrovascular dis-
ease is now standard of care in SCD
to detect silent infarcts secondary to
small-vessel disease and possible
moyamoya.
Transcranial Doppler (TCD) screening to
measure flow and detect large-vessel
arteriopathy and predict stroke risk is
now standard of care of children with
SCD based on strong evidence
[38]. Flow rates above 200 cm/s predict
D. Tibussek et al.
increased stroke risk, and regular trans-
fusion therapy (target hemoglobin S
<20–30 %) significantly reduces the
stroke risk in children with elevated
Doppler velocities. Additional stroke
prevention strategies are currently
under evaluation including hydroxy-
urea and nocturnal oxygen supplemen-
tation. Aspirin should be considered,
especially in cases with large-vessel
vasculopathy and/or recurrent stroke.
Revascularization surgery may be suc-
cessful in SCD-related moyamoya.
(C) Fibromuscular dysplasia
FMD is a segmental vasculopathy of
unknown etiology. The disease predomi-
nantly affects the renal arteries and the
internal carotid arteries. In contrast to
adults, where medial fibrodysplasia is
common, intimal fibrodysplasia is the typ-
ical pathological finding in children
[39]. Involvement of other vessels can be
observed, including hepatic, celiac, mes-
enteric, axillary, and external iliac arteries.
Childhood FMD is believed to be rare and
poorly understood.
Characteristic angiographic findings
have been described; however, they are
nonspecific. The most common finding is
the so-called arterial “string of beads”
appearance, characterized by multifocal
short stenoses (Fig. 11). However, this
finding is also well described in other
arteriopathies of childhood [17]. There-
fore, the diagnosis of FMD is challenging
and cannot be confirmed based solely on
angiographic features, renal arteriopathy,
or clinical findings. Histopathology is cur-
rently the only confirmative diagnostic
tool for childhood fibromuscular dysplasia
[39]. As in other arteriopathies, there is a
risk of recurrence of ischemic stroke
[39]. Preventive treatment, mainly aspirin,
will usually be indicated in cases of
pediatric FMD.
(D) Cervical arterial dissection
Arterial dissection is likely to be
underdiagnosed in children. It accounts
42 Pediatric Stroke
for 14–20 % of CAIS cases and has been
reported in up to 50 % of posterior circu-
lation strokes [40]. Males are more com-
monly affected. Besides head and neck
trauma, several potentially causative fac-
tors have been reported, including preced-
ing infections, inflammation, connective
tissue disorders, and homocystinuria.
Non-accidental trauma should be consid-
ered as an important differential diagnosis,
especially in small children with
dissection.
In children, the dissecting event is usu-
ally followed by a latent period lasting days
or even weeks after trauma. New onset
head or neck pain may be a presenting
feature in children. Compared to adults,
children with dissection have a high likeli-
hood of presenting with an AIS as the first
symptom of dissection. Recurrence rates
are estimated at least 20 % [5, 40].
(a) Imaging Features
In posterior circulation dissection, verte-
bral artery at the level of C1–C2 is the
most typical location. Extracranial ICA
dissections commonly occur near the
origin of the ICA or at the distal
segment before or immediately after
entering the skull base [40]. Intracranial
dissection involving the anterior
circulation in children also occurs
and carries the risk of dissecting
aneurysms secondary to subarachnoid
hemorrhage [40].
Neuroimaging is essential to establish the
diagnosis of dissection [41]. MRI will
detect ischemic lesions, which are
commonly found at multiple locations,
especially in the posterior circulation.
In these cases MRI allows estimation
of the timing of the ischemic events.
Specific arterial wall MRI findings of
dissection include an intramural hema-
toma that is demonstrated as enlarge-
ment of the artery with an eccentrically
located crescent-shaped rim of abnor-
mal signal. Axial fat-suppressed,
non-contrast T1-WI images through
1025
the neck or skull base will best visual-
ize the abnormal intramural signal. In
addition, a tapered narrowing or even
occlusion of the dissected vessel at
characteristic arterial locations may be
visualized on MRA.
However, negative MR and CT imaging of
the cervical arteries does not exclude the
diagnosis of ICAD. It is important to
recognize that time-of-flight MRA can
miss dissection in up to 61 % of cases.
Especially in cases of presumed dissec-
tion where the diagnosis could not be
established by MRA or CTA, CA
remains the gold standard. Typical
angiographic findings include an intimal
flap, a double lumen, or a long tapered
stenosis at characteristic locations (also
referred to as “flame-shaped” stenosis).
The formation of a pseudoaneurysm will
sometimes be identified on follow-up
imaging (Fig. 15).
Follow-up imaging is always indicated
and will guide treatment decisions.
Complete resolution will occur in the
majority of cases [40]. However, per-
sistence (23 %) or even progression
(10 %) has been reported.
2. Cardiac disorders
Congenital heart disease is a risk factor for
embolic stroke [42], especially in the setting of
cardiac surgery, cardiac catheterization, extra-
corporeal membrane oxygenation (ECMO),
and ventricular assist device (VAD) placement.
Cardiac diseases are responsible for about
one-third of all AIS in children [43, 15].
However, ischemic brain lesions related to
congenital heart disease are frequently asymp-
tomatic and may be found on routine
preoperative MRI.
In addition to congenital heart disease,
acquired cardiac etiologies predispose to AIS,
including infective endocarditis, cardiomyop-
athy, valvular heart disease, arrhythmia, and
atrial myxoma.
AIS in multiple, distinct vascular
territories suggests a cardiac source. Infarcts
are usually thromboembolic with typical
1026 D. Tibussek et al.

Fig. 15 Ten-year-old boy with sudden neck pain and vertebral artery. Follow-up angiography after 3 months
dizziness during basketball game. Conventional angiogra- revealed a pseudoaneurysm
phy revealed a filling defect and irregular stenosis of right

peripheral infarcts involving the cortex and Table 7 Role of thrombophilia in childhood arterial
ischemic stroke and cerebral sinovenous system (Trenor
subcortical white matter of both the anterior and Michelson [44] with permission)
and posterior circulation. Vascular imaging
Odds ratio Odds ratio
may show abrupt occlusion(s) indicating (95 % (95 %
emboli. confidence confidence
Hypoperfusion patterns of infarction lim- interval) for interval) for
ited to watershed distributions are commonly Thrombophilia AIS CSVT
Two or more 18.75 6.12
seen after acute changes in cardiac output. The
genetic traits (6.49–54.14) (0.87–43.07)
role of PFO as a conduit for paradoxical embo-
Protein C 11.0 6.12
lism remains a concern, and to date the risk of deficiency (5.13–23.59) (0.87–43.07)
stroke in isolated childhood PFO remains Antiphospholipid 6.95 –
unknown. antibodies (3.67–13.14)
3. Prothrombotic conditions Lipoprotein 6.53 –
Inherited or acquired thrombophilia has been (a) elevation (4.46–9.55)
Factor V Leiden 3.70 2.74
identified in up to 50 % of children with
(2.82–4.85) (1.73–4.34)
AIS (Table 7). The overall impact of Antithrombin 3.29 18.41
thrombophilia in childhood AIS is still deficiency (0.70–15.48) (3.25–104.29)
unclear [44]. Prothrombotic disorders Factor II 2.60 1.95
likely play a role as additional “predisposing” G20210A (1.66–4.08) (0.93–4.07)
risk factors working in concert with other MTHFR 1.58 –
“triggering” factors at the time of AIS. thermolabile (1.20–2.08)
Protein S 1.49 5.27
They also predict increased risk of recurrence.
deficiency (0.32–6.92) (1.53–18.21)
Coagulation testing is generally recommended
in any child with AIS including children
with other identified risk factors. However,
accurate testing and interpretation of test 4. Genetics and stroke
results requires an experienced laboratory The role of genetic factors in causing child-
with well-established age-appropriate norma- hood stroke is increasingly recognized but
tive values. complex [45]. It is important to recognize that
42 Pediatric Stroke
Fig. 16 Example of a patient with PHACES syndrome
and multiple vascular abnormalities on MR angiography
certain genetic syndromes are associated with
an increased risk of strokes in childhood that is
usually related to cerebral vasculopathies (e.g.,
Neurofibromatosis Type 1, PHACES (Fig. 16),
Alagille, Down, Williams, moyamoya).
More recently, monogenetic disorders asso-
ciated with arteriopathies have been identified,
among these COL4A1, ACTA2 (Fig. 17), and
RNF213 in familial moyamoya cases.
Characteristic imaging patterns have been
described in some of these genetic disorders.
COL4A1 mutations may be a relatively
common cause for a genetic small-vessel dis-
ease affecting the brain, eyes, and kidneys,
leading to intracerebral hemorrhage, cerebral
calcification, and ischemic damage typically
with ante-, peri-, or postnatal onset. Character-
istically, but not exclusively, the disorder is
associated with porencephaly.
Several inborn errors of metabolism predis-
pose to AIS. Homocystinuria, a potentially
treatable condition, can lead to both arterial
and venous thrombosis and may be recognized
by clinical features and laboratory markers
[46]. In mitochondrial diseases, including
mitochondrial encephalopathy, lactic acidosis,
and stroke-like episodes (Fig. 18), associated
1027
Fig. 17 Patient with ACTA2 mutation: symmetric prom-
inence of the distal cervical segments of the ICAs and
fusiform aneurysmal dilatation of the petrous and cavern-
ous segments. Symmetric stenoses and tapering noted at
the mid cavernous ICA segments and extending to the
carotid terminations, involving the anterior and middle
cerebral arteries with straightening of the major branches
stroke-like episodes mimic AIS; however, the
mechanism is metabolic rather than primarily
ischemic [47].
Treatment of Arterial Ischemic Stroke
The main focus of early stroke management is
neuroprotection, prevention of stroke evolution,
and recurrence and early rehabilitation. Treatment
of CAIS is based on consensus guidelines for
acute and chronic treatment of CAIS [48–50].
Neuroprotection is a crucial mainstay of acute
CAIS treatment. This includes adequate oxygen-
ation, maintenance of normoglycemia, mainte-
nance of blood pressure between the 50th and
90th percentiles, and, importantly in children,
aggressive seizure and fever control. Especially
in the child with reduced level of consciousness,
continuous EEG should be considered in order to
detect nonconvulsive seizures.
Further treatment options for acute therapy of
CAIS include anticoagulation (ultra-fractionated
or low-molecular weight heparin) or antiplatelet
therapy with acetylsalicylic acid (ASA). No com-
parative studies are available to prove superiority
of one or the other drug.
1028
Fig. 18 Example of a patient with proven MELAS syndrome who developed three stroke-like episodes within 6 months.
MRI revealed areas of diffusion restriction not matching arterial territories
While in adults strong evidence supports the
usage of IV tPA, in CAIS no such evidence exists.
Safety concerns lead to the recommendation not
to use tPA in childhood stroke outside clinical
trials, even if a child presents within the 4.5 h
window. However, despite safety concern, tPA is
increasingly administered to children, frequently
outside the recommended 4.5 h window. Time to
treatment after onset up to 52 h has been
reported [51].
Successful thrombectomy has been reported in
childhood basilar artery thrombosis [52]. How-
ever, severe complications have also been
reported [53]. Safety and efficacy are basically
unknown and such device use in children is not
FDA approved. Multiple published pediatric
stroke guidelines all suggest avoidance of
endovascular procedures in the pediatric popula-
tion [48–50]. It is crucial that interventional neu-
roradiologists be aware of this. Since the devices
for endovascular management of acute ischemic
stroke were designed for use in adults, safety is an
even greater concern in the pediatric population.
Also, the likelihood of a better outcome in most
children with CAIS treated conservatively under-
lines the importance of avoiding endovascular
treatments as a general rule. The use of thrombo-
lytic therapies or thrombectomy in children with
CAIS must be considered highly experimental.
D. Tibussek et al.
There may be a role in selected children, for
example, an older child with life-threatening pos-
terior circulation stroke. Recommendations for an
optimal interdisciplinary approach to care of these
children have been recently proposed [52]. A
close collaboration of a neurointerventional team
with childhood stroke neurologists or, when not
available, a team including a pediatric neurologist
and adult stroke neurologist, where at least one
member of the team has extensive pediatric stroke
experience, is required.
Hemorrhagic Stroke
The incidence of hemorrhagic stroke (HS) in chil-
dren is similar to CAIS, approximately 1.4 per
100,000 children per year. Intracerebral hemor-
rhage (ICH) is the most common type of hemor-
rhagic stroke and accounts for around 50 % of
pediatric strokes.
Traumatic head injury, non-accidental injury,
bleeding disorders, arteriovenous malformations
(Fig. 19), and hemorrhagic neoplasms are some of
the causes of hemorrhagic stroke in children [54].
The clinical presentation of HS in childhood is
nonspecific, and symptomatology may be partic-
ularly misleading in young children and infants,
where altered mental status and convulsions are
42 Pediatric Stroke
Fig. 19 Sixteen-year-old girl with sudden onset severe headache. CT angiography revealed a ruptured arteriovenous
malformation. AVM nidus involving the left thalamus and corona radiata
the most common presenting symptoms. Typical
signs in older children include headaches and
sudden or rapidly progressive mental status
changes with focal signs, whereas seizures are
only observed in 16 %. An important diagnostic
hint in all age groups is the sudden onset of focal
symptoms [54]. A high level of suspicion is there-
fore needed to allow early diagnosis and
intervention.
Imaging
CT is generally considered the preferred acute
imaging of choice in children with suspected HS.
In a stable patient initial MRI may be consid-
ered. Blood-sensitive sequences such as SWI will
identify hemorrhages. In addition, MRI may be
helpful for differential diagnosis and identifica-
tion of an underlying etiology, such as AVM or
brain tumor. A thorough evaluation for vascular
anomalies is critical. If CTA and MRA do not
establish the cause of the HS, conventional angi-
ography needs to be considered.
After the diagnosis has been established and
acute treatment initiated, close monitoring in
1029
pediatric intensive care is always indicated. Any
further clinical deterioration should prompt a
follow-up CT scan to look for secondary lesions
such as hydrocephalus, extension of an intracere-
bral hematoma, herniation, or vasospasm.
Prognosis
Mortality in childhood hemorrhagic stroke is
about 23 %, usually as a result of the initial hem-
orrhage. Importantly, a recurrent cerebral bleed
resulted in death in 33 % of children in one
study. This underlines the importance of a thor-
ough etiological evaluation. More than 2/3 of
children will have motor impairment and/or cog-
nitive deficits at follow-up. Epilepsy developed in
11 % of children.
Prognosis of HS is linked to the intrapar-
enchymal component of the intraparenchymal
hemorrhage. Lesions including beyond 2 % of
total brain volume resulted in moderate disabil-
ities and [44] at or beyond 4 % in severe disability
or death. Another reported predictor of poor out-
come has been altered mental status within 6 h of
hospital arrival.
1030
Beslow et al. [55] recently demonstrated that a
pediatric ICH score derived from volume of hem-
orrhage, infratentorial location, and presence of
either hydrocephalus or herniation can predict the
extent of disability or death.
Perinatal Stroke
The perinatal period is a focused high-risk period
for stroke. Perinatal strokes include arterials
ischemic stroke, periventricular venous infarc-
tion, hemorrhagic stroke, and CSVT. Strokes can
present acutely in the newborn period or later in
infancy with the evolution of chronic hemiplegic
cerebral palsy as the brain matures.
Mortality of perinatal stroke is low, and fatal
cases are often related to an underlying disease.
However, the risk for long-term neurological def-
icits is significant, including unilateral spastic
cerebral palsy (up to 50 %) and impairments in
vision, cognition, language, and behavior [3, 56].
The lifelong risk for epilepsy is increased. Recur-
rence of stroke in the child with a history of
perinatal stroke is rare.
Perinatal Arterial Ischemic Stroke
Perinatal AIS comprises approximately 25 % of
pediatric AIS, with the remainder in older infants
and children. Studies report an incidence of peri-
natal strokes of approximately 1 per 1,600–5,000
live births [6]. However, the fact that MRI is not
routinely done in all neonates under intensive care
suggests that current estimates of incidence may be
too low. Also the frequency of delayed presentation
of infants diagnosed beyond 6 months of age with
an emerging severe hemiparesis and chronic AIS
indicates that silent perinatal infarcts having more
normal outcomes are likely missed frequently.
Male neonates are at higher risk for stroke.
Clinical Presentation
PAIS in the newborn often manifests with sei-
zures, typically unilateral. Commonly the babies
D. Tibussek et al.
are initially well, and seizures tend to occur later
in the first day than in neonates with
hypoxic–ischemic encephalopathy (HIE). Other
symptoms are nonspecific, including poor feed-
ing, hypotonia, and apnea. Presumed perinatal
ischemic stroke (PPIS) is a term that has been
introduced to characterize children, who typically
present at the age of 3–6 months with early signs
of hemiparesis (e.g., early hand preference) or
seizures [57].
Imaging
Perinatal arterial ischemic AISs are more typically
observed in the left hemisphere, mostly involving
the MCA territory. A brain MRI after neonatal
seizures is always indicated. In fact, a recent
study found that MRI after neonatal seizures con-
tribute to a final diagnosis in >90 % of cases, with
strokes being the second most common diagnosis
(12 %) after HIE [58].
Serial cranial ultrasound is now routinely
performed in most NICU. Ultrasound may reveal
a wedge-shaped area of increased echogenicity
with a linear demarcation line. However, stroke
detection rates are not satisfactory, and further
imaging is required to exclude stroke [8].
MRI is the most sensitive imaging modality for
detection of perinatal strokes. The most frequently
used sequences include T1-WI, T2-WI, and
diffusion-weighted imaging (DWI). As in older
age groups, DWI plays the most important role in
the diagnosis of an acute PAIS. While an ischemic
lesion will be dark on ADC in an early stage, it
normalizes at 6–10 days (“pseudonorma-
lization”). Within the first week after PAIS, T2
will show high signal intensity in the affected
cortex and white matter. Cortical highlighting is
often visible on T1 (Fig. 8). Typically a cystic
evolution will be seen on follow-up.
An important imaging finding early after an
ischemic insult is the “acute Wallerian”
degeneration, characterized by restricted diffusion
on DWI at the level of the ipsilateral cerebral
peduncle. This finding has been identified as a
reliable predictor of poor motor outcome [59]
(Fig. 20).
42 Pediatric Stroke
Fig. 20 Neonate with focal neonatal seizures due to
stroke. DWI revealed pre-Wallerian degeneration of the
corticospinal tract and corpus callosum
In children with PPIS neuroimaging shows
evidence of a remote, chronic infarction with
absent diffusion restriction, encephalomalacia,
gliosis, and atrophy of the cerebral peduncle
reflecting chronic Wallerian degeneration.
Potential risk factors of perinatal strokes are
commonly identified and significantly differ
from older children. However, they are diverse
and the exact causal role is not always
definable. Risk factors may be categorized in
prenatal/maternal (e.g., diabetes mellitus, thyroid
disease, infertility), peripartum (e.g., fetal distress,
prolonged rupture of membranes, maternal fever),
and neonatal conditions (e.g., low APGAR
scores, hypoglycemia, early onset bacterial
infection) [60].
Periventricular Venous Infarction
Periventricular venous infarction (PVI) has been
described as a distinct and previously underappre-
ciated stroke pattern predominantly seen in PPIS
[57]. PVI are typically localized in the
periventricular white matter, sparing subcortical
white matter, cortex, and basal ganglia, and occur
in premature brain rather than term babies. Contrary
to MRI findings of patients with periventricular
1031
Fig. 21 Periventricular venous infarction noted on MRI in
a 3-month-old infant with early signs of hemiparesis
leukomalacia that show rather diffuse ex vacuo
enlargement of the posterior aspects of the lateral
ventricles, MRI of patients with PVI shows a focal
gliosis at the site of infarct frequently with accom-
panying blood products. In addition, the lateral
ventricles typically show a very focal expansion
adjacent to the gliosis or a porencephalic cyst.
It is observed in term babies often without obvi-
ous perinatal risk factors. PVI represents a common
subgroup of presumed perinatal ischemic strokes
and is typically diagnosed among populations of
children with hemiplegic cerebral palsy [57].
Perinatal Hemorrhagic Stroke
Perinatal hemorrhagic stroke (PHS) is observed in
6.2 in 100,000 live births [38]. PHS is typically
unifocal and unilateral (Fig. 21). Most newborn
present with signs of encephalopathy (poor feed-
ing, apnea, respiratory distress) and seizures. Etiol-
ogies are difficult to identify and may be
multifactorial [61]. While an underlying vascular
malformation is often suggested, these are typically
not detectable, even on follow-up. Other risk
factors include thrombocytopenia, fetal distress,
1032
and postmaturity. Bleeding disorders are rarely
been identified. However, it is also important
to recognize that approximately 20 % of appar-
ently normal newborns have small amounts of
subdural, subarachnoid, or parenchymal hemor-
rhages after birth.
A recent study of hemorrhagic stroke in neo-
nates suggested that hemorrhagic transformation of
an arterial or venous infarction is the leading mech-
anism of PHS [61]. Importantly, PHS is a well-
documented complication of neonatal CSVT.
Pediatric Cerebral Sinovenous
Thrombosis
Epidemiology
Cerebral sinovenous thrombosis (CSVT) in child-
hood has been increasingly recognized during the
past several decades. It accounts for one in four
for cases of pediatric ischemic strokes, with an
incidence of around 1 per 100,000 [62]. The
highest incidence is found in neonates (2.6 per
100,000) [63]. Similar to AIS, important differ-
ences in CSVT in children compared with adults
have to be considered.
Clinical Presentation
The clinical presentation is influenced by the age
of the child, the extent and location of the throm-
bus, and the presence or absence of associated
venous infarction.
Seizures are the most common presentation in
neonates. Additional signs are usually nonspecific
and include lethargy, hypotonia, feeding
difficulties, respiratory distress, and apnea
[63]. Focal neurological deficits are rare [63, 64].
In severe cases, bulging fontanels, splaying of the
cranial sutures, and prominent scalp veins may be
observed, indicating poor cerebral venous drain-
age and increased intracranial pressure.
In older infants and children, signs of raised
intracranial pressure predominate, typically
D. Tibussek et al.
developing gradually over hours, days, or even
weeks. These include headache, papilledema, and
a decreased level of consciousness. Additional
visual disturbances are present in 18 % of children
(diplopia, visual field deficits, and progressive
blindness due to prolonged compression of the
optic nerves). Seizures are present in nearly half
of children [63]. Focal deficits are a presentation
of CSVT when focal hemorrhages or infarcts
accompany the thrombus.
Risk Factors
Risk factors for CSVT are age dependent. During
the first 28 days of life, these include maternal
fever/chorioamnionitis, hypoxic–ischemic injury,
dehydration, infection, and thrombophilia. Often,
neonates have more than one risk factor.
In childhood CSVT, thrombosis results from a
combination of systemic and local vascular
factors. Within individual patients, certain
underlying risk factors including prothrombotic
states may predispose to thrombosis, and other
states including acute illnesses or prothrombotic
medications act as triggering factors. Head and
neck infections including meningitis or otitis
media and mastoiditis are a common cause for
septic thrombophlebitis and CSVT among
toddlers.
Imaging of CSVT
The diagnosis of CSVT in infants and children can
be challenging. Findings and radiologic appear-
ances are more variable compared with adults.
Understanding of normal variations and possible
artifacts and correctly applying newer available
techniques are crucial.
MRI with venography (MRV) or CT venogra-
phy (CTV) are the methods of choice for
investigation of CSVT. The diagnosis is
established by demonstration of a causal throm-
bus, a lack of flow in the affected cerebral veins,
and possibly additional venous infarction or
42 Pediatric Stroke
hemorrhage. Brain parenchymal injuries can
result from venous congestions leading to
impaired parenchymal venous drainage, with
resultant venous hypertension, vasogenic and
cytotoxic edema, and possibly hemorrhagic
venous infarction.
The superficial venous system is more
frequently involved than the deep system with
the transverse, superior sagittal, sigmoid,
and straight sinuses being the most common
sites of CSVT. Multiple sites of obstruction
within the sinovenous system at the time of diag-
nosis may be present. In severe cases, the superfi-
cial and deep systems are thrombosed in
combination.
Nearly 60 % of pediatric patients with CSVT
have associated parenchymal infarcts
[63]. Thrombosis of the superficial system will
cause infarcts in the cortex and white matter and
may be bilateral. A thrombosis located in the deep
venous system usually affects one or both thalami
or the cerebellum [65].
Intraventricular or thalamic hemorrhage in
term babies strongly suggests sinovenous throm-
bosis involving the deep system.
Ultrasound
The presence of an open fontanel allows radiolo-
gist and neonatologists to image the brain and
sinovenous system with ultrasound. Cranial ultra-
sound may detect a midline thrombus in the supe-
rior sagittal sinus.
Doppler ultrasound can define absent or reduced
flow in the sinovenous channels in CSVT. A uni-
lateral thalamic hemorrhage or venous infarcts as
well as intraventricular hemorrhages can also be
detected by cranial ultrasound. Although cranial
ultrasound is highly operator dependent, it can be
very useful and effective in skilled hands. Supple-
mental cranial ultrasound with color and pulsed
Doppler may also improve diagnostic accuracy in
equivocal cases. One study showed that color
Doppler ultrasound is highly specific for detecting
true negative cases of cerebral venous thrombosis in
clinically suspected neonates. In this study, excel-
lent reading agreement was noted [66].
1033
Magnetic Resonance Imaging
and Venography
MR imaging is currently the diagnostic study of
choice in neonatal and childhood CSVT. MRI
with MRV is able to visualize flow, the actual
thrombus, and areas of infarction and may also
identify an underlying abnormality [65]. The
appearance of the thrombus on MRI is age depen-
dent. An acute thrombus within a sinus is isodense
or hyperintense with brain on T1-WI images and
dark on T2-WI images. In the subacute stage, the
clot is isointense to brain on T1-WI images and
hypointense on T2-WI images. A chronic throm-
bus becomes hyperintense on both T1-WI and
T2-WI images (Table 8).
Time-of-flight (TOF) MR venography is com-
monly used to visualize sinuses and cerebral veins
(Fig. 22). However, since the signal is flow depen-
dent, it is prone to artifacts, particularly in neo-
nates and infancy due to the slower venous flow
and anatomically smaller veins. Flow gaps are
also common in the nondominant transverse
sinus. In neonates, compression of posterior por-
tion of the superior sagittal sinus by the occipital
bone is another reason for a localized flow artifact.
Another commonly misinterpreted finding in neo-
nates is subdural blood adjacent to the tentorium
and the torcula that can be confused with lateral
sinus thrombosis. Imaging in different planes may
be necessary to differentiate both conditions.
Susceptibility-weighted imaging (SWI) has
been reported as another useful sequence in
confirming the presence of CSVT and its
complications. Deoxygenated hemoglobin within
cerebral veins serves as an intrinsic contrast that
helps to visualize even tiny cerebral veins. Abnor-
mal prominence of cerebral veins may be seen on
SWI especially when a CSVT involves deep
venous structures. This sign can be visible in
early CSVT. Blooming on SWI within
thrombosed cortical veins may be another useful
imaging finding indicating cortical vein thrombo-
sis. Finally, SWI is the most sensitive pulse
sequence for the detection of small areas of paren-
chymal hemorrhage.
1034 D. Tibussek et al.

Table 8 Appearance of thrombus at various stages on different MRI sequences (From Bracken et al. [67] with
permission)
Contrast-
Fluid- Susceptibility- enhanced Time-of-
Thrombus age T1-WI sensitive weighted T1-WI flight Contrast-enhanced
(stage) images images images images MRV MRV
<24 h SI$(#) SI(") +/SI# Filling Absent Filling defect
(hyperacute) defect flow
1–3 days (acute) SI$ SI# may SI# Filling Absent Filling defect
mimic normal defect flow
SI# flow void
3–14 days SI" Early subacute Early Late Late Late subacute SI" may
(subacute; early, SI# may subacute SI#. subacute subacute mimic normal flow
days 3–7; late, mimic normal Late subacute SI" may SI" may
days 3–7; late, flow void. may not show mimic mimic
days 7–14) Late subacute SI# normal normal
SI" flow flow
>14 days SI# SI# may May not show Chronic Absent Filling defect. Delay
(chronic) may mimic normal SI# clot may flow between contrast
mimic flow void enhance administration and
normal imaging may result in
flow thrombus enhancement
void
SI signal intensity, $ isointense, # hypointense, " hyperintense, (#) slightly hypointense, (") slightly hyperintense, +/
may or may not be

Fig. 22 Ten-year-old girl with a history of otitis media, right jugular bulb are noted. Corresponding bright signal
headache, and new onset blurred vision. On MRV flow on T2/FLAIR images (arrow)
gaps involving the right transverse and sigmoid sinus and
42 Pediatric Stroke
Computed Tomography
Computed tomography (CT), including CTV, is
still often the initial test used in children with
suspected CSVT. Easy access, short scanning
time, and ability to detect acute hemorrhage are
some advantages. On the unenhanced CT,
increased attenuation within cortical veins,
superior and inferior sagittal sinuses, or the deep
midline venous structures should raise suspicion
of CSVT. A number of CT signs have been
described as suggestive of CSVT (cord sign,
hyperdense dural sinus sign, empty delta sign);
however, sensitivity is insufficient. Normal cere-
bral veins and dural venous sinuses may be
hyperdense in newborns due to physiological
polycythemia.
However, in equivocal cases, particularly if deep
venous infarction or cortical venous thrombosis is
suspected, high-resolution CT venography as an
endoluminal technique may be required. At CTV a
thrombus is seen as an area of non-enhancement
within a venous structure. In addition, CTV may
help to clarify the venous anatomy in cases where
TOF MRV suggests absent flow.
Follow-Up Imaging
The timing and type of follow-up imaging after
confirmed CSVT varies with the clinical status
and type of therapy used in the individual patient.
When anticoagulation has been initiated, an early
follow-up CT or blood-sensitive MRI technique is
suggested to exclude hemorrhagic complications.
Further follow-up studies using MRV are indi-
cated to evaluate whether the thrombus is resolv-
ing or extending under therapy and before
termination of treatment.
Treatment
If applicable, specific primary treatment for all
reversible underlying risk factors, including mas-
toiditis or meningitis, is indicated. In addition,
neuroprotection, including maintenance of blood
pressure, correction of hyper- or hypoglycemia
1035
and fever, and prevention of recurrent seizures,
is essential. In older children treatment for
intracranial hypertension may be required.
This frequently includes acetazolamide and, if
persistent, serial lumbar punctures with removal
of CSF. Careful monitoring of headache and
visual impairment, the degree of papilledema,
and visual field defects is important. Rarely, in
resistant cases, placement of a lumboperitoneal
shunt or optic nerve sheath fenestration may be
necessary.
Available data from neonatal childhood studies
supports that anticoagulant therapy should be con-
sidered in children and neonates with CSVT. The
safety of anticoagulation therapy in children and
neonates with CSVT previously was supported by
numerous studies [68, 63]. Anticoagulation aims
to prevent the propagation of an existing thrombus
and to enable the dissolution of existing thrombus
by the activity of the fibrinolytic system. Without
anticoagulation propagation of the thrombus is
documented in 30 %, associated with worse
outcome [69].
However, in the treatment decisions, several
factors must be considered. These include the
extent and location of the thrombus, the presence
of intracranial hemorrhage, and the reversibility of
risk factors for the CSVT. If anticoagulants are not
given, it is crucial to reimage and assess for prop-
agation of the initial thrombus or evolution of
parenchymal venous infarcts 5–7 days later.
Outcome
CSVT is a potentially life-threatening condition.
In the Canadian Registry, a mortality rate of 8 %
was found [63]. Recurrent CSVT or other sys-
temic thromboses occur in 13 % of children with
CSVT. Recurrence may even occur despite
receiving initial anticoagulation therapy.
Few studies document long-term outcomes in
more detail. However, motor and cognitive
impairments, as well as epilepsy, can be fre-
quently observed [64]. The presence of a cerebral
venous infarction is one predictor of poor out-
come, particularly when infarction is bilateral
[63, 8].
1036
Prediction of Outcomes from Pediatric
Stroke: The Role of Imaging
Despite the widely held belief that the increased
plasticity of the young brain protects against the
effects of injury, there is now considerable evidence
that early brain injury results in significant long-
term impairment. Recent studies provide evidence
that early stroke-related brain injury commonly
leads to widespread motor, sensory, behavioral,
and cognitive dysfunction [70, 3, 71]. However,
outcomes are variable in all age groups.
Prediction of clinical outcomes of children
with ischemic stroke is of major importance for
clinical management including selection of more
aggressive therapeutic options, counseling of par-
ents, understanding the child’s health-related
needs, and planning of rehabilitation and early
intervention programs.
A number of cognitive outcome studies
suggested worse outcomes after neonatal or peri-
natal stroke compared to later in childhood. A
U-shaped curve relating cognitive deficits with
age is also reported. Lesions that occurred in
middle childhood (5–10 years old) led to better
cognitive outcome than early lesions (0–5 years
old) or late (10–18 years old). Age and brain mat-
uration likely play an important role when looking
at vulnerability and repair mechanisms [4].
Available studies are somewhat inconsistent
with regard to stroke volume and location and
cognitive outcome. Current evidence suggests
that combined cortical and subcortical lesions
are linked to a worse cognitive outcome [4].
Limited evidence suggests that motor outcome
may be predicted by type, localization, and extent
of the ischemic zone. In neonates concurrent
involvement of basal ganglia, posterior limb of
the internal capsule, and cerebral cortex predicted
hemiparesis compared with other patterns of
injury. Kirton et al. [72] showed that an abnormal
signal on diffusion-weighted MRI in the
descending corticospinal tract (“acute Wallerian
degeneration”) was predictive of hemiparesis
after neonatal stroke. Chronic Wallerian degener-
ation was seen in all children with hemiparesis
[72] (Fig. 20).
When studying children with stroke outside the
neonatal period, Domi et al. [59] were able to
D. Tibussek et al.
demonstrate the ability of acute DWI changes
within the descending corticospinal tracts to pre-
dict poor motor outcome [59].
The addition of studying brain plasticity to
studies of stroke lesion characteristics in
predicting long-term motor outcome in childhood
is a promising area of imaging research. Asym-
metry of diffusion tensor imaging (DTI) parame-
ters has been recently correlated with an increase
in severity of motor impairments after neonatal
strokes. DTI-based tractography performed
3 months after a perinatal stroke showed an excel-
lent positive predictive value in terms of later
unilateral motor dysfunction [73].
Novel methods of imaging classification of
pediatric stroke are just emerging which combine
type, location, and volume assessment,
including Perinatal Stroke Classification system
[57] and a pediatric modification of an adult
method, namely, the modified pediatric ASPECT
scoring [74]. Recognizable patterns of presumed
perinatal stroke may predict neurological morbid-
ity in [57].
With the rapid progress and availability of
sophisticated brain imaging techniques, more
detailed analyses of type, localization, and extent
of ischemic lesions as well as cerebral reorgani-
zation and functional consequences after cerebral
ischemia will be made possible.
Some imaging observations of yet unknown
significance may deserve further research.
A frequent finding in posterior circulation
strokes in neonates is a diffusion restriction of
the optic radiation, sometimes also affecting the
corpus callosum (Fig. 23).
A high percentage of these children suffering
from visual field abnormalities suggests a func-
tional significance. Another example is cerebellar
atrophy associated with a contralesional pontine
DWI signals that was recently demonstrated on
MRI after CAIS [75].
Post-stroke dystonia is a severe and common
movement disorder affecting approximately 25 %
of children after basal ganglia stroke. However,
research focusing on imaging prediction of
post-stroke movement disorders is not available.
In the future, functional imaging techniques such
as resting state fMRI and other methods of
42 Pediatric Stroke
Fig. 23 Neonatal stroke involving the right posterior cir-
culation. On DWI highlighting of the optic radiation is
noted, likely representing acute Wallerian degeneration
connectivity measurement will help to better under-
stand the underlying networks and to identify chil-
dren at risk.
Summary
Important progress has been made during recent
years contributing to a better understanding of
pediatric stroke. It is now clear that significant
differences between neonates (birth-28 days),
children (1 month–18 years), and adults need to
be considered when dealing with stroke patients.
Due to different underlying mechanisms and
important developmental changes throughout
childhood, neuroimaging and management deci-
sions in pediatric stroke cannot be simply extrap-
olated from adults.
Imaging, usually by means of MRI, plays a
crucial role in pediatric stroke in order to detect
and characterize an ischemic lesion and exclude
common stroke mimics. It also provides valuable
information about the potential etiology and prog-
nosis. This in turn guides further diagnostic steps
and assists in treatment decisions.
1037
In the near future more sophisticated imaging
techniques such as vessel wall imaging, DTI, and
functional imaging techniques will likely emerge
into the clinical arena and play an important role
in enhancing our understanding of pediatric
stroke.
References
1. Mallick AA, Ganesan V, Kirkham FJ et al (2014)
Childhood arterial ischaemic stroke incidence,
presenting features, and risk factors: a prospective
population-based study. Lancet Neurol 13(1):35–43
2. Agrawal N, Johnston SC, Wu YW et al (2009) Imaging
data reveal a higher pediatric stroke incidence than
prior US estimates. Stroke 40(11):3415–3421
3. Kirton A, deVeber G (2013) Life after perinatal stroke.
Stroke 44(11):3265–3271
4. Westmacott R, Askalan R, MacGregor D et al (2010)
Cognitive outcome following unilateral arterial
ischaemic stroke in childhood: effects of age at stroke
and lesion location. Dev Med Child Neurol
52(4):386–393
5. Fullerton HJ, Wu YW, Sidney S et al (2007) Risk of
recurrent childhood arterial ischemic stroke in a
population-based cohort: the importance of cerebro-
vascular imaging. Pediatrics 119(3):495–501
6. Kirton A, deVeber G (2009) Advances in perinatal
ischemic stroke. Pediatr Neurol 40(3):205–214
7. Srinivasan J, Miller SP, Phan TG et al (2009) Delayed
recognition of initial stroke in children: need for
increased awareness. Pediatrics 124(2):e227–e234
8. Golomb MR, Dick PT, MacGregor DL et al (2003)
Cranial ultrasonography has a low sensitivity for
detecting arterial ischemic stroke in term neonates. J
Child Neurol 18(2):98–103
9. Groenendaal F, van der Ground J, Witkamp TD
et al (1995) Proton magnetic resonance spectroscopic
imaging in neonatal stroke. Neuropediatrics
26:243–248
10. Mackay MT, Wiznitzer M, Benedict SL et al (2011)
Arterial ischemic stroke risk factors: the international
pediatric stroke study. Ann Neurol 69:130–140
11. Benseler S, Pohl D (2013) Childhood central nervous
system vasculitis. Handb Clin Neurol 112:1065–1078
12. Meoded A, Poretti A, Benson JE et al (2014) Evalua-
tion of the ischemic penumbra focusing on the venous
drainage: the role of susceptibility weighted imaging
(SWI) in pediatric ischemic cerebral stroke.
J Neuroradiol 41(2):108–16
13. Bosemani T, Verschuuren SI, Poretti A et al (2013)
Pitfalls in susceptibility-weighted imaging of the pedi-
atric brain. J Neuroimaging 24:221–225
14. Hoffman CE, Santillan A, Rotman L et al (2014) Com-
plications of cerebral angiography in children younger
than 3 years of age. J Neurosurg Pediatr 13(4):414–419
1038
15. deVeber G, Canadian Paediatric Ischemic Stroke Study
Group (2000) Canadian paediatric ischemic stroke reg-
istry: analysis of children with arterial ischemic stroke
(Abstract). Ann Neurol 48(3):514
16. Askalan R, Laughlin S, Mayank S et al (2001)
Chickenpox and stroke in childhood: a study of fre-
quency and causation. Stroke 32(6):1257–1262
17. Braun KP, Bulder MM, Chabrier S et al (2009) The
course and outcome of unilateral intracranial
arteriopathy in 79 children with ischaemic stroke.
Brain 132(Pt 2):544–557
18. Mackay MT, Prabhu SP, Coleman L (2010) Childhood
posterior circulation arterial ischemic stroke. Stroke
41(10):2201–2209
19. Goeggel SB, Ritter B, Gautschi M et al (2013) Basilar
artery stroke in childhood. Dev Med Child Neurol
55(1):65–70
20. Mangla R, Kolar B, Almast J et al (2011) Border zone
infarcts: pathophysiologic and imaging characteristics.
Radiographics 31(5):1201–1214
21. Beslow LA, Smith SE, Vossough A et al (2011) Hem-
orrhagic transformation of childhood arterial ischemic
stroke. Stroke 42(4):941–946
22. Schechter T, Finkelstein Y, Ali M et al (2012)
Unfractionated heparin dosing in young infants: clini-
cal outcomes in a cohort monitored with anti-factor Xa
levels. J Thromb Haemost JTH 10(3):368–374
23. Rasalkar DD, Chu WC (2012) Imaging in children
presenting with acute neurological deficit: paediatric
stroke mimics. Postgrad Med J 88(1045):639–648
24. Smith SE, Kirkham FJ, deVeber G et al (2011) Out-
come following decompressive craniectomy for malig-
nant middle cerebral artery infarction in children. Dev
Med Child Neurol 53(1):29–33
25. Rahme R, Jimenez L, Bashir U et al (2013) Malignant
MCA territory infarction in the pediatric population:
subgroup analysis of the Greater Cincinnati/Northern
Kentucky Stroke Study. Childs Nerv Syst 29(1):
99–103
26. Beslow LA, Vossough A, Dahmoush HM et al (2012)
Modified pediatric ASPECTS correlates with infarct
volume in childhood arterial ischemic stroke. Front
Neurol 3:122
27. Mackay MT, Chua ZK, Lee M et al (2014) Stroke and
nonstroke brain attacks in children. Neurology
82(16):1434–1440
28. Shellhaas RA, Smith SE, O’Tool E et al (2006) Mimics
of childhood stroke: characteristics of a prospective
cohort. Pediatrics 118(2):704–709
29. Amlie-Lefond C, Bernard TJ, Sebire G et al (2009)
Predictors of cerebral arteriopathy in children with
arterial ischemic stroke: results of the International
Pediatric Stroke Study. Circulation 119(10):
1417–1423
30. Bernard TJ, Manco-Johnson MJ, Lo W et al (2012)
Towards a consensus-based classification of childhood
arterial ischemic stroke. Stroke 43(2):371–377
31. Dlamini N, Freeman JL, Mackay MT et al (2011) Intra-
cranial dissection mimicking transient cerebral
D. Tibussek et al.
arteriopathy in childhood arterial ischemic stroke. J
Child Neurol 26(9):1203–1206
32. Thomas SL, Minassian C, Ganesan V et al (2014)
Chickenpox and risk of stroke: a self-controlled case
series analysis. Clin Infect Dis 58(1):61–68
33. Mineyko A, Kirton A (2013) Mechanisms of pediatric
cerebral arteriopathy: an inflammatory debate. Pediatr
Neurol 48(1):14–23
34. Kurokawa T, Chen YJ, Tomita S et al (1985) Cerebro-
vascular occlusive disease with and without the
moyamoya vascular network in children.
Neuropediatrics 16(1):29–32
35. Mikulis DJ, Krolczyk G, Desal H et al (2005) Preop-
erative and postoperative mapping of cerebrovascular
reactivity in moyamoya disease by using blood oxygen
level-dependent magnetic resonance imaging. J
Neurosurg 103(2):347–355
36. Bernaudin F, Verlhac S, Arnaud C et al (2011) Impact
of early transcranial Doppler screening and intensive
therapy on cerebral vasculopathy outcome in a new-
born sickle cell anemia cohort. Blood 117(4):
1130–1140
37. Prengler M, Pavlakis SG, Prohovnik I et al (2002)
Sickle cell disease: the neurological complications.
Ann Neurol 51(5):543–552
38. Armstrong-Wells J, Grimes B, Sidney S et al (2009)
Utilization of TCD screening for primary stroke pre-
vention in children with sickle cell disease. Neurology
72(15):1316–1321
39. Kirton A, Crone M, Benseler S et al (2013)
Fibromuscular dysplasia and childhood stroke. Brain
J Neurol 136(Pt 6):1846–1856
40. Rafay MF, Armstrong D, deVeber G et al (2006)
Craniocervical arterial dissection in children: clinical
and radiographic presentation and outcome. J Child
Neurol 21(1):8–16
41. Orman G, Tekes A, Poretti A et al (2013) Posttraumatic
carotid artery dissection in children: not to be missed!
42. Giglia TM, Massicotte MP, Tweddell JS et al (2013)
Prevention and treatment of thrombosis in pediatric
and congenital heart disease: a scientific statement
from the American Heart Association. Circulation
128(24):2622–2703
43. Ganesan V, Prengler M, McShane MA et al (2003)
Investigation of risk factors in children with arterial
ischemic stroke. Ann Neurol 53(2):167–173
44. Trenor CC III, Michelson AD (2010) Thrombophilia
and pediatric stroke. Circulation 121(16):1795–1797
45. Munot P, Crow YJ, Ganesan V (2011) Paediatric
stroke: genetic insights into disease mechanisms and
treatment targets. Lancet Neurol 10(3):264–274
46. Bodensteiner JB, Hille MR, Riggs JE (1992) Clinical
features of vascular thrombosis following varicella.
Am J Dis Child 146(1):100–102
47. Mudd SH, Skovby F, Levy HL et al (1985) The natural
history of homocystinuria due to cystathionine beta-
synthase deficiency. Am J Hum Genet 37(1):1–31
48. Monagle P, Chalmers E, Chan A et al (2008)
Antithrombotic therapy in neonates and children:
42 Pediatric Stroke
American College of Chest Physicians evidence-based
clinical practice guidelines (8th Edition). Chest
133(6 Suppl):887S–968S
49. Paediatric Stroke Working Group (2004) Stroke in
childhood: clinical guidelines for diagnosis, manage-
ment and rehabilitation. Clinical Effectiveness
and Evaluation Unit, Royal College of Physicians,
London
50. Roach ES, Golomb MR, Adams R et al (2008) Man-
agement of stroke in infants and children: a scientific
statement from a Special Writing Group of the Amer-
ican Heart Association Stroke Council and the Council
on Cardiovascular Disease in the Young. Stroke
39(9):2644–2691
51. Amlie-Lefond C, deVeber G, Chan AK et al (2009) Use
of alteplase in childhood arterial ischaemic stroke: a
multicentre, observational, cohort study. Lancet Neurol
8(6):530–536
52. Ellis MJ, Amlie-Lefond C, Orbach DB (2012)
Endovascular therapy in children with acute ischemic
stroke: review and recommendations. Neurology
79(13 Suppl 1):S158–S164
53. Felker MV, Zimmer JA, Golomb MR (2010) Failure of
a clot retrieval device in an adolescent stroke patient.
Pediatr Neurol 43(6):435–438
54. Lo WD (2011) Childhood hemorrhagic stroke: an
important but understudied problem. J Child Neurol
26(9):1174–1185
55. Beslow LA, Ichord RN, Gindville MC et al (2014)
Pediatric intracerebral hemorrhage score: a simple
grading scale for intracerebral hemorrhage in children.
Stroke 45(1):66–70
56. Westmacott R, MacGregor D, Askalan R et al (2009)
Late emergence of cognitive deficits after unilateral
neonatal stroke. Stroke 40(6):2012–2019
57. Kirton A, deVeber G, Pontigon AM et al (2008) Pre-
sumed perinatal ischemic stroke: vascular classifica-
tion predicts outcomes. Ann Neurol 63(4):436–443
58. Osmond E, Billetop A, Jary S et al (2014) Neonatal
seizures: magnetic resonance imaging adds value in the
diagnosis and prediction of neurodisability. Acta
Paediatr 103:820–826
59. Domi T, DeVeber G, Shroff M et al (2009)
Corticospinal tract pre-wallerian degeneration: a
novel outcome predictor for pediatric stroke on acute
MRI. Stroke 40(3):780–787
60. Van der AN, Benders M, Groenendaal F et al (2014)
Neonatal stroke: a review of the current evidence on
epidemiology, pathogenesis, diagnostics and therapeu-
tic options. Acta Paediatr 103(4):356–364
1039
61. Bruno CJ, Beslow LA, Witmer CM et al (2014)
Haemorrhagic stroke in term and late preterm neonates.
Arch Dis Child Fetal Neonatal Ed 99(1):F48–F53
62. Mallick AA, O’Callaghan FJ (2010) The epidemiology
of childhood stroke. Eur J Paediatr Neurol 14(3):
197–205
63. deVeber G, Andrew M, Adams C et al (2001) Cerebral
sinovenous thrombosis in children. N Engl J Med
345(6):417–423
64. Jordan LC, Rafay MF, Smith SE et al (2010)
Antithrombotic treatment in neonatal cerebral
sinovenous thrombosis: results of the International
Pediatric Stroke Study. J Pediatr 156(5):704–710,
710.e1–710.e2
65. Shroff M, deVeber G (2003) Sinovenous
thrombosis in children. Neuroimaging Clin N Am
13(1):115–138
66. Miller E, Daneman A, Doria AS et al (2012) Color
Doppler US of normal cerebral venous sinuses in neo-
nates: a comparison with MR venography. Pediatr
Radiol 42(9):1070–1079
67. Bracken J, Barnacle A, Ditchfield M (2013) Potential
pitfalls in imaging of paediatric cerebral sinovenous
thrombosis. Pediatr Radiol 43(2):219–231
68. Moharir MD, Shroff M, Stephens D (2010) Anticoag-
ulants in pediatric cerebral sinovenous thrombosis: a
safety and outcome study. Ann Neurol 67(5):590–599
69. Moharir MD, Shroff M, Pontigon AM et al (2011) A
prospective outcome study of neonatal cerebral
sinovenous thrombosis. J Child Neurol 26(9):1137–1144
70. Ganesan V (2013) Outcome and rehabilitation after
childhood stroke. Handb Clin Neurol 112:1079–1083
71. Murias K, Brooks B, Kirton A et al (2014) A review of
cognitive outcomes in children following perinatal
stroke. Dev Neuropsychol 39(2):131–157
72. Kirton A, Shroff M, Visvanathan T et al (2007) Quan-
tified corticospinal tract diffusion restriction predicts
neonatal stroke outcome. Stroke 38(3):974–980
73. van der Aa NE, Leemans A, Northington FJ
et al (2011) Does diffusion tensor imaging-based
tractography at 3 months of age contribute to the pre-
diction of motor outcome after perinatal arterial ische-
mic stroke? Stroke 42(12):3410–3414
74. Beslow LA, Kasner SE, Smith SE et al (2012) Concur-
rent validity and reliability of retrospective scoring of
the Pediatric National Institutes of Health Stroke Scale.
Stroke 43(2):341–345
75. Mah S, deVeber G, Wei XC et al (2013) Cerebellar
atrophy in childhood arterial ischemic stroke: acute
diffusion MRI biomarkers. Stroke 44(9):2468–2474
 Congenital Vascular Syndromes and
Diseases 43
Sarah Milla, Jennifer Vaughn, and Nilesh K. Desai

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1041 There are several neurovascular diseases and
syndromes that are present at birth and may
Moyamoya Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1042
present in the pediatric age range. This chapter
PHACE Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1043 will describe the most common congenital
AVM/AVF Syndromes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1047 neurovascular entities and help to classify
Heriditary Hemorrhagic Telangiectasia . . . . . . . . . . . . 1047 them by dominant etiology (arterial or venous
Wyburn-Mason . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1049 abnormality) when known. The following enti-
Vein of Galen Malformation . . . . . . . . . . . . . . . . . . . . . . . 1050
Sturge-Weber Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . 1051
ties will be discussed: moyamoya, PHACE
syndrome, arteriovenous malformations and
Venous Malformations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1054
associated syndromes, Sturge-Weber syn-
Developmental Venous Anomalies . . . . . . . . . . . . . . . . . 1054
drome, venous malformations, and mitochon-
Cavernous Malformations . . . . . . . . . . . . . . . . . . . . . . . . 1055 drial encephalopathy with lactic acidosis and
Mitochondrial Myopathy, Encephalopathy, stroke-like symptoms (MELAS).
Lactic Acidosis, and Stroke-Like Symptoms
(MELAS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1057
Keywords
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1059 Pediatric stroke • Congenital neurovascular
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1059 syndromes • Moyamoya • PHACE • Sturge
Weber • MELAS

Introduction
S. Milla (*)
Department of Radiology, Emory University School of
Medicine Egleston Hospital, Children’s Healthcare of Several neurovascular abnormalities have an
Atlanta, Atlanta, GA, USA underlying congenital or genetic etiology. Abnor-
e-mail: sarahsarvismilla@gmail.com malities during embryology and genetic predispo-
J. Vaughn sitions may cause significant neurovascular
Department of Radiology, Boston Children’s Hospital, abnormalities such as hemorrhage and stroke.
Boston, MA, USA
Patients may present asymptomatically for
e-mail: Jennifer.Vaughn@childrens.harvard.edu
screening, such as a patient with a facial heman-
N.K. Desai
gioma, while others present acutely after stroke or
Department of Radiology, Texas Children’s Hospital,
Baylor College of Medicine, Houston, TX, USA hemorrhage, such as MCA territory infarct in a
e-mail: ndesai26@gmail.com patient with moyamoya disease or rupture of an
# Springer Science+Business Media New York 2016 1041
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_33
1042
arteriovenous malformation. When encountering
imaging being performed in infants and children
with neurologic symptoms, even as routine as
headache, knowledge of patient history or under-
lying syndromes, as well as the imaging evalua-
tion of the vasculature, parenchyma and
leptomeninges are important in making the appro-
priate and timely diagnosis of a neurovascular
disorder.
Moyamoya Disease
Background: Moyamoya arteriopathy manifests
as an acquired narrowing of the intracranial arter-
ies; typically involving the anterior circulation to
include the supraclinoid internal carotid arteries
(ICAs), the middle cerebral arteries (MCAs), and
the anterior cerebral arteries (ACAs). This
narrowing is a gradual process, which typically
results in enlargement and increasing number of
small basilar and perforating arteries to create
collateral vasculature. Early diagnosis of
moyamoya arteriopathy is critical for preservation
of neurologic function.
Moyamoya disease is defined as an idiopathic
bilateral arteriopathy in the absence of a systemic
disorder. This term is distinguished from
moyamoya syndrome, which is used when the
arteriopathy is found in the presence of a known
disease, syndrome or unilateral arteriopathy.
There are many pathologies and syndromes that
predispose patients to developing arteriopathies.
Most commonly encountered predispositions
include sickle cell anemia, Down’s syndrome,
neurofibromatosis type 1, and previous radiation
to the area (e.g., radiation therapy for brain
tumor). Less common associations include
Alagille syndrome, hyperthyroidism, congenital
cardiac anomalies, cervicofacial hemangiomas
(see section “PHACE Syndrome”), and renal
artery stenosis. Certain prothrombotic genes
have also been described as risk factors.
Imaging: Early findings of moyamoya
arteriopathy can be subtle, so a heightened aware-
ness of these subtle findings on conventional MR
imaging is critical. A subtle finding concerning for
moyamoya arteriopathy is curvilinear
S. Milla et al.
Fig. 1 “Ivy sign” within the leptomeninges on T2 FLAIR
(thin arrows) and white matter changes related to silent
ischemia often seen in watershed areas (thick arrows)
hyperintensity along the leptomeninges on T2
FLAIR imaging typically in the MCA territory,
also known as the “ivy sign” (Fig. 1). This should
be distinguished from the occasionally seen inci-
dental diffuse symmetric leptomeningeal
hyperintensity related to a variety of causes, such
as hyperoxygenation and propofol administration
in sedated children. Another subtle sign is the loss
of the normal flow void and decreased caliber of
the distal ICAs and MCAs, as seen best on the T2
weighted images (Fig. 2). Often the T2 images or
thin section T1 postcontrast images will show
multiple tiny basilar collateral vessels especially
along the sylvian fissures, suprasellar cistern and
the perimesencephalic cistern. Other findings seen
on conventional MR imaging are T2/FLAIR
hyperintensities suggesting foci of gliosis in the
subcortical or deep white matter, particularly in
the ACA/MCA watershed distribution, consistent
with ischemic changes, often referred to as clini-
cally silent infarcts (CSI) (Fig. 3). If moyamoya
arteriopathy is not detected and treated early, large
territorial infarctions can occur.
On MRA, narrowing of the distal ICAs and/or
the proximal MCA/ACAs can be seen (Fig. 4).
43 Congenital Vascular Syndromes and Diseases
Fig. 2 An 8-year-old with headaches. Decreased visuali-
zation of normal MCA flow voids on T2 (arrows)
Fig. 3 Same patient as Fig. 2. Eight-year-old with head-
aches. T2 FLAIR image demonstrates a focus of chronic
white matter ischemia in the left ACA/MCA watershed
territory (arrow)
MRA may show basilar collateral formation.
Digital subtraction angiography will show these
findings well, demonstrating the characteristically
described “puff of smoke” as the basilar collateral
1043
Fig. 4 Time of flight MRA, anterior projection MIP, in a
10-year-old with moyamoya demonstrates the absence of
flow in MCAs (thin arrows) and multiple tiny collaterals
(thick arrows)
formation causing thin wispy densities to appear
like strands of smoke (Fig. 5).
TREATMENT: The treatment for moyamoya
is a revascularization procedure. Many centers are
utilizing indirect or direct techniques to create
collaterals or reestablish arterial blood supply to
the cerebral areas at risk for infarct. A current com-
mon technique is an indirect revascularization,
encephalo-duro-arterio-synangiosis (EDAS), which
can be combined with other techniques such as pial
synangiosis, dural inversion, or split dural tech-
niques in order to optimize collateral formation
(Fig. 6). Some centers may include perfusion
imaging techniques to determine presurgical and
postsurgical appearances.
PHACE Syndrome
The presence of an infantile hemangioma in the
head and neck region, particularly if in a segmen-
tal distribution, should raise the clinical concern
for associated intracranial and vascular abnormal-
ities. The acronym PHACE (or PHACES) repre-
sents the spectrum of findings in this rare
neurocutaneous condition.
1044 S. Milla et al.

Fig. 5 Eleven year old with new stroke found to have (arrowhead). (b). Follow-up 2 years later demonstrates
moyamoya. (a). At initial diagnosis, right ICA injection complete occlusion of MCA with formation of basilar
(lateral) projection, showing narrowing of MCA collaterals, aka “moyamoya vessels” (arrows)

Fig. 6 Early (a) and delayed (b) images from ECA injection (AP projection). 13-year-old with moyamoya, status post
revascularization with pial synangiosis using right superficial temporal artery (arrow)
43 Congenital Vascular Syndromes and Diseases 1045

Posterior fossa/intracranial abnormalities

Hemangioma (cervicofacial)
Arterial cerebrovascular abnormalities
Cardiac abnormalities and aortic coarctation
Eye abnormalities

PHACE association is listed in Online Mende-

lian Inheritance in Man (OMIM 606519) and may
be termed PHACES when referring to the associ-
ation with the involvement of Sternal cleft or
supraumbilical raphe abnormalities.
Patients with hemangiomas of the head or neck
often undergo evaluation for PHACE syndrome
with imaging examinations. It is important to
understand the diagnostic imaging criteria for
making the clinical diagnosis. MRI of the brain
with and without contrast to evaluate for structural
abnormalities as well as optimized MRA of the
aorta, neck and head with and without contrast are Fig. 7 Axial T2-weighted image in child with right facial
important for accurate determination of arterial hemangioma demonstrates ipsilateral cerebellar hypopla-
abnormalities commonly seen in PHACE sia, enlarged posterior fossa, and absence of ipsilateral
patients. internal carotid artery (arrow)

Posterior fossa/intracranial abnormalities – A

major criterion for PHACE syndrome is poste-
rior fossa abnormalities. This is most typified
by cerebellar hypoplasia. This may manifest as
unilateral cerebellar hemispheric hypoplasia,
usually ipsilateral to the cervicofacial heman-
gioma, however the hypoplasia may be bilat-
eral or involving the cerebellar vermis (Fig. 7).
Intracranial hemangiomas have typical imag-
ing appearances for infantile hemangiomas and
are considered minor criteria (Fig. 8). Other
structural anomalies, in particular midline and
neuronal migration abnormalities, are consid-
ered minor criteria.
Hemangioma – The infantile hemangioma is the
most common tumor in infancy and typically
presents within the first few weeks of life, has a Fig. 8 Three-month-old infant with bilateral facial hem-
angiomas with right posterior fossa hemangiomas, largest
rapid proliferative phase within the first year of
at the right porus acousticus (arrow)
life, then begins spontaneous regression. As
defined in a consensus statement published in
2009, PHACE syndrome requires a facial hem- two minor criteria or may have no hemangi-
angioma >5 cm (plus one major or two minor oma if two major criteria are present. Imaging
criteria). Possible PHACE syndrome may have features of hemangioma include significant T2
only one minor criterion or may include a hyperintensity, presence of flow voids, and
chest/neck hemangioma with one major or early arterial enhancement with persistent
1046 S. Milla et al.

Fig. 9 Six-month-old girl with bilateral facial masses. (a). homogeneous enhancement of the masses is seen on coro-
Coronal T2 fat sat image demonstrates hyperintense bilat- nal T1 fat saturated post contrast image
eral parotid masses with internal flow voids. (b). Intense

blood pool enhancement (Fig. 9). During the

proliferative phase, lesions are homogeneous
in enhancement and well defined. The presence
of an infantile cutaneous lesion with similar
intense enhancement but with infiltrative or
aggressive imaging features should question
the diagnosis of hemangioma and raise con-
cern for a kaposiform hemangioendothelioma.
Arterial cerebrovascular abnormalities –
Craniocervical arterial abnormalities are
major criteria in PHACE syndrome. Common
abnormalities include congenital absence,
excessive tortuosity, looping, kinking,
dolichoectasia, aneurysm or stenosis of a por-
tion of the carotid, vertebrobasilar, or intrace-
rebral vasculature (Fig. 10). Moyamoya
vasculopathy has also been reported. Persistent
embryonic arterial vasculature is also common,
with persistent trigeminal artery being a major
criterion for PHACE. While arterial strokes
have been reported, they are rare but reflect a Fig. 10 Axial image from noncontrast time of flight MRA
significant possible sequela of the syndrome. shows significant tortuosity of the left ICA
Cardiac abnormalities and aortic coarctation –
Aortic arch anomalies are a major criterion,
with dysplasia and coarctation being the classic presence of a right-sided aortic arch are minor
appearance (Fig. 11). Aberrant subclavian ori- criteria.
gin and aortic aneurysm are also considered Eye abnormalities – Posterior segment abnormal-
major criteria. Ventricular septal defects and ities are major criteria, specifically persistent
43 Congenital Vascular Syndromes and Diseases
Fig. 11 3D MRA reconstruction (posterior view) demon-
strating aortic dysplasia, kinking, and coarctation in an
infant with PHACES
hyperplastic primary vitreous (PHPV), retinal
vascular anomalies, morning glory disc anom-
aly, optic nerve hypoplasia, peripapillary
staphyloma, and colobomas. Anterior segment
abnormalities are minor criteria, specifically
cataract, anterior coloboma, sclerocornea, and
microphthalmia.
TREATMENT: Multiple therapeutic options
can be utilized for the management of infantile
hemangiomas, including steroids, surgical resec-
tion, laser treatments, and most recently the use of
beta blockers, in particular propranolol. Heman-
gioma therapy may be related to size and location
of the mass. Additional treatments are related to
the individual patient’s manifestations of PHACE
syndrome, particularly cardiovascular abnormali-
ties. While the cerebrovascular abnormalities are
well-defined in the literature, the long-term course
and necessity for treatment and management of
intracranial arteriopathies are being evaluated and
1047
are typically assessed by individual case review.
Recommendations for routine follow-up imaging
relate to severity of arteriopathy and concern for
progression of disease and potential need for
treatment.
AVM/AVF Syndromes
In reference to the terminology of vascular
malformations, reference to ISSVA classification
is recommended. In general, arteriovenous
malformations are considered congenital disor-
ders, with underlying etiology of the malforma-
tion due to abnormal developmental
communications of arteries and veins. Classifica-
tion and discussion of AVMs and AVFs are
covered more extensively in Chapter “▶ Brain
Arteriovenous Malformations”.
Many children may have CNS abnormalities in
association with their syndrome/disease or as
reportable variant anomalies. The syndromes
below are highlighted because of their primary
neurovascular manifestation.
Heriditary Hemorrhagic Telangiectasia
Heriditary hemorrhagic telangiectasia (HHT),
also known as Rendu-Osler-Weber or Osler-
Weber-Rendu disease, is an autosomal dominant
condition that manifests as mucocutaneous telan-
giectasias and visceral arteriovenous
malformations. Central nervous system abnormal-
ities in HHT are less common than involvement of
the body. Arteriovenous malformations and cav-
ernous malformations are the most common CNS
manifestations (Fig. 12). Dural arteriovenous fis-
tulas (AVFs) may be present. Commonly, CNS
findings may relate to sequela of systemic
AVMs, particularly embolic strokes and abscesses
(Fig. 13).
Gene mutations have been identified, allowing
subclassification of HHT. The most frequent
mutation is HHT type 1 (HHT1) caused by muta-
tion in ENG gene (chromosome 9q34). HHT2 has
been mapped to ALK1 gene (12q), HHT3 (5q31),
AND HHT4 (7p14), and juvenile polyposis/HHT
1048
Fig. 12 Small AVM in a child with HHT. (a). Axial SWI
shows blooming artifact in right frontal lobe. (b). Post
contrast T1 weighted image shows associated focal
enhancement. (c). Arterial phase injection of right ICA
syndrome has been linked to mutation in SMAD4
gene (18q21).
HHT patients often present in childhood, with
clinically significant epistaxis (most common),
evaluation of mucosal or skin telangiectasias, GI
bleeding, or visceral angiodysplasias (Fig. 14).
Screening for cerebral AVMs is often performed
in patients with genetic or clinical diagnosis of
HHT. Neurologic symptoms are occasionally the
presentation of HHT, typically from embolic
S. Milla et al.
showing the right frontal AVM (white arrow). (d).
Draining vein of AVM seen on subsequent image (black
arrow)
strokes or abscesses from pulmonary AVMs,
which are a more typical manifestation. Cerebral
AVMs are less common, but are more typical in
ENG mutations than other known mutations.
Imaging for HHT patients typically includes
workup and screening with MRI and MRA. Addi-
tion of susceptibility weighted imaging (SWI) or
gradient echo sequences will maximize sensitivity
to blood products. CTA is a useful tool, however,
in the era of limiting pediatric exposure to
43 Congenital Vascular Syndromes and Diseases
Fig. 13 DWI shows a punctate embolic stroke in a teen-
ager with HHT
Fig. 14 Same patient as Fig. 13. Teenager with HHT.
Coronal CT image demonstrates right-sided pulmonary
AVM and angiodysplasia within the spleen. Hemorrhagic
fluid is seen in the abdomen
1049
Fig. 15 Left vertebral injection in a teenage HHT patient
showing a left thalamic AVM supplied by a PCA branch
(arrow). This was not identified on conventional
MRI/MRA
ionizing radiation (As Low As Reasonably
Achievable- ALARA principle), the use of MRA
has increased. There are occasions when small
AVMs are not well seen on MRA and may only
be seen on digital subtraction angiography
(Fig. 15), which is still considered the gold stan-
dard for diagnosis and also allows potential for
treatment.
Wyburn-Mason
Wyburn-Mason syndrome (also known as
Bonnet-Dechaume-Blanc syndrome or retinoen-
cephalofacial angiomatosis) is classically
described as the presence of an AVM of the visual
pathway (typically retinal or orbital AVM), ipsi-
lateral midbrain AVM and often an ipsilateral
facial angioma. Wyburn-Mason studied patients
with retinal AVMs and found a strong association
with concomitant intracranial AVMs. The current
classification does not necessitate all three find-
ings of the classical description. The embryology
is thought to relate to a fetal abnormality of the
1050
primitive mesoderm before 7 weeks gestational
age causing persistence of early fetal vascular
tissues shared by the developing optic cup and
anterior neural tube. Abnormal vascular develop-
ment leads to anomalous communications
between arteries and veins, thus AVM formation.
While midbrain AVMs are classically described,
hypothalamic/thalamic and basal ganglia AVMs
are also common. The ipsilateral facial angioma is
only seen in 30–50 % of patients classified with
Wyburn-Mason.
Initial presentation may be related to ocular
and orbital involvement, including decreased
visual acuity, proptosis, afferent pupillary defect,
or pale optic disk. Retinal AVMs are the most
common orbital finding in published cases and
reportedly are more stable than intracranial
AVMs. Neurologic symptoms may be mild to
severe, ranging from headaches and retroorbital
pain to hemiparesis and seizures. Treatment
options may be reduced to embolization given
location of central CNS AVMs.
Vein of Galen Malformation
The vein of Galen malformation (VOGM) forms
in early fetal life and involves abnormal arterial
communication to the embryonic prosencephalic
vein of Markowski, which is intimately associated
with the development of the vein of Galen, thus
the naming of the disorder. Embryologic study
correlates this time period of choroid and mid-
brain arterial development with the prosence-
phalic vein of Markowski, explaining the typical
feeding arterial supply involving the anterior cere-
bral artery branches, anterior and posterior cho-
roidal arteries (ACHA and PCHA), and the dorsal
midbrain arterial plexus. The timing of early
embryologic venous development also correlates
to early fetal life, as most patients with VOGM
will have drainage from the dilated vein into a
persistent falcine sinus, with absence of or a
diminutive straight sinus.
Historically, two classifications have been uti-
lized to describe the characteristics of individual
anatomic arterial to venous connections. The
Yasargil classification is based on the location of
S. Milla et al.
the arterial feeders and type of arterial to venous
communication. Type I–III are AV fistulas (Type
I: Small pure cisternal fistula between pericallosal
arteries or posterior cerebral artery and vein; Type
II: multiple fistulas between thalamoperforating
vessels and vein; Type III: high flow, mixed type
I and II) and Type IV being a true arteriovenous
malformation with a nidus and dominant draining
vein. The Lasjaunias classification describes cho-
roidal and mural subtypes. The choroidal classifi-
cation accounts for the dominant percentage of
cases. Choroidal type has numerous feeding arter-
ies including choroidal, pericallosal, and
thalamoperforating arteries, which communicate
with the anterior aspect of the embryonic median
prosencephalic vein of Markowski. These patients
typically present as neonates with shunt physiol-
ogy and congestive heart failure because of
high flow dynamics without outflow restriction.
In contradistinction, the mural subtype typically
has fewer arterial feeders and is associated with
outflow restriction associated with stenosis of
dural sinuses and/or at the level of the jugular
foramen. If not prenatally diagnosed,
these patients can present later in infancy, often
with hydrocephalus, developmental delay, or
seizures. Hypoxic ischemic injury to brain
parenchyma can occur due to shunting of
oxygenated blood away from tissue into venous
drainage.
Due to prevalence of fetal sonography and fetal
MRI, many VOGM cases are prenatally diag-
nosed. Sonographic imaging demonstrates a
hypoechoic rounded structure in the expected
location of the vein of Galen which demonstrates
avid color flow on Doppler imaging, confirming a
dilated vascular structure. Waveform interroga-
tion shows a dominant venous waveform, which
may be arterialized. Sagittal midline Doppler
imaging typically demonstrates the varix and
often the associated persistant falcine sinus
(Fig. 16). Fetal MRI better demonstrates the arte-
rial supply and also better determines parenchy-
mal damage that may be present (Fig. 17).
Postnatal imaging with sonography and MRI
with MRA/MRV are routinely performed. Direct
angiography and interventional treatment are typ-
ically necessary in VOGM, occasionally
43 Congenital Vascular Syndromes and Diseases 1051

Fig. 16 (a) Sagittal

Doppler on fetal ultrasound
and (b) sagittal and (c) axial
fetal brain MRI shows large
varix in the expected
location of the vein of Galen

Fig. 17 Postnatal MRI in newborn with vein of Galen malformation showing (a) varix, (b) arterial feeders, and (c) MRV
showing persistent falcine sinus (arrow)

necessitating urgent intervention in the neonatal Sturge-Weber Syndrome

period in the “choroidal” type, if congestive heart
failure and shunt physiology do not respond to Sturge-Weber syndrome (SWS), also known as
medical management. encephalotrigeminal angiomatosis, is a
1052
neurocutaneous syndrome classically involving a
facial port-wine stain (PWS, also known as nevus
flammeus) associated with ipsilateral
leptomeningeal angiomatosis and often ipsilateral
glaucoma from ocular choroid involvement. This
classic triad is considered Type I utilizing the
Roach Scale, with Type II being facial PWS and
often glaucoma without CNS involvement. Type
III consists of intracranial leptomeningeal
angiomatosis without facial or orbital involve-
ment. Recent studies suggest SWS and
nonsyndromic PWS correlation with a somatic
mosaic mutation of the Guanine Nucleotide Bind-
ing Protein, Q polypeptide (GNAQ) gene on
chromosome 9q21.
Intracranial manifestation of Sturge-Weber
syndrome is thought to be related to the lack of
normal development of cortical draining veins
and proliferation of pial capillaries and small
venous channels during early fetal life. This is
postulated to be embryologic in etiology due to
similar timing of the development of the ectoderm
and the vascular development of the related por-
tion of the neural tube at that time. The term
encephalotrigeminal is due to the suggested cor-
relation of trigeminal nerve distribution of the
facial port-wine stain with intracranial involve-
ment. Correlation has been suggested of V1 (oph-
thalmic branch) facial distribution with ipsilateral
occipital lobe involvement, V2 (maxillary branch)
with parietal lobe involvement, and V3 (mandib-
ular branch) with frontal lobe involvement. How-
ever, there has been recent research suggesting
that the distribution of the port-wine stain may
not be related to trigeminal nerve development
but rather the intrinsic pattern of facial vascular
development. Multisegment facial port-wine stain
with ipsilateral lobar involvement is commonly
seen in more extensive intracranial cases. Approx-
imately 10–20 % of patients have bilateral facial
and intracranial involvement. Occipital-parietal
involvement is most common.
The diagnosis of a facial port wine stain is
typically made after birth, and appropriately, par-
ents are counseled about the possibility of SWS.
While a facial PWS is common, reportedly up to
1 in 300 patients, approximately 8 % of patients
with facial PWS will have SWS. Forehead
S. Milla et al.
location, larger areas of involvement, and/or bilat-
eral PWS have suggested higher risks for SWS,
reportedly from 20 % to 70 % depending on the
study population. Most patients with SWS will
have infantile spasms which progress to seizure
disorder. Between 75 % and 100 % of classic
SWS patients will have seizures, with more exten-
sive and bilateral disease portending a worse
prognosis. Developmental delay and migraine-
like headaches are also very common in SWS
patients. More recently, endocrine problems,
such as central hypothyroidism and growth hor-
mone deficiency, have been reported as well.
Contrast-enhanced MRI is now the standard of
care study to initially evaluate for the presence of
intracranial abnormalities in SWS. Intracranial
findings reflect the underlying pathophysiology
of SWS. The lack of normal cortical veins with
the proliferation of leptomeningeal capillaries and
small veins leads to slow flow and increased
enhancement in the leptomeninges, which is best
demonstrated by postcontrast T2 FLAIR and
postcontrast T1 techniques (Fig. 18). An enlarged
and avidly enhancing choroid plexus is com-
monly seen within the ipsilateral lateral ventricle
due to similar vascular malformation (Fig. 19).
Developmental venous anomalies and dilated
medullary veins may be present as well, due to
underdevelopment of the normal cortical venous
drainage. This lack of normal cortical venous
drainage is thought to cause venous stasis and
ischemia to the subcortical white matter and
within the cortex itself. Over time, these ischemic
changes can lead to subcortical and cortical calci-
fications and parenchymal atrophy. As time pro-
gresses, these changes become more obvious.
Occasionally, involvement of the globe will be
visualized on MRI with buphthalmos or choroidal
enhancement. CT studies will demonstrate corti-
cal/subcortical calcification and atrophy in
affected areas over time (Fig. 20). These calcifi-
cations on radiography and subsequently CT have
been referred to as “gyral” or “tramtrack.” PET
studies will often show areas of reduced FDG
uptake in affected areas. In dynamic MR perfu-
sion studies, the involved areas of the brain dem-
onstrate abnormal perfusion and mean transit
times thought to be due to impaired venous
43 Congenital Vascular Syndromes and Diseases
Fig. 18 MRI in two infants with Sturge-Weber. (a).
Postcontrast T1 weighted image in an infant with bilateral
port-wine stains showing extensive abnormal bilateral
leptomeningeal enhancement with relative sparing of a
Fig. 19 MRI of an 8-month-old with right-sided port-
wine stain. (a) Gradient echo and (b) precontrast T1
shows right parietal calcification (thin arrows). (c)
drainage. MR spectroscopic findings support the
theory of ongoing ischemic change in affected
areas.
Treatment depends on clinical symptomatol-
ogy, location, and extent of disease. Treatment
with antiepileptic medication is important in the
management of patients with seizures. Aspirin has
been suggested to be helpful and its use in SWS is
1053
portion of the right frontal lobe. (b). Postcontrast FLAIR
image in an infant with right sided port-wine stain showing
ipsilateral leptomeningeal enhancement
Postcontrast T1 weighted image shows leptomeningeal
enhancement (thin arrow) as well as ipsilateral choroid
plexus enlargement (thick arrow)
being investigated. In patients with seizures and
focal SWS in a noneloquent area, surgical
resection may be performed. In patients with
more extensive unilateral lobar SWS, a hemi-
spherectomy may be performed early in child-
hood to allow the contralateral normal
hemisphere to take over functions from the
affected hemisphere.
1054
Fig. 20 (a) Soft tissue and (b) bone windows from a CT scan in a 4-year-old with bilateral port-wine stains,
demonstrating cortical and subcortical calcifications and prominent left hemiatrophy
Venous Malformations
Developmental Venous Anomalies
Developmental venous anomalies (DVAs) are
common venous malformations or perhaps ana-
tomic variations of the venous drainage system,
reported in up to 2.5 % of the population in
autopsy study. DVAs consist of a network of
small medullary veins that converge into a larger
draining vein, which then communicates with the
superficial or deep venous system. Although not
distinctly known, it is thought that they may form
due to maldevelopment of superficial or deep
local veins or possibly relate to localized variation
in capillary venous pressures. They may be seen
in patients with other vascular anomalies, such as
cerebral cavernous malformations, but are most
commonly seen without any other associated con-
dition. While DVAs themselves are thought to be
benign, cases of venous thrombosis of DVAs with
infarctions of associated parenchyma have been
described in patients with inflammatory condi-
tions, dehydration, diabetes, and pregnancy.
S. Milla et al.
DVAs are typically incidentally noted on imag-
ing. They can be well seen on contrast-enhanced
studies and have a “Medusa-like” appearance,
with contrast enhancement of the medullary
veins converging into the larger draining vein
and subsequently into the superficial or deep
venous system (Fig. 21). The imaging appearance
is classic and similar on CT and MRI when thin-
slice imaging techniques are utilized. Digital sub-
traction angiography also has a similar appear-
ance. On noncontrast studies, DVAs may be
difficult to see on CT, and may be faintly visible
on MRI, best seen on susceptibility weighted
imaging (SWI) sequences. Recent study of pedi-
atric DVAs suggests that up to 11.6 % of DVAs
not associated with cavernous malformations had
adjacent T2/FLAIR signal abnormality and 4.1 %
had parenchymal atrophy. These findings adjacent
to DVAs have also been reported in adults and
may be related to the altered hemodynamics found
adjacent to DVAs or possibly associated venous
hypertension. As mentioned previously, rarely
DVAs may thrombose, and precontrast linear T1
hyperintensity of a DVA should be concerning for
thrombosis. In these cases, diffusion weighted
43 Congenital Vascular Syndromes and Diseases 1055

Fig. 21 (a, b) Two-contrast enhanced T1 images illustrate the “medusa like” appearance of an incidental developmental
venous anomaly (arrows) in the right cerebellar hemisphere of a 19-month-old female

Fig. 22 (a) Coronal SWI and (b) coronal postcontrast T1 demonstrate heterogeneous appearance of a left frontal
cavernous malformation. (c) Sagittal postcontrast T1 illustrates associated developmental venous anomaly (arrow)

imaging should be evaluated for evidence of
ischemia.
Cavernous Malformations
Cerebral cavernous malformations (CCMs, also
known as cavernous angiomas and informally as
cavernomas) are abnormal capillary-venous com-
munications, appearing as sinusoidal-shaped
blood vessels lined with a single layer of
endothelium, and are prone to hemorrhage.
CCMs can be seen in sporadic or familial forms.
A sporadic CCM is typically solitary, small,
and is often associated with an adjacent develop-
mental venous anomaly (DVA) (Fig. 22). Familial
forms (fCCMs) have been linked to autosomal
dominant heterozygous germline mutations in
the CCM1 (approximately 50 % of familial
cases), CCM2, or CCM3 genes. These proteins
are important for control of endothelial permeabil-
ity and are also important in angiogenesis.
1056
Fig. 23 Sixteen-year-old with CCM3/PDCD10 mutation.
SWI demonstrates numerous cavernous malformations, as
evidenced by foci of hemosiderin blooming. Right frontal
artifact from prior surgeries
Patients with fCCM demonstrate multiple cavern-
ous malformations with the number of lesions
increasing over time. These patients can have
hundreds of lesions (Fig. 23). It is thought that
fCCM lesions may not have the same association
with DVAs as with sporadic CCM lesions. A
history of craniospinal radiation from CNS
tumor also may increase the likelihood of devel-
oping multiple cavernous malformations.
Sporadic CCM patients often present after an
episode of hemorrhage, with an acute neurologic
symptom such as seizure or loss of consciousness.
Occasionally, these patients are diagnosed
because of a history of headaches, either poten-
tially related to prior hemorrhage or incidentally
found. fCCM patients may present symptomati-
cally or may have known family history with
lesions found on screening. Cases of de novo
mutations of fCCM genes, as well as children of
asymptomatic carriers, have been documented. If
more than one CCM is found in a patient, there is
approximately an 80 % chance of fCCM gene
mutation.
As the presentation may be due to acute neu-
rologic event, patients may have a noncontrast CT
S. Milla et al.
scan that may show acute parenchymal and/or
intraventricular hemorrhage if the CCM has
bled. In this acute scenario, further imaging is
typically performed with CT angiogram to deter-
mine if the hemorrhage is due to an AVM or a
cavernous malformation, as they are the most
common causes of acute hemorrhage in children
in the absence of trauma. In the absence of acute
hemorrhage, noncontrast CT may detect an inci-
dental well-defined blood density lesion with or
without calcification. Imaging finding such as
lack of edema and stability over time mitigate
against acute hemorrhage. CTA may show an
associated developmental venous anomaly with-
out any additional suggestion of abnormal arterial
structures, which may help make the diagnosis.
MRI is the most specific imaging modality to
diagnose CCM with the imaging appearance
depending on the stage of hemorrhage associated
with the lesion, e.g., hyperintensity on T1
weighted imaging suggests subacute hemorrhage.
Gradient echo and susceptibility weighted imag-
ing (SWI) sequences highlight blooming artifacts
from hemorrhage which are classic to the diagno-
sis. Similarly, a T2 hypointense peripheral ring
around the lesion reflects deposition from hemo-
siderin from remote hemorrhage. Internal lesional
T2 hypointensity and hyperintensity may come
from various states of hemoglobin breakdown.
Imaging appearance on T1, T2, and FLAIR have
often been described as “popcorn-like” due to the
heterogeneous appearance in the central core of
the lesion (Fig. 24). FLAIR signal abnormality is
commonly seen surrounding a CCM after acute or
subacute hemorrhage, which typically resolves
over time. After contrast administration, many
CCMs do not demonstrate enhancement, while
others may demonstrate punctate areas of
enhancement, possibly due to internal capillary
telangiectasia within the CCM. Enhancement of
an associated DVA is again helpful to make the
diagnosis of CCM. In patients with fCCM, lesions
demonstrate typical CCM characteristics, but over
time, new lesions typically appear. Lesions are
seen throughout the supra and infratentorial
brain. Brainstem and spinal lesions are most wor-
risome, as they tend to have significant morbidity
after hemorrhage. While spinal cavernous
43 Congenital Vascular Syndromes and Diseases
Fig. 24 MRI in 16-year-old with CCM3 mutation. (a)
Axial T1 image demonstrates multiple foci of T1
hyperintensity representing hemorrhage within cavernous
malformations may occur with similar imaging
characteristics, they are less common than
CCMs, even in fCCM patients.
In patients with solitary cavernous
malformations, surgical resection is often
entertained, particularly in the setting of recurrent
hemorrhage. In fCCM families, surgery is often
not considered prophylactically due to number of
lesions, but may be performed as indicated after
evidence of hemorrhage.
Mitochondrial Myopathy,
Encephalopathy, Lactic Acidosis,
and Stroke-Like Symptoms (MELAS)
MELAS is a rare mitochondrial disorder, which
can be caused by mutation in one of many genes,
with approximately 80 % of cases due to mutation
in the MT-TL1 gene. Other involved genes may
include MT-ND1, MT-ND5, MT-TH, and MT-TV.
These mutations impair the ability of mitochon-
dria to make proteins, use oxygen, and produce
energy, yet it is still unclear exactly how these
mutations cause strokes. As this relates to mito-
chondrial DNA, the inheritance pattern is
1057
malformations. (b) T2 weighted image demonstrates “pop-
corn” heterogeneity and shape of the cavernous
malformations
maternal. Manifestation is variable, and clinical
presentation often includes severe headache, nau-
sea/vomiting, and neurologic symptoms (“stroke-
like” episodes) such as ataxia, hemianopsia,
hemiparesis, and cortical deafness. Presentation
is often in the second decade of life, and paired
clinical presentation and imaging appearance
should suggest the diagnosis. Blood and CSF
lactate levels are elevated in the acute presenta-
tion. Confirmation of the diagnosis with muscle
biopsy, looking for ragged-red fibers, COX nega-
tive fibers, and mitochondrial inclusions with sub-
sequent genetic testing is routine.
While the term “stroke-like” is utilized, imag-
ing characteristics are fairly similar to typical
stroke imaging appearance. The biggest differ-
ence between the appearance in MELAS and clas-
sic stroke findings is the lack of distinct vascular
territory correlating with the ischemic area. The
most common locations for ischemia in MELAS
are parietal and occipital cortex, however any lobe
can be affected and grey matter involvement can
be seen. CT may show loss of grey white differ-
entiation in area of cortical involvement (Fig. 25).
Basal ganglia calcification may be seen. In the
acute phase, cortical edema on MRI is
1058
Fig. 25 Ten-year-old with seizure. Genetic testing con-
firmed MELAS. (a) Axial CT with hyperdensity
representing abnormal mineralization in globi palladi and
subtle loss of grey white differentiation in the left cerebral
Fig. 26 Eight-year-old female with abnormal vision and
headache with subacute right temporal occipital and acute
left occipital infarcts. (a) Axial T2 weighted image show-
ing cortical and white matter hyperintensity with volume
demonstrated with T2/FLAIR hyperintensity and
reduced diffusion noted on DWI, consistent with
cytotoxic edema (Fig. 26). Some studies report
T2/FLAIR abnormalities with associated
increased diffusion on DWI, which suggests
vasogenic edema in some cases. In serial
S. Milla et al.
hemisphere posteriorly concerning for stroke. (b) Axial
FLAIR and (c) DWI demonstrate abnormal signal and
reduced diffusion in the left occipital and posterior tempo-
ral lobes consistent with acute stroke
loss in right temporal and occipital lobes and loss of grey
white differentiation in left occipital lobe. (b) DWI dem-
onstrated restricted diffusion consistent with acute ische-
mia in left occipital lobe
examinations, there can be evolution of previ-
ously seen findings with normalization of signal
and/or increased cortical T1 hyperintensity with
development of cortical atrophy. Additionally, on
serial imaging, new areas may develop acute
T2/FLAIR and DWI abnormalities. MRAs in
43 Congenital Vascular Syndromes and Diseases
these patients are normal. MR Spectroscopy
shows lactate peaks in areas of ischemia, but
may also show lactate in areas that may not have
any other abnormality on conventional imaging,
including over CSF containing spaces. These
parenchymal areas that appear normal on conven-
tional T2/FLAIR and DWI yet have abnormal
lactate on MRS may indicate areas which will
later develop abnormal MRI findings. This obser-
vation on MRS and the lack of vascular territories
corresponding to the ischemic areas should raise
the diagnosis of MELAS on initial presentation
and guide clinicians to clinical diagnosis by lac-
tate levels, muscle biopsy, and genetic workup.
Reports of improvement after treatment with
L-arginine, an amino acid involved in endothelial-
dependent vascular relaxation, have been
published. Additional supportive therapies
including vitamin regimens, coenzyme CoQ10,
and riboflavin have been utilized; however, no
consensus treatment regimen has currently been
identified.
Summary
• Many congenital neurovascular abnormalities
present during childhood.
• The spectrum of congenital neurovascular
abnormalities can often be categorized into
arterial or venous disorders.
• There are known genetic mutations that cause
many of the congenital neurovascular
syndromes.
• Moyamoya disease is idiopathic, while
moyamoya syndrome is associated with a
known underlying disease or syndrome. The
“ivy sign” or watershed ischemic changes
should prompt evaluation of ICA/MCA T2
flow voids.
• Infants with hemangiomas of the head or neck
should be evaluated for PHACES association
with MRI/MRA.
• Most AVMs and AVFs are sporadic; however,
there are syndromes that should be considered,
particularly in the presence of other
angiodysplasias of other organs.
1059
• Vein of Galen malformations are typically AV
fistula with complications of hypoxic injury
due to a steal phenomenon.
• Most patients with port-wine stains do not have
intracranial abnormalities; however, close
attention to the postcontrast T1 and
postcontrast T2 FLAIR sequences are useful
in detecting the pial angiomatosis classically
seen in Sturge-Weber syndrome.
• Cavernous malformations are commonly asso-
ciated with developmental venous abnormali-
ties. SWI sequence is optimal in the detection
of cavernous malformations.
• MELAS is a genetic syndrome that causes
strokes or stroke like symptoms not distinctly
in typical arterial distributions. Lactate peaks
in CSF or in normal appearing parenchyma
may be helpful in making the diagnosis.
References
1. Metry D et al. (2009) Consensus statement on diagnos-
tic criteria for PHACE syndrome. Pediatrics
124:1447–1456
2. Hess CP, Fullerton HJ, Metry DW, Drolet BA, Siegel
DH, Auguste KI, Gupta N, Haggstrom AN, Dowd CF,
Frieden IJ, Barkovich AJ (2010) Cervical and intracra-
nial arterial anomalies in 70 patients with PHACE
syndrome. AJNR Am J Neuroradiol 31(10):
1980–1986
3. Michelle L, Bayer et al. (2013) Congenital cardiac,
aortic arch, and vascular bed anomalies in PHACE
syndrome (from the International PHACE Syndrome
Registry). Am J Cardiol 112(12):1948–1952
4. Dayani PN, Sadun AA (2007) A case report of
Wyburn-Mason syndrome and review of the literature.
Neuroradiology 49(5):445–456
5. Guey S, Tournier-Lasserve E, Hervé D, Kossorotoff M
(2015) Moyamoya disease and syndromes: from genet-
ics to clinical management. Appl Clin Genet 8:49–68
6. Beth A, Drolet et al. (2013) Initiation and use of pro-
pranolol for infantile hemangioma: report of a consen-
sus conference. Pediatrics 131(1):128–140
7. Ito H, Mori K, Kagami S (2011) Neuroimaging of
stroke-like episodes in MELAS. Brain Dev
33:283–288
8. Linscott LL, Leach JL, Zhang B, Jones BV (2014)
Brain parenchymal signal abnormalities associated
with developmental venous anomalies in children and
young adults. AJNR Am J Neuroradiol 35(8):
1600–1607
9. Meyers PM, Halbach VV, Barkovich AJ (2012) Anom-
alies of cerebral vasculature: diagnostic and
1060
endovascular considerations. In: Barkovich AJ,
Raybaud C (eds) Pediatric neuroimaging, 5th edn.
Lippincott Williams & Wilkins, Philadelphia
10. Nozaki T, Nosaka S, Miyazaki O, Makidono A,
Yamamoto A, Niwa T, Tsutsumi Y, Aida N,
Masaki H, Saida Y (2013) Syndromes associated with
vascular tumors and malformations: a pictorial review.
Radiographics 33(1):175–195
11. Pascual-Castroviejo I (1978) Vascular and nonvascular
intracranial malformations associated with external
capillary hemangiomas. Neuroradiology 16:82–84
12. Raybaud C (2010) Normal and abnormal embryology
and development of the intracranial vascular system.
Neurosurg Clin N Am 21:399–426
13. Smith ER (2012) Moyamoya arteriopathy. Curr Treat
Options Neurol 14(6):549–556
S. Milla et al.
14. Smith ER, Scott RM (2010) Moyamoya: epidemiol-
ogy, presentation, and diagnosis. Neurosurg Clin N Am
21(3):543–551
15. Sudarsanam A, Ardern-Holmes SL (2014) Sturge-
Weber syndrome: from the past to the present. Eur J
Paediatr Neurol 18(3):257–266
16. Vanaman MJ, Hervey-Jumper SL, Maher CO (2010)
Pediatric and inherited neurovascular diseases.
Neurosurg Clin N Am 21(3):427–441
17. Waelchli R, Aylett SE, Robinson K, Chong WK,
Martinez AE, Kinsler VA (2014) New vascular
classification of port-wine stains: improving
prediction of Sturge-Weber risk. Br J Dermatol
171(4):861–867
18. www.OMIM.org (Online Mendelian Inheritance in
Man)
Part IX
Spine
 Spinal Vascular Imaging: Technique
44
Maria Isabel Vargas, Fabrice Bing, Joanna Gariani,
and Jean-Louis Dietemann

Contents Preoperative Vascular Imaging

of Spinal Vascular Tumors and Malformations . . . . 1074
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1064 Abdominal Aortic Aneurysm, Thoracoabdominal
Vascular Anatomy of the Spine . . . . . . . . . . . . . . . . . . 1064 Aortic Aneurysm, and Aortic Dissection:
Arterial Blood Supply . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1064 Does Preoperative Imaging of the SC
Venous Drainage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1066 Blood Supply Present an Interest? . . . . . . . . . . . . . . . . . 1075

Arteriography and CT Angiography Magnetic Resonance Imaging (MRI)

of the Vasculature of the SC: Techniques,
Results, Advantages, and Disadvantages . . . . . . . . 1067
Arteriography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1067
Angio-CT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1069
Spine and Spinal Cord Vascularization
Imaging: Main Indications . . . . . . . . . . . . . . . . . . . . . . . 1071
Spinal Surgery and Percutaneous
Spinal Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1074
and Magnetic Resonance
Angiography (MRA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1078
Clinical Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1078
Technical Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1079
Patterns to Know in Vascular Lesions . . . . . . . . . . . . . . 1080
Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1083
Imaging of Spinal Vascular Malformations . . . . . . . . 1086
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1089
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1089

M.I. Vargas (*)

Department of Neuroradiology, Geneva University
Hospitals, Geneva, Switzerland
e-mail: maria.i.vargas@hcuge.ch
F. Bing • J.-L. Dietemann
Department of Radiology, University Hospital of
Strasbourg, Strasbourg, France
e-mail: fabrice.bing@chru-strasbourg.fr;
jean-louis.dietemann@chru-strasbourg.fr
J. Gariani
Department of Radiology, Geneva University Hospitals,
Geneva, Switzerland
e-mail: joanna.gariani@hcuge.ch

# Springer Science+Business Media New York 2016 1063

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_13
1064
Abstract
In this chapter, the principal imaging tech-
niques of vascular pathologies of the spine
and spinal cord are described, including com-
puted tomography angiography (CTA), mag-
netic resonance imaging (MRI), magnetic
resonance angiography (MRA), and digital
subtraction angiography (DSA) as well as the
technical protocols, the different vascular pat-
terns to recognize, and the clinical indications
where these techniques can be applied.
MRI is the technique of choice for the first-
line approach of vascular lesions and their
follow-up. It is the only non-radiating tech-
nique that allows visualization of the spinal
cord and spinal vessels. It also permits the
differential diagnosis with other nonvascular
diseases. Currently, DSA is used mainly for
therapeutic purposes and for detailed analysis
particularly in small vascular malformations.
Finally, the advantages of each technique will
be illustrated with clinical examples.
Keywords
Spine • Spinal cord • Technical MR • Angio-
MR • CT • Angiography • DTI • Diffusion •
Tumors • Arteriovenous malformations •
Ischemia
Introduction
The visualization of vessels supplying the spinal
cord requires high spatial and temporal resolution
imaging. For that reason and due to the complex-
ity of the spinal cord vasculature, spinal digital
subtraction angiography (DSA) is still considered
the gold standard for evaluating spinal arteries.
However, the improvement in technology of non-
invasive imaging modalities such as computed
tomography angiography (CTA) and magnetic
resonance angiography (MRA) allows examina-
tion of spinal vessels with good results in the
reported series.
Currently, magnetic resonance imaging (MRI)
is the first-line study for the diagnosis of vascular
anomalies in the spine and spinal cord because it
M.I. Vargas et al.
allows the visualization of vessels and the analysis
of the spinal cord and other structures such as
bones, meninges, etc.
Performing vascular MRI techniques at the
level of spinal structures is more challenging
than at the level of the brain due to the small size
of the anatomical structures of the spinal cord and
spinal vessels.
Firstly in this chapter, the anatomy of the spinal
vasculature is described. Secondly, different tech-
niques such as DSA, CTA, and MRI are presented
with their advantages and disadvantages for the
exploration of the spinal vessels. Thirdly, the main
indications of vascular spine imaging are discussed
as well as the different imaging patterns and a brief
description of the principal vascular lesions.
Finally, the treatment of spinal vascular
malformations requires imaging of very high
quality for the exact determination of the lesion
anatomy, the systematic imaging of the main
arteries supplying the spinal cord before surgical
or endovascular repair of aortic disease, or for
percutaneous spinal procedures and surgery of
the spine.
Vascular Anatomy of the Spine
Arterial Blood Supply
The blood supply of the spinal cord (SC) is
formed by longitudinal vessels extending along
the axis of the SC from the foramen magnum to
the conus medullaris. Three longitudinal arteries
are described: one anterior spinal artery running in
the anterior median fissure of the SC and two
posterior (or posterolateral) spinal arteries,
located on the posterolateral surface of the
SC. The anterior spinal artery supplies the two
anterior thirds of the SC, and the posterior spinal
artery supplies the posterior third. At the surface
of the SC, a transversal pial plexus provides a
communication between these main longitudinal
axes. The intrinsic SC vascular system is formed
by a central system fed by the sulcal arteries
(centrifugal system) and a peripheral system
forming a pial network, giving perforating arteries
(centripetal system) [1, 2].
44 Spinal Vascular Imaging: Technique
Table 1 Arterial anatomy of the SC
Name of the artery [2] Vascularization
Radicular Nerve root, dura mater
(or radiculomeningeal)
arteries (0.2–0.8 mm)
Anterior Join the ASA (0.2–0.5
radiculomedullary arteries mm) and supply the two
(6–8) and the great anterior thirds of the SC
radiculomedullary artery
(0.5–1.0 mm)
Posterior Join the posterior and
radiculomedullary arteries posterolateral spinal
(11–16) (0.1–0.4 mm) arteries and supply the
posterior third of the SC
Along their course, spinal arteries receive con-
tributions from radiculomedullary arteries, which
arise from segmental arteries (intercostal and lum-
bar arteries) in the majority of cases. Segmental
arteries originate from the posterior wall of the
aorta. They can present an ascending portion,
usually described at the thoracic level, and hence
vascularize a higher metamer: in that case, the
name of the artery is the name of the level
vascularized, not the level of its origin. At the
level of the transverse process, segmental arteries
divide into a ventroparietal branch (which sup-
plies the transverse process and the rib) and a
dorso-spinal branch. At the level of the neural
foramen, the dorso-spinal branch divides into a
muscular branch (which supplies the paraspinal
muscles) and a radicular artery, dividing into ante-
rior and posterior branches when entering the
spinal canal. At each level, radicular arteries
(also called radiculomeningeal arteries) supply
the bony walls of the spinal canal.
Concerning the radicular supply and the super-
ficial SC arteries, Lasjaunias classification differ-
entiates three types of spinal radicular arteries:
radicular, radiculopial, and radiculomedullary
arteries [3]. Thron [2] proposed a different ana-
tomical classification of the spinal radicular arter-
ies (Table 1). Radicular arteries giving branches
for the spinal arteries are called radiculomedullary
arteries: 6–8 anterior and 11–16 posterior
radiculomedullary arteries are usually described.
The branches of the radiculomedullary artery
divide into ascending and descending branches
1065
at the contact of the SC, to join the anterior and
posterior spinal arteries.
The junction of the anterior radiculomedullary
artery with the ASA forms the characteristic hair-
pin configuration, located on the midline, identi-
fied on high-resolution DSA or CTA coronal
reconstruction. The posterior radiculomedullary
artery also gives this hairpin turn, but the junction
with the PSA is off the midline.
The metameric supplementation of the ASA
and PSA is not regular. Cervicothoracic,
midthoracic, and thoracolumbar regions present
anatomic variations of the SC vasculature
(Table 2). At the upper thoracic level, a dominant
radiculomedullary artery, described by von Haller
in 1754, exists most of the time. This artery was
observed in 86 % of patients with conventional
angiography [4].
The great radiculomedullary artery (GRA) or
Adamkiewicz artery, described by Adamkiewicz,
originates from the segmental (intercostal or lum-
bar) arteries [5]. The course and morphology of
the GRA are described as stable: in 70 % of cases,
the GRA originates on the left side, frequently at
the T8-L1 vertebral level [6]. Adamkiewicz
described three anatomical types of the spinal
vascular system [5]. Firstly, the ASA presents an
uninterrupted course from the cervical portion to
the lumbar level, with supplementation coming
from intercostal arteries and the GRA. The second
system presents the same configuration, but the
ASA is poorly developed. In the third one, the
ASA is interrupted, mostly at the thoracic level.
The second and third systems present the highest
risk of SC ischemia after clamping radicular arter-
ies and consequently the GRA. Biglioli
et al. reported an anatomic study on 51 adult
cadavers without aortic disease, showing continu-
ity between the vertebral artery, the anterior spinal
artery, and the GRA [7]. This result supports that
the sacrifice of segmental arteries during surgery
or endovascular treatment of aortic disease may be
justified without the risk of SC injury [8, 9]. How-
ever, this anatomic continuity may become a dan-
gerous stealing pathway for the SC, explaining
paraplegia occurring after surgery when
prolonged cross-clamping of the aorta is
performed [7].
1066 M.I. Vargas et al.

Table 2 Main feeders of the SC blood supply at the different spine levels
Anterior feeders (for
ASA) Posterior feeders (for PSA) Main feeder and particularity
Cervical Intracranial section Vertebral artery (Fig. 2) Artery of the cervical enlargement (level
region of the vertebral C5-C6) arising from the vertebral artery
arteries (Fig. 2) PICA Radiculomedullary arteries arising from the
deep and ascending cervical arteries and the
vertebral artery
Variation: anastomoses with the occipital and
ascending pharyngeal artery (external carotid
artery). Artery of the cervical enlargement can
originate from the costocervical and thyrocervical
trunk. Duplication of the anterior spinal artery is
possible
Thoracic Supreme and Two dorsal feeders One radiculomedullary artery (level T4-T5)
region posterior intercostal entering the spine above arising from the intercostal artery
arteries and below the GRA One great anterior radiculomedullary artery
(level T9 to T12 on the left side (75 %)) (Fig. 1)
Variation: two (48 %) or three (7 %) anterior
radiculospinal artery [13]
Discontinuity of the ASA is possible
Lumbar Lumbar arteries Artery of the lumbar enlargement
region Anastomosis: at the level of the conus medullaris,
connection of the anterior and posterior spinal
arteries forming a basket shape anastomotic
network
Variation: duplication of the radiculomedullary
artery
Sacral Lateral sacral artery and iliolumbar artery from the Variation: spinal arteries caudal to the aortic
region internal iliac artery bifurcation (rare)
Medial sacral artery
GRA great radiculomedullary artery, PICA posterior inferior cerebellar artery, ASA anterior spinal artery, PSA posterior
spinal artery

Venous Drainage
Understanding the extensive anastomosing venous
channels draining the SC and the spine is manda-
tory before a surgical or endovascular approach of
spinal vascular malformations (SVM). In fact,
abnormal visibility of the spinal veins is one of
the main signs of arteriovenous fistula and can be
identified either on CTA or DSA. Dilatation of the
veins will motivate the radiologist to find the path-
ological communication between the artery and the
vein: for that purpose, DSA remains the most sen-
sible imaging modality.
One of the major characteristics of the spinal
venous system is its variability, described as being
greater than that of the arterial system [1]. The
venous system is composed of an intrinsic and
extrinsic system:
– The intrinsic veins are divided into sulcal and
radial veins.
– The extrinsic or superficial venous system is
composed of anterior and posterior spinal
veins.
• The anterior median spinal vein (diameter
0.4–1.5 mm) [2] runs with the ASA and
continues to the filum terminale (vein of
the filum terminale).
• One posterior median, the greatest spinal
vein, is accompanied by two posterolateral
veins, located next to the posterolateral spi-
nal arteries.
The extrinsic network system is in contact with
the spinal pia matter. Spinal veins are character-
ized by a high anatomical variability: longitudinal
44 Spinal Vascular Imaging: Technique
veins can be replaced by a secondary network.
Spinal veins drain into the anterior (numbers
8–20) [10] and posterior (numbers 5–10) [2]
radiculomedullary veins that drain into the
paravertebral and intervertebral plexuses. These
venous plexuses drain in the segmental veins
draining into the ascending lumbar veins, azygos
system, and the pelvic venous plexuses. At the
thoracolumbar level, the great anterior radiculo-
medullary vein (GRV; diameter 1.5–2.0 mm) may
be mistaken for the GRA, but the angle of the
junction of the GRV with the median vein is
described as more obtuse than the hairpin config-
uration of the artery [11] (Fig. 1).
Fig. 1 DSA, AP view. Fifty-year-old man, suspicion of
arteriovenous fistula. Injection into the L2 left lumbar
artery: opacification at the arterial phase (a) and venous
phase (b). (a) The great radiculomedullary artery (GRA)
(double arrow), its hairpin turn (star), and the anterior
spinal artery (ASA) (arrow) are opacified. (b) The great
radiculomedullary vein (GRV) (double arrow), the hairpin
turn (star), and the spinal vein (arrow) can present the same
aspect as the arterial distribution
1067
The radiculomedullary veins communicate
with the epidural venous plexus, into which the
intradural veins are draining. Intradural and epi-
dural veins are separated by valves at the level of
the dura. Groen et al. [12] differentiate three levels
of intercommunicating veins:
– The internal vertebral venous plexus, formed
by intradural and epidural veins, communicat-
ing with the intracranial venous system and
draining in the external vertebral veins
– The external vertebral venous plexus, located
around the vertebra
– The basivertebral veins, within the vertebra
Arteriography and CT Angiography
of the Vasculature of the SC:
Techniques, Results, Advantages,
and Disadvantages
Arteriography
Technique
Spinal DSA can be performed without general
anesthesia, usually reserved for children and
patients who cannot cooperate or lie flat. The
pain induced by a tumoral vertebral lesion has to
be evaluated during the consultation with the radi-
ologist and the anesthesiologist: in fact, a painful
lesion can motivate the sedation of the patient, as
the patient may present difficulties to remain
immobile during the image acquisition. Most of
the time, local anesthesia with intravenous
midazolam and fentanyl is enough. Drugs are
injected for the comfort of the patients, as they
will be asked to hold their breath during the pro-
cedure to decrease artifacts. After local anesthesia,
a 4 or 5 F sheet introducer is placed in the femoral
artery. Catheterization of segmental arteries is
performed from the higher to the lower level,
mostly with a Cobra C2 (Terumo) or a Simmons
1 (Terumo) catheter. When the ostium of a seg-
mental artery is catheterized, manual injection of
contrast material is performed. To catheterize the
lower ostium, the radiologist has to gently retrieve
1068
Fig. 2 DSA, lateral view (a, b) and AP view (c). Fifty-
five-year-old man presenting a renal tumor cervical spine
metastasis. Injection in the right vertebral artery. The ante-
rior spinal artery (ASA) (black arrow) takes its origin at the
level of V3, and one posterior spinal artery (PSA) (white
the catheter without turning it until he gets a new
stabilization, considering the ostia of the segmen-
tal arteries as ladder rungs. Hence, it is
recommended to begin the catheterization of seg-
mental arteries from one side before doing the
other side.
Bone relief of the spine is essential to get a
good projection of the spinal artery: the spinous
process has to project between the pedicles on the
anteroposterior (AP) view. The GRA is well rec-
ognized thanks to its hairpin configuration that
projects on the midline, at the level of the junction
with the ASA. If the ASA mostly presents as a
straight vessel situated on the midline in young
patients (Fig. 3), it can present loops in older
patients (Fig. 4). The radiologist has to be precise
on the level of the segmental artery giving rise to
the GRA. The posterolateral radiculomedullary
artery presents a smaller course with a lateral
projection. During the injection of a segmental
artery, a blush of the vertebra is clearly identified
(Fig. 3). Difficulties to catheterize segmental
arteries can be explained by an agenesia (Fig. 5)
or thrombus of the artery in the case of an ather-
omatous aorta.
Vessels that may supply the spinal arteries have
to be known and explored according to the level of
M.I. Vargas et al.
arrow) comes from V4 at the level of the posterior cere-
bellar artery (PICA) (arrow head). The star shows the
space occupied by the spinal cord, between the ASA and
the PSA
the spinal lesion (Table 3). Except for the diagno-
sis of arteriovenous fistula, which may require a
global visualization of the spine and SC blood
supply, the arterial exploration of a spinal lesion
concerns the level of the lesion and one or two
levels above and below.
Advantages and Disadvantages
DSA presents a higher spatial and temporal reso-
lution than CTA and MRA. Moreover, angiogra-
phy may be the first step before an endovascular
treatment such as embolization of an arteriove-
nous fistula or preoperative embolization of ver-
tebral tumors. Concerning the vascular disease of
the spine, DSA remains the best diagnostic imag-
ing tool for the anatomic comprehension and ther-
apeutic planification. Even if new generation CT
scanners can improve the anatomic comprehen-
sion of arteriovenous fistula, cases reporting the
feasibility of CT scanner in dynamic assessment
of spinal fistula remain anecdotal [14].
DSA presents the main disadvantage of being
invasive, time consuming, complicated, and con-
sequently associated with a risk of SC ischemia
[15]. Moreover, DSA exposes patients to poten-
tially high doses of radiation and contrast. The
risk of paraplegia after arteriography was
44 Spinal Vascular Imaging: Technique
Fig. 3 DSA, AP view. Thirty-five-year-old woman, sus-
picion of medullary arteriovenous fistula. The GRA (dou-
ble arrow) is visible after injection into the left T9
intercostal artery. The hairpin turn (star) and the ASA
(arrow) run on the midline. The intercostal artery presents
a short ascension and divides in a ventral branch and a
dorso-spinal branch. Enhancement of the vertebra body
(brace) and of the paraspinal muscles (white arrow) is
clearly identified
estimated to be 1.2 % in a series of 480 patients,
which is significant compared with the 3–5 %
occurrence of paraplegia after aortic surgery in
a surgical series [16]. However, these high rates
of complications are described for patients
suffering from dissection or aneurysm of the
aorta. Other retrospective case series reported a
minimal rate of neurologic complications and
emphasized the important role that spinal DSA
still presents for the diagnosis of spinal vascular
malformations or preoperative spine surgery
imaging (Table 4) [17, 18]. Hence, with the use
of new endovascular technologies and nonionic
1069
Fig. 4 DSA, AP view. Seventy-year-old patient, suspicion
of arteriovenous fistula. Catheterization of the left T10
intercostal artery. GRA (double arrow), the hairpin (star),
and the ASA (arrow) are clearly identified. Note the loops
of the ASA, on the contrary to case described on Fig. 3
contrast agents, it can be assumed that modern
spinal DSA carries a very low risk of neurologic
and systemic complications [18].
Angio-CT
Technique
Many reports propose protocols of injection and
technical points to visualize the segmental artery
giving rise to the GRA. Many case series describe
different times of injection, aortic triggering, and
contrast agents for different aortic diseases.
Detection rates of the GRA are variable, ranging
from 18 % to 100 % [22–29] (Table 5). This
variability may be attributable to the patient’s
hemodynamic characteristics, the type of aortic
disease explored, and technical factors such as
imaging delay, volume, and flow of contrast
medium. Authors tried to determine the optimal
1070 M.I. Vargas et al.

Fig. 5 DSA, AP view with (a) and without (b) subtrac- arrow) is opacified. This artery presents an angulation and
tion. Thirty-three-year-old woman. Preoperative imaging a hairpin configuration (star) as the ASA, but is localized
for an intercostal tumor. Injection into the left T8 intercos- off the midline (image b). The PSA is visible (arrow)
tal artery. The posterior radiculomedullary artery (double

Table 3 Arteries supplying the spine and potentially the Some authors have evaluated the interest of
SC depending on regions multidetector-row CT angiography with intra-
Upper cervical Vertebral artery arterial contrast injection (IA-CTA) to depict the
spine Ascending cervical artery GRA. Authors reported very high percentages of
Deep cervical artery visualization of the GRA and were able to precise
Occipital arterya the origin of the GRA in almost all cases [29, 32].
Cervicothoracic Vertebral artery Only one study attempted to compare the sensi-
spine Thyrocervical and costocervical tivity of CT angiography and conventional angi-
trunks
ography in the visualization of the GRA before
Supreme intercostal artery
Intercostal arteries
thoracoabdominal aorta repair for aneurysms
Bronchial arterya or dissection [32]. The CT angiography was
Lumbosacral Lumbar arteries performed after IA (IA-CTA) injection of
spine Iliolumbar artery contrast, and a conventional angiography was
Medial sacral artery done just after the CT. It was possible to visualize
Lateral sacral artery the GRA in 90 % with IA-CTA. Moreover, in
a
Artery that may supply less frequently the spine 23 % of the patients, the GRA was visualized
with the IA-CTA but not with conventional
angiography.
New algorithms such as model-based iterative
protocols for the detectability of the GRA. reconstruction (MBIR) have been proposed to
Nakyama et al. proposed a contrast medium injec- improve the quality and spatial resolution
tion using an imaging delay of 18 s after triggering [33]. Nishida et al. [34] reported a series of
and an iodine dose of 720 mg l/kg body weight 34 patients presenting different aortic diseases
[30]. Dual-phase CT angiography with a catheter with an improvement of the detectability of the
inserted into the right atrium has been evaluated GRA in comparison with CT reconstruction algo-
for preoperative visualization of GRA in patients rithms (adaptive statistical iterative reconstruction
with aortic aneurysms [31]. Results were promis- and filtered back projection). This improvement
ing, but the problem of the double dose of radia- of delineation of the ASA has been reported in a
tion and contrast medium remains (Table 5). short series of ten patients [35].
44 Spinal Vascular Imaging: Technique 1071

Table 4 Case series of spinal DSA

Number of Localization of the Complications
Study patient Indication GRA (%) (%)
Forbes 1988 96 AVM (88), tumor (8) NA 3.7 systemic
[19] 2.2 neurologic
Savader 1993 50 Thoracoabdominal aortic aneurysm NA 4.6
[20]
Heinemann 46 Aneurysm and chronic dissection 65 0
1998 [21]
Kieffer 2002 487 Aortic aneurysm 86 0.8 neurologic
[15] 0.4 renal failure
0.6 groin
complication
Williams 2004 151 Aortic aneurysm and dissection 43 0
[14]
Chen 2011 [18] 302 Spinal vascular malformation (most 97.4 1 systemic
common indication) 0.7 groin
complication
Nair 2013 [17] 228 Spinal tumor embolization NA 0

Ou et al. [36] reported a series of 40 children
(mean age, 7.5  5 years) explored with CT angi-
ography for the assessment of the GRA. The GRA
was visualized in 95 % of the children. One dis-
advantage of CT angiography is the artifacts gen-
erated by the proximity of bone structures. This
problem may be encountered and overcome with
MRI because the bone shows no signal.
Advantages and Disadvantages
CTA presents the advantage of being noninvasive
in comparison with conventional arteriography.
New generations of scanner allow very fast acqui-
sition, with an excellent spatial resolution, which
is necessary for the visualization of the vessels of
the SC. Moreover, rates of detection of the GRA
reported are extremely high, even if there is a
publication bias and no comparison with arteriog-
raphy most of the time. For aortic disease, CTA
allows a precise visualization of an aneurysm or a
dissection (extent, diameter, type of aneurysm, or
dissection). CTA is logistically more accessible
than MRI in most institutions. Moreover, the
availability of the software Osirix, dedicated to
DICOM image, allows the detection of the GRA
with good agreement in comparison to the stan-
dard workstation [24].
However, CTA involves ionizing radiation,
and the exploration of the abdominal cavity is
associated with a relatively high effective dose
of 20 mSv [37]. The larger the abdomen, the
lower the contrast-to-noise ratio. In consequence,
it is possible that the high rate of detection of the
GRA, mostly reported in the Japanese population,
is not the same with Caucasian patients.
It is difficult to find a dedicated timing for the
visualization of the GRA (Table 5). High concen-
tration of contrast is mandatory, but the difficulty
is to determine the exact timing of the acquisition,
as the enhancement of the GRA is visible for
approximately 10 s. After that very precise tem-
poral window, the image is polluted by venous
drainage, which can be mistaken for the SC arte-
rial supply. This drawback is more obvious in the
case of a spinal vascular malformation.
Spine and Spinal Cord Vascularization
Imaging: Main Indications
The knowledge of the main arteries supplying the
SC, in particular the GRA, is one of the major
points of spinal surgical planning. The aim is to
 1072

Table 5 Main case series reporting the success rates of detection of the GRA and the technical aspects of injection with CTA
GRA
Aortic luminal visualization Continuity GRA
Study N= Indication MDCT Contrast material Injection attenuation (%) and aorta
Zhao 2009 99 Thoracic aortic disease 16-row Iohexol 120–150 mL; IV 358  62 UH 52 NA
[25] 350 mgI/mL; 4–5 mL/s
Clarençon 30 Aneurysm and dissection 16-row Iodixanol 120 mL; IA NA 90 84
2013 [32] 320 mg/mL; 15 Ml/S;
Nojiri 2007 27 Aneurysm (13) and dissection 4-row Iomeprol 400; iopamidol IA 4.5–7 s after 100 90 for the detection
[29] (14) 370 or iohexol 300; start of inj of the level: 61.5
100 ml;4–5 ml/s (A); 120 ml;7 for the exact
ml/s (D) continuity
Ou 2007 40 Congenital disease of the aorta 16-row Iohexol 300; 1.5 ml/kg; 2.5–3 IV 85 UH 95 95
[36] ml/s
Uotani 2008 32 Aneurysm (21) and dissection 16-row Iopamidol 370; 100 mL;5 IV 140–160 UH IACT:94.1 87.5
[22] (11) mL/s (15) and IVCT:60 55.6
IA (17)
Kudo 2003 19 Liver disease 4-row Iohexol 350; 100 mL;5 mL/s IV NA 68 NA
[28]
Nakayama 80 Aneurysms 64-row Iohexol 540 mg/mL; 6 mL/s IV 150UH 15 s 56.3 NA
2008 [28] 18 s
Iohexol 720 mg/mL; 6 mL/s 21 s
18 s
Boll 2006 100 40 Iomeprol 400 mg/mL;80 mL; IV 150 100 100
[38] 3.5 mL/s
M.I. Vargas et al.
 44

Utsunomiya 80 Aneurysms 64 Iomeprol 300 mg/mL;100 mL; IV 250 75 50 (best protocol)

2008 [27] 5 mL/s
350 mg/mL;100 mL; 5 mL/s 80
300 mg/mL;100 mL; 3.5 mL/s 45
350 mg/mL;100 mL; 3.5 mL/s 50
Yoshioka 30 Dissection (12) 16-row Iopamidol 370 mg/mL; IV 200 83 58
2006 [27]a TAA (18) 2 mL/kg; 3.5 mL/s
Yoshioka 30 Dissection (10) 4-row Iopamidol IV 130 80 62.5
2003 [39]a TAA (20) 370 mg/mL;2.5 mL/kg; 3 Ml/s
Nishida 34 TAAA (6) 64 Iopamidol 370; 100 mL; IV 250 90b NA
2014 [34] Follow up TAAA surgical 5 mL/s
grafting (9), endovascular repair
Spinal Vascular Imaging: Technique

(15), chronic dissection (3),

Takayashu (1)
Amako 2012 110 TAA and TAAA 8 and Iopamidol 370; 120 mL; IV (right Two 100 90
[31] 16-row 8 mL/s atrium) acquisitions:
20 and 55 s
after the first
injection
Takase 2002 70 TAA (20) NA Iopamidol 300; 100 mL; IV 85 90 % 31.7
[23] AAA (16) 3.5 mL/s
Dissection (19)
Atherosclerosis (15)
a
MRA and CTA study: the depiction of the GRA is reported to be better with MRA
b
Score of 3 and 4 of the scale of depiction of the GRA proposed by the authors
1073
1074
Fig. 6 DSA, AP view with (a) and without (b) subtraction. Catheterization of the left T6 intercostal artery. A longitudinal
anastomosis coming from T6 supplies the level below (arrow). No left T5 intercostal artery was found for that patient
avoid neurologic complications, such as paraple-
gia, anesthesia, loss of bladder and bowel control,
and sexual dysfunction, described after the liga-
ture of the GRA or the segmental artery giving the
GRA. Radiculomedullary arteries have to be
clearly identified before an embolization. How-
ever, the interest of searching for a radiculo-
medullary artery before endovascular or surgical
treatment of aortic disease or even before spine
surgery continues to be a controversial subject.
Indications of such preoperative imaging are
discussed.
Spinal Surgery and Percutaneous
Spinal Procedures
Concerning the vascular preoperative imaging of
spinal tumors, selective catheterization of seg-
mental arteries feeding the tumor allows the visu-
alization of spinal arteries, which can be helpful
for surgical planification. In these cases, the aim of
DSA is to identify a radiculomedullary artery at
the level concerned by the surgery or before a
percutaneous treatment (Figs. 6 and 7). Koshino
et al. showed that the GRA does not specially
originate from big segmental artery and proposed
that all the segmental arteries have to be preserved
during surgery [6].
M.I. Vargas et al.
There is no interest to systematically search the
GRA at levels far from the operative site. How-
ever, longitudinal anastomoses allow a new blood
supply for the SC: hence, ablation of vertebras and
segmental arteries giving rise to spinal artery is of
course not necessarily associated with paraplegia.
Murakami et al. reported their experience in total
en bloc spondylectomy associated with an inter-
ruption of the GRA. Fifteen cases in which the
GRA was described on DSA and ligated did not
present alteration of their neurologic function.
They concluded that surgeons are allowed to sac-
rifice up to three pairs of segmental arteries and
that the presence of the GRA at the level of the
surgery is not a contraindication for surgery
(Fig. 7) [40].
Preoperative Vascular Imaging
of Spinal Vascular Tumors
and Malformations
Imaging of Spinal Vascular Tumors
The interest of arteriography is not only to search
for a radiculomedullary artery but also to perform
an embolization in case of hypervascular
tumors. Even if the embolization of thyroid
carcinoma [41–43] or in general [44] remains
controversial, the embolization of metastasis of
44 Spinal Vascular Imaging: Technique
Fig. 7 DSA, AP view with (a) and without (b) subtrac-
tion. Thirty-three-year-old woman presenting a
schwannoma at the level of left T10 (double star). Injection
into the left T10 intercostal artery. Because of the presence
of the GRA (double arrow) coming at the level of the
renal cell carcinoma before surgery has been
reported to be useful [45, 46]. A prospective ran-
domized controlled study is warranted to precise
the benefits of preoperative embolization for spi-
nal tumors.
Thiex et al. reported no statistical difference in
intraoperative blood loss between partial and
complete transarterial tumor embolization
[47]. Interestingly, the authors reported a poor
correlation between MRI enhancement and angio-
graphic vascularity. In consequence, spinal DSA
may bring new information on the tumor vascu-
larity which could motivate the realization of the
embolization. The interventional radiologist has
to be sure that the segmental artery providing the
tumor does not supply a radiculomedullary artery
before embolization: in that case, it is considered
as an absolute contraindication to embolization
(Fig. 8) [17, 47–51].
It is necessary to explore levels above and
below the segmental artery that have to be closed:
in fact, collateral supplies coming from the rich
anastomotic network may explain accidental
embolization of a radiculomedullary artery
(Fig. 9) [52].
1075
schwannoma, surgery was canceled, and a cryotherapy
was performed, taking precautions not to injure the inter-
costal artery. Hairpin turn (star) and ASA (arrow) are
identified. Note the mass effect on the intercostal artery
which is upraised by the lesion
Abdominal Aortic Aneurysm,
Thoracoabdominal Aortic Aneurysm,
and Aortic Dissection: Does
Preoperative Imaging of the SC
Blood Supply Present an Interest?
Spinal cord ischemia after surgical aortic repair
and endovascular treatment of aortic disease
(Fig. 10) has been largely reported [53, 54]. Para-
plegia and paraparesis rates for descending tho-
racic aneurysms and thoracoabdominal aneurysm
resection are about 1 % and between 5 % and
10 %, respectively [55]. The incidence of spinal
injury after treatment with endovascular grafts is
between 3 % and 12 % and may be due to the
occlusion of the GRA associated with a drastic
decrease of the residual collateral blood supply
[56]. This hypothesis is comforted by an article
describing more SC ischemia with the number or
the length of aortic stent grafts [57].
Aneurysm and dissection have to be distin-
guished as the repercussion of these different aor-
tic mural diseases on the SC blood supply is
different. In fact, the mural thrombus of the aneu-
rysm is associated with an occlusion of the
1076
Fig. 8 DSA, AP view with (a) and without (b) subtrac-
tion. Thirty-year-old woman presenting paravertebral
chondroblastoma. Catheterization of the left T4 intercostal
artery at the level of the lesion (vascular blush, double
star). As a radiculomedullary artery (double arrow) was
injected, no preoperative embolization was performed. (c)
segmental arteries, and collateral channels devel-
oped allowing blood supply for the SC. On the
contrary, in the case of dissection, most of the
segmental arteries remain patent, and no collateral
network develops. Hence the surgical or
endovascular occlusion of the segmental artery
giving the GRA may acutely disrupt the spinal
cord vascularization. Williams et al. reported a
series of preoperative selective intercostal arteri-
ography in 131 patients undergoing treatment for
the dissection or aneurysm of the aorta. SC ische-
mia was observed in 4.6 %, and all were observed
in patients with dissection [14].
M.I. Vargas et al.
Radiograph of resected specimen of T3, T4, and T5. The
radiculomedullary artery has been resected, and the patient
did not present postoperative spinal cord ischemia (Image
courtesy of Pr Steib, Service de Chirurgie du Rachis, CHU
de Strasbourg)
The systematic imaging exploration of the seg-
mental arteries in patients undergoing
thoracoabdominal repair can be criticized, as
there is a controversy on the necessity for the
reimplantation of the intercostal and lumbar arter-
ies [58]. Heinmann reported a series of 46 patients
treated for thoracoabdominal aneurysms and
chronic dissection and precised that preoperative
selective angiography was helpful in the
planification of the surgery [21]. When the sur-
geon knows the level of the GRA and the appro-
priate branch is within the aortic segment to be
replaced, reimplantation of the segmental artery is
44 Spinal Vascular Imaging: Technique
Fig. 9 DSA, AP view. Thirty-five-year-old woman, sus-
picion of medullary arteriovenous fistula (same patient as
Fig. 3). Injection into the right L1 intercostal artery. A
radiculomedullary artery (double arrow) is opacified
through anastomoses coming from the left T12 intercostal
artery supplying a spinal artery (arrow). This case illus-
trates the rich interconnecting longitudinal extravertebral
network between segmental arteries
indicated. When the level of the GRA is unknown
but with a risk to be included in the ill segment of
the aorta, more aggressive techniques to preserve
the blood flow in the segmental artery are
recommended [21]. Nijenhuis [59] insisted on
the necessity of preoperative visualization of the
GRA, showing that no decrease in SC function
was observed when the clamped aorta did not
concern the segmental supplier of the GRA. On
the contrary, a 32% decline in SC function was
observed if the supplier of the GRA was located
inside the cross-clamped area [59].
However, case series showing very low rates of
SC ischemia without segmental artery
reimplantation and with several sacrificed seg-
mental arteries have been published [60]. More-
over, the occlusion of the segmental artery giving
rise to the spinal artery is obviously not systemat-
ically associated with spinal ischemia: this is due
1077
to the longitudinal anastomoses between segmen-
tal arteries and the collateral circulation of the
thoracic wall allowing the perfusion of the SC
[61]. This fully efficient collateral circulation has
also been demonstrated in a postmortem study of
the spinal blood supply after the occlusion of
radicular arteries [62]. The first arterial system
described by Adamkiewicz presents a low risk of
SC ischemia, as many radiculomedullary arteries
supply an uninterrupted spinal artery.
The interest of preoperative DSA and its rela-
tionship to postoperative neurologic complica-
tions in patients operated for TAAA have been
evaluated. Savader et al. [20] reported a study
with two groups of 50 patients, one with positive
spinal artery and one without spinal artery
identified on DSA. The prevention of spinal
ischemia consists of the maintenance of SC
circulation and spinal protection. For the group
of patients with positive spinal artery, spinal
protection was performed, and the authors report
no significant difference between the two groups
in the occurrence of neurologic complications.
They concluded that DSA reduced the risk of
neurologic complications, even if the rate of
morbidity of DSA in that series was very high
(4.6 %) [20].
Feeding arteries of the SC have been explored
before and after thoracic endovascular aortic
repair (EVAR). Von Tengg-Kobligk
et al. evaluated the visibility of posterior intercos-
tal arteries, dorsal branches, and the GRA with CT
angiography, before and after EVAR [63]. The
authors showed that the majority of the segmental
arteries were not occluded after EVAR, due to
retrograde perfusion. On 17 patients followed,
one GRA was not visible after EVAR, but no
neurological events were observed. Amabile
et al. reported a series of 67 patients with lesions
of the descending thoracic aorta treated by stent
graft and showed that the length of aorta covered
was the only independent predictive factor for SC
ischemia [57]. The authors identified a threshold
of 205 mm above which the risk of SC ischemia is
predictable with a sensitivity of 80 % and a spec-
ificity of 92.5 %. Delayed SC ischemia has also
been described. Etz et al. reported this rare situa-
tion, concerning 10 on 858 patients operated for
1078
Fig. 10 Ischemia. VRCT reformatting shows an aortic aneurysm (a); unfortunately, the patient becomes paraplegic after
surgery in rapport with ischemia of the spinal cord appearing as a high signal on T2 (b) with restricted diffusion (c, d)
thoracoabdominal aneurysms, has the conse-
quence of subtle hemodynamic differences
[64]. In those cases, radiculomedullary arteries
may have been injured, but collaterals may have
allowed a fragile blood supply of the SC, until the
blood pressure decreases, revealing the insuffi-
ciency of the vascularization.
In conclusion, these case series comfort the
idea that systematic CTA or DSA exploration
before surgery or endovascular treatment of aortic
disease is not necessary. Methods to protect the
SC from ischemia have been proposed (Table 6).
This emphasizes the theory of many collaterals
rather than a single dominant radiculomedullary
artery supplying the SC [8].
Magnetic Resonance Imaging (MRI)
and Magnetic Resonance Angiography
(MRA)
The other noninvasive vascular imaging of the SC
blood supply is magnetic resonance angiography
(MRA). Nijenhuis et al. [37, 59] reported
M.I. Vargas et al.
interesting results of MRA used for the preopera-
tive localization of the GRA in thoracoabdominal
aortic aneurysm patients: comparing CTA and
MRA, the authors showed that both techniques
allowed the localization of the GRA but that MRA
was more favorable because of a higher contrast-
to-noise ratio. Moreover, the quality of the image
was independent of patient thickness. On con-
trary, Zao et al. reported that the detection rate of
the ASA and the Adamkiewicz artery (AKA) with
CTA was influenced by the vertebral mass index
and the contrast-to-noise ratio [25]. Vargas
et al. [65] showed the utility of MRI in the analysis
of vascular lesions (such as cavernomas and
hemangioblastomas), in the surgical planning of
lesions, as the first-line method of analysis of
vascular malformations and for planification of
DSA in patients with vascular malformations.
Clinical Indications
Until the last decade, the visualization of vessels
of the spine by MRI was difficult due to their
44 Spinal Vascular Imaging: Technique 1079

Table 6 Etiology of SC ischemia and methods to protect
the spinal cord from ischemia during and after surgery of
the aorta
Etiology of SC Coverage of the segmental
ischemia artery with stent graft
Clamping of the aorta and
reperfusion injury with
cytotoxic metabolites
released and flushed back in
the spinal arteries
Clamping of the aorta and
steal of the blood flow
Clamping of the aorta
resulting in an increase in
intracranial pressure and a
drop in SC perfusion,
consequence of a rise in the
perfusion of the brain
Systematic hypotension
Embolization
Internal iliac artery
circulation
Methods to protect the Identification and
SC from ischemia reimplantation of the
segmental arteries giving the
spinal arteries
Systemic hypothermia
Adjunctive pharmacologic
therapy
CSF drainage
small size. Actually, the use of high-field MR in
clinical routine imaging allows the realization of
dynamic MRA and the analysis of different
malformations.
The real advantage of 3D dynamic vascular
sequences is the capacity to noninvasively evalu-
ate the shunt location and sometimes feeding
arteries and drainage veins which is extremely
helpful to guide invasive conventional digital
angiography.
This technique is extremely useful for the fine
analysis, characterization, and classification of
different vascular lesions; currently, we can iden-
tify and analyze normal vessels of the spinal cord
such as AKA (Fig. 11) as well as pathological
arteries in the case of arteriovenous
malformations [66] and dural fistulas.
The localization of AKA is an essential data for
surgical planning of the spine and vascular
surgery, for example, in the case of the surgery
of aortic aneurysms (Fig. 10), endovascular treat-
ment in the case of a thrombus (Fig. 11), spinal
tumors (Figs. 12 and 13) localized between Th8
and L1, or surgery for scoliosis and/or kyphosis.
The difficulties that remain in some cases
include the differentiation between arteries and
veins, the visualization of small shunts, and the
visualization of normal AKA and accessory spinal
cord arteries.
In ischemia [67], spinal cord anomalies are
visible early only on diffusion imaging, but unfor-
tunately, the implicated vessel is not detectable by
MRA (Fig. 10).
On the other hand, the radiographer has to be
very familiar with the MR technique because
timing is crucial.
The reformatting of this sequence is time con-
suming but is indispensible for understanding the
vascular malformations.
Technical Protocol
The technical protocol recommended is a T2-WI
sagittal (TE 120, TR 3,800, 3.5 mm slice thick-
ness) and axial (TE 120, TR 3,800, 3 mm slice
thickness), T1-WI sagittal (TE 10, TR
600, 3.5 mm thickness), diffusion tensor imaging
(DTI) with 30 directions, sagittal CISS (construc-
tive interference steady state), FIESTA (fast imag-
ing employing steady-state acquisition) or 3D T2
SPACE sequence (sampling perfection with appli-
cation optimized contrasts using different flip
angle evolutions) (TE 5.9 ms; TR 11.9 ms, slices
72, slice thickness 700.00 μm), and fat-saturated
T1-WI in the sagittal (TE 12, TR 400, 3.5 mm
slice thickness) and axial (TE 12, TR 400, 3 mm
thickness) planes after the administration of con-
trast in the case of tumors and an angio-MR (three
MRA acquisitions) with late high-resolution
sequences are added in cases of suspicion of a
dural fistula or an arteriovenous
malformation [65].
The MRA sequence consists of three MRA
acquisitions based on a 3D contrast-enhanced
gradient echo sequence sequentially performed
after a bolus injection of an intravascular contrast
1080
Fig. 11 Visualization of AKA. Patient with extensive
thrombus of the aorta, renal, and first segment of iliac
arteries very well illustrated by the coronal MPR (a) and
agent. The sequence is fat saturated, and therefore
it is not centric reordered. A test bolus is then
performed in order to determine the arrival time
of the contrast agent in the feeding arteries of the
spinal cord, and the acquisition is started ade-
quately so that the acquisition of the k-space cen-
ter matches the bolus arrival time. For the
arterial and venous 3D gradient echo MR
sequence, the parameters were TE/TR/flip
angle = 1.52 ms/3.98 ms/20 , a matrix size
of 384, pixel size = 1.2  1.0  0.9 mm,
FOV = 380 mm, no rectangular FOV, 88 slices,
0.9 mm slice thickness. The phase encoding direc-
tion is in the craniocaudal direction to avoid flow
and motion artifacts from the ascending
hyperintense aorta projecting on the spinal cord.
The sequence is acquired at the isocenter and
requires 1 min for a single phase.
For the high-resolution late enhanced 3D
gradient echo MR sequence, the parameters
were TE/TR/flip angle = 1.89 ms/4.95 ms/20 ,
two averages, a matrix size of 512, pixel
size = 0.7  0.6  0.6 mm, FOV = 330 mm,
M.I. Vargas et al.
VR reformatting (b). The MRI shows no anomaly of the
conus medullaris and hyperintensity of the aorta due to the
thrombus (c). MRA illustrates the AKA (d)
no rectangular FOV, 72 slices, 0.6 mm slice thick-
ness. The sequence was also performed systemat-
ically at the isocenter.
When your machine does not allow vascular
sequences, CISS, FIESTA, 3D SPACE, and
CUBE sequences permit a better visualization of
the enlarged pial vessels around the spinal cord.
Finally, MPR reformatting and thin MIP are
realized for all acquisitions.
Patterns to Know in Vascular Lesions
The following systematic approach is proposed to
facilitate the diagnosis of vascular malformations.
The following checklist of signs must be used
for all vascular studies of the spinal cord:
1. First, centromedullary high signal of the spi-
nal cord on T2-WI: this can be localized in the
medullary conus, thoracic, or thoracolumbar
and cervical level. The anomaly may vary
and reflects venous hypertension (Fig. 14).
44 Spinal Vascular Imaging: Technique
Fig. 12 Localization of AKA for presurgical planning
of a schwannoma. Note the displacement of the spinal
cord well demonstrated by DTI (c): The lesion is
2. The presence or absence of vessels: They can
be located within or around the spinal cord.
3. The presence of blood in the subarachnoid
space or a hematoma within the spinal cord.
4. An enhancement of the spinal cord due to a
breakdown of the blood-spinal cord barrier in
cases of venous hypertension or arterial ische-
mia or due to the presence of dilated spinal
cord vessels, in the extradural space or in the
paravertebral tissues.
The analysis of signs is tricky and their combi-
nation may cause the following different patterns:
1. Centromedullary high signal of the spinal cord
associated with thin perimedullary corkscrew
dilated vessels in rapport with dilated pial
1081
heterogeneous on T2 (a) with strong enhancement on T1
and the AKA is localized below the lesion (b)
veins, the vascular transverse structures to the
extradural space at root level, and possible
enhancement of the spinal cord reflect a break-
down of the blood-spinal cord barrier [68] and
should evoke the diagnosis of a dural fistula
(Fig. 14). The shunt itself between the vein and
a radicular artery is sometimes difficult to visu-
alize by MRI.
2. If the MRI shows a ball (nidus) of
intramedullary or intra- and perimedullary loca-
tion, with hyperintensity and the
enlarged perimedullary vessels which may
be associated with subarachnoid hemorrhage or
a hematoma, the most probable diagnosis is an
arteriovenous malformation of the spinal cord
(Fig. 15). If it is in contact with the bone,
remodeling may result.
1082 M.I. Vargas et al.

Fig. 13 Hemangioblastoma. MRI shows the vascular the mass very well illustrated by DTI and the enlarged
lesion localized in the spinal cord with a perimedullary vessels around the mass. MRA is used to make a cartogra-
extension. Note the displacement of the spinal cord by phy for presurgical planning

Fig. 14 Dural fistula. Note the high signal of the thoracic spinal cord and the enlarged perimedullary vessels without any
anomaly on the DTI. MRA and axial slices nicely show the dural fistula similar to DSA

3. Another pattern consists of dilated vessels
associated with a hyperintensity of the spinal
cord due to a communication between the pos-
terior and anterior spinal artery with the veins
via direct fistulas, located close to the conus
medullaris; in this case, the most probable
diagnosis is an arteriovenous fistula of the spi-
nal cord. It is important to note that this diag-
nosis by MRI is difficult due to the size of the
shunt (1–2 mm) and can be confounded with a
dural fistula. If the arteriovenous fistula of the
spinal cord is larger, the diagnosis is easier, and
44 Spinal Vascular Imaging: Technique
Fig. 15 Arteriovenous malformation. MR reveals
intramedullary flow voids representing feeding arteries,
nidus, and draining veins (a, b), note the feeder, nicely
in the axial plane, you can see the vascular
structures as pointed by a pencil that embrace
the anterior and posterior spinal cord.
4. If you observe a rounded intramedullary lesion,
well defined with heterogeneous central T1 and
T2 signal (the signal anomaly will depend on the
stage of hemoglobin degradation) and a periph-
eral hypointense rim on T1 and particularly on T2
images, you must think of a cavernoma. Never-
theless, in the case of recent bleeding, MRI pat-
tern can change, and the lesion appears
hyperintense on T2 and isointense or hypointense
on T1. The “popcorn” appearance is the most
characteristic, and the T2 gradient echo sequences
are essential and show a drop of signal due to a
magnetic susceptibility effect. Contrast enhance-
ment is absent or minimal (Fig. 16).
5. The last pattern regarding spinal vascular
malformations is related to aggressive bone
hemangioma. MRI shows a bone lesion in the
vertebral body which may extend to the pedi-
cles, heterogeneously hypointense on T1,
hyperintense on T2, and STIR sequences with
remodeling of the walls of the vertebra, asso-
ciated with an important enhancement and epi-
dural or paravertebral tissue mass. The most
common complication is spinal cord
compression [69].
1083
demonstrated in the MRA reconstructions (c),
corresponding to what is seen on DSA (d)
The imaging of spinal vascular malformations
will be described in the second part of this chapter
including their origins, clinical signs, pathophys-
iology, treatment, and new imaging techniques in
tumors and vascular malformations.
Tumors
Tumors may be divided by their location into
intradural (hemangioblastoma, cavernoma) and
extradural tumors (hemangiomas, angiolipoma).
Spinal cavernomas (Fig. 16), the correct
definition of these lesions are “hamartomas”
and are not true tumors however are in this
section because are angiographically occult
and often produce a mass effect. Represent
5 % of lesions in adults and 1 % in children
[70–73]. They can be sporadic, familial
(multiple), or acquired (radiation induced).
Mutations in genes (CCM1, CCM2, and
CCM3) have been identified [70, 74–76]. The
incidence is about 5 % in adults [70, 72, 77]
and 1 % in children [70, 75] of all spinal
vascular malformations [78]. These are more
common in women than men with a ratio of
2:1 [79] and become symptomatic between the
third and sixth decade of life.
1084
Fig. 16 Cavernoma. Fusiform lesion of the spinal cord,
hyperintense at the center with a hypointense rim on T2
without enhancement associated with high signal around
the lesion due to a cavernoma with bleeding. Note also
The symptoms can vary from asymptomatic,
insidious, or sudden, the latter when the lesion
bleeds.
The most common symptoms are a sensorimo-
tor deficit, progressive paraparesis, or sudden
paraplegia in a young adult.
The location in order of frequency is thoracic
(50 %), cervical (40 %), and spinal cord and at the
medullary cone (10 %) [73].
Cavernomas are angiographically silent and
composed of dilated capillaries without a patho-
logical arterial supply. The scanner shows only
calcified cavernomas; plain radiography is unnec-
essary. MRI is the reference imaging method and
may be performed in an emergency setting when
the onset of symptoms is sudden, due to cavern-
ous bleeding.
Spinal DSA is indicated in case of medullary
hemorrhage, which can reveal cavernomas as well
as glomerular AVM [77]. When a cavernoma is
M.I. Vargas et al.
other lesions in the brain. Cavernomas are visible on mor-
phologic sequences and silent in MRA and conventional
angiography; they do not require MRA
suspected and has not bled, DSA can be useful
with the use flat-panel catheter angiotomography
(FPCA): the aim is to characterize “cryptic venous
anomalies” by the analysis of the abnormal spinal
veins [80].
Pattern number 4 describes the imaging char-
acteristics of cavernoma in MRI (Fig. 16).
The advances in techniques such as diffusion
tensor imaging and tractography can demonstrate
the relationship between the vascular malforma-
tion and spinal cord tracts, which can be useful for
surgical planning [81]. A brain MRI should be
part of the imaging protocol of this type of pathol-
ogy to exclude multiple cavernomas.
Radiation-induced cavernomas may be associ-
ated with signs of postradiation myelitis with a
hyperintense signal on T2 which may extend over
several vertebral levels.
Spinal cavernomas are usually intra-axial, but
there are reports in the literature that describe an
44 Spinal Vascular Imaging: Technique
epidural location with extension to the paraspinal
muscles and foraminal nerve roots [82–93].
The differential diagnosis of intramedullary
cavernomas is mainly with lesions likely to have
a hemorrhagic component such as ependymoma
and hemorrhagic metastases and more rarely with
astrocytomas and hemangioblastomas.
Standard treatment may be abstention and fol-
low up by MRI if asymptomatic or surgical resec-
tion if symptomatic [70, 73, 94, 95].
Spinal hemangioblastoma (Fig. 13) is a large
vascular mass in conventional angiography with a
high vascular nodule; generally, preoperative
embolization is not necessary. Spinal hemangio-
blastoma represents 2–10 % of intramedullary
spinal cord tumors [96–98]. The most frequent
location is intramedullary with some lesions
presenting an exophytic component. Lesions
may also be exclusively intramedullary in loca-
tion or in a minority of cases extramedullary
[96]. Spinal cord hemangioblastomas are associ-
ated with von Hippel-Lindau disease in 25 % [99]
and are typically formed by a nodule and/or a cyst,
both hyperintense on T2; the nodule is isointense
to the spinal cord and the cyst slightly
hyperintense compared to cerebrospinal fluid
(CSF). After contrast, the nodule enhances
strongly and enhancement of the cyst wall is
rare. MRA clearly illustrates the enlarged vessels
in the periphery of the tumor and is useful for the
presurgical planning, as well as DTI, to guide the
neurosurgeon and provide a vascular map in order
to avoid the enlarged vessels and bleeding.
Hemangiomas are the most common spine
tumor; they are often asymptomatic and are usu-
ally an incidental finding. These lesions may be
solitary or multiple; clinical manifestations are
due to compression by the hemangioma,
hemorrhage, or a compression fracture of the ver-
tebra. Symptoms go from back pain to acute
paraplegia. Malignant hemangioma has not been
reported. The thoracic spine is usually
involved, and the hemangioma is located within
the vertebral body, but they can extend to the
neural arch.
The differential diagnosis has to be made with
a solitary plasmacytoma, metastasis, fatty islands,
and Paget’s disease.
1085
In hemangiomas that have a fatty involution,
further investigation is unnecessary. However,
aggressive hemangiomas require a thorough
review and monitoring, and when they become
symptomatic, treatment is usually necessary.
Hemangiomas have a fairly characteristic
appearance on radiographs: round, well-defined
lesions with thick vertical trabeculae.
In case of doubt, a CT scan with millimetric
slices and high resolution is performed; it is the
method of choice for the analysis of bone structure
usually not requiring injection of contrast. If the
lesion shows a typical appearance, i.e., that of a
well-defined lesion, with a stippled appearance,
often with fatty infiltration, other investigations
are unnecessary. If a deformation of the posterior
wall and enhancing epidural soft tissue mass is
visualized on CT, an MRI is essential for better
demonstration of the extension to the spinal canal
and to evaluate possible spinal cord compression.
Magnetic resonance appearance of a
nonaggressive hemangioma is hyperintense on
T1-WI and T2-WI with no or very discrete con-
trast enhancement in the periphery on T1-WI
sequence with FAT SAT.
The aggressive vertebral hemangioma has a
prevalence of 10 % and a predilection for young
women at the thoracic level [100].
The aggressive hemangioma is hypointense or
heterogeneous in T1 and hyperintense in T2, with
significant contrast enhancement. Tissue mass is
frequently associated; the most common compli-
cation is spinal cord compression by the mass
which requires emergency treatment.
Vertebroplasty may be associated with
alcoholization of the epidural component and sur-
gical decompression only when necessary.
Vertebroplasty strengthens the vertebral body
and decreases the risk of bleeding [69]. In some
cases, a presurgical embolization is also used.
Multiple aggressive hemangiomas can be
found in the “PTEN hamartoma tumor
syndrome” [101].
Spinal angiolipoma Are rare tumors mainly
located in the epidural space and more rarely
intradural with a vascular enhancing component
and a fatty part. Clinical symptoms are slowly
progressive. MRI demonstrates a fatty epidural
1086 M.I. Vargas et al.

mass associated with a soft tissue enhancing com-
ponent. Fat saturation is required to confirm the
vascular part of the tumor.
Imaging of Spinal Vascular
Malformations
If MRI is the first imaging modality to perform
when a spinal vascular malformation (SVM) is
suspected, spinal DSA is mandatory to precise
the type of SVM and to decide the therapy. As
the enhancement of the veins is very fast, it is
difficult to precise the anatomy of the SVM with
noninvasive imaging modalities. CTA can be use-
ful to guide DSA in the aim to more rapidly
identify the feeding artery of spinal vascular
malformations or fistulas, but it cannot be a sub-
stitute [102]. CTA seems performant in the detec-
tion of dilated perimedullary veins but, as for
MRI, is not precise enough to determine the
exact type of fistula. Moreover, CTA cannot
exclude the possibility of multiple feeders [103].
The classification of SVM proposed by Krings
et al. presents the advantage of being based on the
vascular anatomy of the SC [77]. The distinction
is made between arteriovenous malformation
(AVM) fed by radiculomedullary arteries and
dural arteriovenous fistula (DAVF) fed by
radiculomeningeal arteries.
SVM are classified as inborn lesions –
represented by AVM – and DAVF considered as
acquired lesions. Epidemiology and clinical pre-
sentation of these SVM are different.
Spinal cord AVMs are high-flow
intramedullary and/or perimedullary
malformations. Three types are described: glo-
merular, fistulous, and juvenile. This distinction
is based on the presence (glomerular) or absence
(fistulous) of a nidus. The juvenile type presents
both fistulous and glomerular compartments, with
an extension to the neighboring tissues (dura,
vertebral body, and paravertebral muscles)
[77]. DSA shows typical feeding by multiple
branches coming from both anterior and posterior
spinal arteries. The malformation can also be

Fig. 17 DSA, AP view.

Eight-month-old boy
presenting an
intramedullary hematoma
revealed by a paraplegia.
There is a familial context
of hereditary hemorrhagic
telangiectasia. Injection into
the left T12 intercostal
artery ((a) arterial phase and
(b) venous phase). (a) A
dilated ASA (arrow) is
directly connected to a
perimedullary venous
pouch. No nidus on MRI
was visible. Diagnosis: pial
arteriovenous fistula. (b)
The dilated venous pouch
drains into an anterior spinal
vein (arrow), which drains
into the intradural (double
arrow) and extradural
(arrowhead) plexuses.
Drainage in the plexuses
and the vein is orthograde
(Image courtesy of Pr Roy,
Service de Radiologie,
CHU de Montréal)
44 Spinal Vascular Imaging: Technique
Fig. 18 Sixty-year-old man presenting hypoesthesia of
the legs and sphincter disturbance. Vertebral fracture was
known and the patient was initially addressed for
vertebroplasty. (a) MRI, sagittal T2 spin echo. MRI
shows increased signal within the conus and multiple
flow voids along the spinal cord (white arrow). (b) DSA,
AP view. Injection into the right L1 lumbar artery. The
intradural point of fistulization (double star) is supplied by
alimented by one feeder, but always coming from
a spinal artery (Fig. 17). An intramedullary hema-
toma which can extend into the subarachnoid
space is common.
Dural arteriovenous fistula (DAVF) is an
intradural shunt between a radiculomeningeal
artery (coming from segmental arteries) and a
radicular vein; a traumatic or iatrogenic origin is
common. Spinal DSA is necessary to precise the
location of the fistula (Figs. 18 and 19).
DAVF represent 70 % of spinal arteriovenous
malformations [68, 77, 104, 105] and are five
times more common in men [104] and may be
associated with neurofibromatosis [106]. Most fis-
tulas are located at the thoracolumbar level [68];
1087
a radiculomeningeal artery. A retrograde opacification of a
radiculomedullary vein (arrow) and the hairpin turn of the
vein (star) which connects with other perimedullary veins
(white arrow) are shown. Drainage from the shunt con-
cerns the intradural plexus also (double arrow). The etiol-
ogy of the progressive myelopathy is a dural
arteriovenous fistula (Image courtesy of Dr Bessou, Ser-
vice de Radiologie, CHU de Bordeaux)
multiple fistulas in the same patient are extremely
rare [68, 104].
DAVF presents in most cases as a progressive
myelopathy resulting in weakness of the lower
extremities, paresthesia, back pain, and urinary
symptoms. These small arteriovenous fistulas
reverse the flow of the root veins, which drain
into the intradural spinal cord venous system,
thereby resulting in venous hypertension and cen-
tral edema of the spinal cord, a stagnant blood
flow and disruption of the blood–brain barrier.
This translates as edema on MRI with a
hyperintense signal on T2 or STIR which extends
far beyond the fistula itself. Once [101] the fistula
is treated, the largest part of the signal abnormality
1088
Fig. 19 Seventy-four-year-old man presenting dyses-
thesia in both feet since 1 year. Vesical sphincter distur-
bance since 6 months. (a) MRI, axial T2 spin echo. A left
anterior epidural lesion (double star) is described. (b, c)
Spinal DSA, AP view. (b) Injection into the left L1 lumbar
artery shows an arterial connection (arrow) with an epidu-
ral venous poach (double star). (c) Arterial hyperselective
injection at the entry of the venous poach shows that the
venous pouch extends on three vertebral levels and drains
M.I. Vargas et al.
through a perimedullary vein (retrograde flow) explaining
the cone edema and the symptoms. Diagnosis: epidural
fistula draining intradurally. (d) Spine X-ray, after embo-
lization with glue of the epidural fistula. The venous pouch
is filled with glue (double star). (e) Non-enhanced CT at
the level of the epidural venous pouch, after embolization.
The hypersignal corresponds to the glue inside the venous
(double star) (Image courtesy of Pr Weill, Service de
Radiologie, CHU de Montréal)
44 Spinal Vascular Imaging: Technique
disappears, but it is possible to find a small abnor-
mal residual signal after several months related to
venous infarction.
A very rare presentation of dural fistulas can be
Foix–Alajouanine syndrome, also known as sub-
acute necrotizing myelitis, which was reported for
the first time by Foix and Alajouanine. Young
patients developed rapid progressive and upward
paralysis and even death. The autopsy revealed
wide areas of necrosis in the lumbar and lower
thoracic cord and the medullary cone with severe
and incomplete destruction of the gray and white
matter.
Summary
MRI is the first-line study of spinal vascular
malformations. CT may play a role in the evalua-
tion of some pathologies, such as bone hemangi-
omas, as well as in the evaluation of aortic
aneurysms before surgery.
Finally, with the emergence of high fields and
new noninvasive vascular imaging techniques,
the detection of vascular malformations becomes
easier and their localization more accurate.
Dynamic sequences allow an approach to the
dynamic study obtained with arteriography
although it is still difficult to differentiate between
arteries and veins especially if they are small and
due to temporal resolution.
Angiography is the second-line study and the
next step to precise the type of malformation and
for the treatment. A multidisciplinary approach is
needed to determine which vascular
malformations will receive endovascular treat-
ment and which will be treated by surgery.
References
1. Santillan A, Nacarino V, Greenberg E et al (2012)
Vascular anatomy of the spinal cord. J Neurointerv
Surg 4(1):67–74
2. Thron AK (2008) Vascular anatomy of the spine and
spinal cord. In: Hurst RW, Rosenwasser RH (eds)
Interventional neuroradiology. Informa Healthcare,
New York, pp 39–55
1089
3. Lasjaunias P, Berenstein A, Ter Brugge K (2002)
Clinical vascular anatomy and variations, Surgical
neuroangiography. Springer, New York
4. Gailloud P (2013) The artery of von Haller: a constant
anterior radiculomedullary artery at the upper thoracic
level. Neurosurgery 73(6):1034–1043
5. Skalski JH, Zembala M (2005) Albert Wojciech
Adamkiewicz: the discoverer of the variable vascu-
larity of the spinal cord. Ann Thorac Surg
80(5):1971–1975
6. Koshino T, Murakami G, Morishita K et al (1999)
Does the Adamkiewicz artery originate from the
larger segmental arteries? J Thorac Cardiovasc Surg
117(5):898–905
7. Biglioli P, Roberto M, Cannata A et al (2004) Upper
and lower spinal cord blood supply: the continuity of
the anterior spinal artery and the relevance of the
lumbar arteries. J Thorac Cardiovasc Surg 127(4):
1188–1192
8. Griepp RB, Ergin MA, Galla JD et al (1996) Looking
for the artery of Adamkiewicz: a quest to minimize
paraplegia after operations for aneurysms of the
descending thoracic and thoracoabdominal aorta.
J Thorac Cardiovasc Surg 112(5):1202–1213; discus-
sion 13–15
9. Biglioli P, Spirito R, Roberto M et al (2000) The
anterior spinal artery: the main arterial supply of the
human spinal cord–a preliminary anatomic study.
J Thorac Cardiovasc Surg 119(2):376–379
10. Lasjaunias P, Berenstein A (1990) Functional vascu-
lar anatomy of the brain, spinal cord and spine,
Surgical neuroangiography. Springer, New York
11. Bowen BC, DePrima S, Pattany PM et al (1996) MR
angiography of normal intradural vessels of the
thoracolumbar spine. AJNR Am J Neuroradiol
17(3):483–494
12. Groen RJ, du Toit DF, Phillips FM et al (2004) Ana-
tomical and pathological considerations in percutane-
ous vertebroplasty and kyphoplasty: a reappraisal
of the vertebral venous system. Spine (Phila Pa
1976) 29(13):1465–1471
13. Bert S, Iyriboz AT, Barret F et al (1995) An angio-
graphic study of spinal vascularization at the thoracic-
lumbar level. J Neuroradiol 22(1):12–19
14. Williams GM, Roseborough GS, Webb TH
et al (2004) Preoperative selective intercostal angiog-
raphy in patients undergoing thoracoabdominal aneu-
rysm repair. J Vasc Surg 39(2):314–321
15. Kieffer E, Fukui S, Chiras J et al (2002) Spinal cord
arteriography: a safe adjunct before descending
thoracic or thoracoabdominal aortic aneurysmectomy.
J Vasc Surg 35(2):262–268
16. Jacobs MJ, Mess W, Mochtar B et al (2006) The value
of motor evoked potentials in reducing paraplegia
during thoracoabdominal aneurysm repair. J Vasc
Surg 43(2):239–246
17. Nair S, Gobin YP, Leng LZ et al (2013) Preoperative
embolization of hypervascular thoracic, lumbar, and
sacral spinal column tumors: technique and outcomes
1090
from a single center. Interv Neuroradiol 19(3):
377–385
18. Chen J, Gailloud P (2011) Safety of spinal angiogra-
phy: complication rate analysis in 302 diagnostic
angiograms. Neurology 77(13):1235–1240
19. Forbes G, Nichols DA, Jack CR Jr et al (1988) Com-
plications of spinal cord arteriography: prospective
assessment of risk for diagnostic procedures. Radiol-
ogy 169(2):479–484
20. Savader SJ, Williams GM, Trerotola SO et al (1993)
Preoperative spinal artery localization and its relation-
ship to postoperative neurologic complications. Radi-
ology 189(1):165–171
21. Heinemann MK, Brassel F, Herzog T et al (1998) The
role of spinal angiography in operations on the tho-
racic aorta: myth or reality? Ann Thorac Surg 65(2):
346–351
22. Uotani K, Yamada N, Kono AK et al (2008) Preoper-
ative visualization of the artery of Adamkiewicz by
intra-arterial CT angiography. AJNR Am
J Neuroradiol 29(2):314–318
23. Takase K, Sawamura Y, Igarashi K et al (2002) Dem-
onstration of the artery of Adamkiewicz at multi-
detector row helical CT. Radiology 223(1):39–45
24. Melissano G, Bertoglio L, Civelli Vet al (2009) Dem-
onstration of the Adamkiewicz artery by
multidetector computed tomography angiography
analysed with the open-source software OsiriX. Eur
J Vasc Endovasc Surg 37(4):395–400
25. Zhao SH, Logan L, Schraedley P et al (2009) Assess-
ment of the anterior spinal artery and the artery of
Adamkiewicz using multi-detector CT angiography.
Chin Med J (Engl) 122(2):145–149
26. Yoshioka K, Niinuma H, Ogino Y et al (2006) Three-
dimensional demonstration of the collateral circula-
tion to the artery of Adamkiewicz with 16-row
multislice computed tomography. Ann Thorac Surg
81(2):749
27. Utsunomiya D, Yamashita Y, Okumura S et al (2008)
Demonstration of the Adamkiewicz artery in patients
with descending or thoracoabdominal aortic aneu-
rysm: optimization of contrast-medium application
for 64-detector-row CT angiography. Eur Radiol
18(11):2684–2690
28. Kudo K, Terae S, Asano T et al (2003) Anterior spinal
artery and artery of Adamkiewicz detected by using
multi-detector row CT. AJNR Am J Neuroradiol
24(1):13–17
29. Nojiri J, Matsumoto K, Kato A et al (2007) The
Adamkiewicz artery: demonstration by intra-arterial
computed tomographic angiography. Eur
J Cardiothorac Surg 31(2):249–255
30. Nakayama Y, Awai K, Yanaga Y et al (2008) Optimal
contrast medium injection protocols for the depiction
of the Adamkiewicz artery using 64-detector CT angi-
ography. Clin Radiol 63(8):880–887
31. Amako M, Yamamoto Y, Nakamura K et al (2011)
Preoperative visualization of the artery of
Adamkiewicz by dual-phase CT angiography in
M.I. Vargas et al.
patients with aortic aneurysm. Kurume Med J
58(4):117–125
32. Clarencon F, Di Maria F, Cormier E et al (2013) Com-
parison of intra-aortic computed tomography angiog-
raphy to conventional angiography in the presurgical
visualization of the Adamkiewicz artery: first results
in patients with thoracoabdominal aortic aneurysms.
Neuroradiology 55(11):1379–1387
33. Nelson RC, Feuerlein S, Boll DT (2011) New itera-
tive reconstruction techniques for cardiovascular
computed tomography: how do they work, and what
are the advantages and disadvantages? J Cardiovasc
Comput Tomogr 5(5):286–292
34. Nishida J, Kitagawa K, Nagata M et al (2014) Model-
based iterative reconstruction for multi-detector row
CT assessment of the Adamkiewicz artery. Radiology
270(1):282–291
35. Machida H, Tanaka I, Fukui R et al (2013) Improved
delineation of the anterior spinal artery with model-
based iterative reconstruction in CT angiography: a
clinical pilot study. AJR Am J Roentgenol
200(2):442–446
36. Ou P, Schmit P, Layouss W et al (2007) CT angiogra-
phy of the artery of Adamkiewicz with 64-section
technology: first experience in children. AJNR Am
J Neuroradiol 28(2):216–219
37. Nijenhuis RJ, Jacobs MJ, Jaspers K et al (2007) Com-
parison of magnetic resonance with computed tomog-
raphy angiography for preoperative localization of the
Adamkiewicz artery in thoracoabdominal aortic aneu-
rysm patients. J Vasc Surg 45(4):677–685
38. Boll DT, Bulow H, Blackham KA et al (2006) MDCT
angiography of the spinal vasculature and the artery of
Adamkiewicz. AJR Am J Roentgenol 187(4):
1054–1060
39. Yoshioka K, Niinuma H, Ohira A et al (2003) MR
angiography and CT angiography of the artery of
Adamkiewicz: noninvasive preoperative assessment
of thoracoabdominal aortic aneurysm. Radiographics
23(5):1215–1225
40. Murakami H, Kawahara N, Tomita K et al (2010)
Does interruption of the artery of Adamkiewicz dur-
ing total en bloc spondylectomy affect neurologic
function? Spine (Phila Pa 1976) 35(22):
E1187–E1192
41. Smit JW, Vielvoye GJ, Goslings BM (2000) Emboli-
zation for vertebral metastases of follicular thyroid
carcinoma. J Clin Endocrinol Metab 85(3):989–994
42. Court C, Noun Z, Gagey O et al (2000) Surgical
treatment of metastases from thyroid cancer in the
axial skeleton. A retrospective study of 18 cases.
Acta Orthop Belg 66(4):345–352
43. Prabhu VC, Bilsky MH, Jambhekar K et al (2003)
Results of preoperative embolization for metastatic
spinal neoplasms. J Neurosurg 98(2 Suppl):156–164
44. Kobayashi K, Ozkan E, Tam A et al (2012) Preoper-
ative embolization of spinal tumors: variables affect-
ing intraoperative blood loss after embolization. Acta
Radiol 53(8):935–942
44 Spinal Vascular Imaging: Technique
45. Wilson MA, Cooke DL, Ghodke B et al (2010) Ret-
rospective analysis of preoperative embolization of
spinal tumors. AJNR Am J Neuroradiol 31(4):
656–660
46. Robial N, Charles YP, Bogorin I et al (2012) Is pre-
operative embolization a prerequisite for spinal
metastases surgical management? Orthop Traumatol
Surg Res 98(5):536–542
47. Thiex R, Harris MB, Sides C et al (2013) The role of
preoperative transarterial embolization in spinal
tumors. A large single-center experience. Spine J
13(2):141–149
48. Jones KM, Gobin P, Liu A-H (2003) Embolisation of
spinal tumors. Oper Tech Neurosurg 6(3):156–162
49. Smith TP, Gray L, Weinstein JN et al (1995) Preop-
erative transarterial embolization of spinal column
neoplasms. J Vasc Interv Radiol 6(6):863–869
50. Shi HB, Suh DC, Lee HK et al (1999) Preoperative
transarterial embolization of spinal tumor: emboliza-
tion techniques and results. AJNR Am J Neuroradiol
20(10):2009–2015
51. O’Reilly GV, Kleefield J, Klein LA et al (1989)
Embolization of solitary spinal metastases from
renal cell carcinoma: alternative therapy for spinal
cord or nerve root compression. Surg Neurol
31(4):268–271
52. Djindjian R, Cophignon J, Rey A et al (1973)
Superselective arteriographic embolization by the
femoral route in neuroradiology. Study of 50 cases.
II. Embolization in vertebromedullary pathology.
Neuroradiology 6(3):132–142
53. Estrera AL, Miller CC 3rd, Huynh TT et al (2001)
Neurologic outcome after thoracic and
thoracoabdominal aortic aneurysm repair. Ann
Thorac Surg 72(4):1225–1230; discussion 30–31
54. Coselli JS, LeMaire SA, Miller CC 3rd et al (2000)
Mortality and paraplegia after thoracoabdominal aor-
tic aneurysm repair: a risk factor analysis. Ann Thorac
Surg 69(2):409–414
55. Griepp RB, Griepp EB (2007) Spinal cord perfusion
and protection during descending thoracic and
thoracoabdominal aortic surgery: the collateral net-
work concept. Ann Thorac Surg 83(2):S865–S869;
discussion S90–S92
56. Weigang E, Hartert M, Siegenthaler MP et al (2006)
Perioperative management to improve neurologic
outcome in thoracic or thoracoabdominal aortic
stent-grafting. Ann Thorac Surg 82(5):1679–1687
57. Amabile P, Grisoli D, Giorgi R et al (2008) Incidence
and determinants of spinal cord ischaemia in stent-
graft repair of the thoracic aorta. Eur J Vasc Endovasc
Surg 35(4):455–461
58. Fereshetian A, Kadir S, Kaufman SL et al (1989)
Digital subtraction spinal cord angiography in
patients undergoing thoracic aneurysm surgery.
Cardiovasc Intervent Radiol 12(1):7–9
59. Nijenhuis RJ, Jacobs MJ, Schurink GW et al
(2007) Magnetic resonance angiography and
neuromonitoring to assess spinal cord blood supply
1091
in thoracic and thoracoabdominal aortic aneurysm
surgery. J Vasc Surg 45(1):71–77; discussion 77–78
60. Etz CD, Halstead JC, Spielvogel D et al (2006) Tho-
racic and thoracoabdominal aneurysm repair: is
reimplantation of spinal cord arteries a waste of
time? Ann Thorac Surg 82(5):1670–1677
61. Nakai M, Shimizu S, Ochi Y et al (2009)
Thoracodorsal artery as a collateral source to the
artery of Adamkiewicz after endovascular aneurysm
repair for descending thoracic aortic aneurysm. Eur J
Vasc Endovasc Surg 37(5):566–568
62. Maliszewski M, Ladzinski P, Aleksandrowicz R
et al (1999) Occlusion of radicular arteries – reasons,
consequences and anastomotic substitution pathways.
Spinal Cord 37(10):710–716
63. von Tengg-Kobligk H, Bockler D, Jose TM
et al (2007) Feeding arteries of the spinal cord at CT
angiography before and after thoracic aortic
endografting. J Endovasc Ther 14(5):639–649
64. Etz CD, Luehr M, Kari FA et al (2008) Paraplegia
after extensive thoracic and thoracoabdominal aortic
aneurysm repair: does critical spinal cord ischemia
occur postoperatively? J Thorac Cardiovasc Surg
135(2):324–330
65. Vargas MI, Nguyen D, Viallon M et al (2010)
Dynamic MR angiography (MRA) of spinal vascular
diseases at 3 T. Eur Radiol 20(10):2491–2495.
doi:10.1007/s00330-010-1815-6
66. Wang D, Yang N, Zhang P et al (2013) The diagnosis
of spinal dural arteriovenous fistulas. Spine.
doi:10.1097/BRS.0b013e31828a38c4
67. Vargas MI, Delavelle J, Jlassi H et al (2008) Clinical
applications of diffusion tensor tractography of the
spinal cord. Neuroradiology 50(1):25–29
68. Krings T, Geibprasert S (2009) Spinal dural arterio-
venous fistulas. AJNR Am J Neuroradiol 30(4):
639–648. doi:10.3174/ajnr.A1485
69. Ide C, Gangi A, Rimmelin A et al (1996) Vertebral
haemangiomas with spinal cord compression: the
place of preoperative percutaneous vertebroplasty
with methyl methacrylate. Neuroradiology 38(6):
585–589
70. Kharkar S, Shuck J, Conway J et al (2007) The natural
history of conservatively managed symptomatic
intramedullary spinal cord cavernomas. Neurosurgery
60(5):865–872. doi:10.1227/01.NEU.0000255437.3
6742.15; discussion 65–72
71. Deutsch H, Jallo GI, Faktorovich A et al (2000) Spi-
nal intramedullary cavernoma: clinical presentation
and surgical outcome. J Neurosurg 93(1 Suppl):
65–70
72. Spetzger U, Gilsbach JM, Bertalanffy H (1995) Cav-
ernous angiomas of the spinal cord clinical presenta-
tion, surgical strategy, and postoperative results. Acta
Neurochir (Wien) 134(3–4):200–206
73. Babu R, Owens TR, Karikari IO et al (2013) Spinal
cavernous and capillary hemangiomas in adults.
Spine 38(7):E423–E430. doi:10.1097/BRS.0b013e3
18287fef7
1092
74. Bergametti F, Denier C, Labauge P et al (2005) Muta-
tions within the programmed cell death 10 gene cause
cerebral cavernous malformations. Am J Hum Genet
76(1):42–51. doi:10.1086/426952
75. Guzeloglu-Kayisli O, Amankulor NM, Voorhees J
et al (2004) KRIT1/cerebral cavernous malformation
1 protein localizes to vascular endothelium,
astrocytes, and pyramidal cells of the adult human
cerebral cortex. Neurosurgery 54(4):943–949; discus-
sion 49
76. Labauge P, Brunereau L, Coubes P et al (2001)
Appearance of new lesions in two nonfamilial cere-
bral cavernoma patients. Eur Neurol 45(2):83–88.
doi:10.1159/000052100
77. Krings T, Mull M, Gilsbach JM et al (2005) Spinal
vascular malformations. Eur Radiol 15(2):267–278.
doi:10.1007/s00330-004-2510-2
78. Biondi A, Clemenceau S, Dormont D et al (2002)
Intracranial extra-axial cavernous (HEM) angiomas:
tumors or vascular malformations? J Neuroradiol
29(2):91–104
79. Pechstein U, Zentner J, Van Roost D et al (1997)
Surgical management of brain-stem cavernomas.
Neurosurg Rev 20(2):87–93
80. Pearl MS, Chen JX, Gregg L et al (2012) Angio-
graphic detection and characterization of “cryptic
venous anomalies” associated with spinal cord cav-
ernous malformations using flat-panel catheter
angiotomography. Neurosurgery 71(1 Suppl
Operative):125–132
81. Vargas MI, Delavelle J, Jlassi H et al (2008) Clinical
applications of diffusion tensor tractography of the
spinal cord. Neuroradiology 50(1):25–29.
doi:10.1007/s00234-007-0309-y
82. Rovira A, Rovira A, Capellades J et al (1999) Lumbar
extradural hemangiomas: report of three cases. AJNR
Am J Neuroradiol 20(1):27–31
83. Hemalatha AL, Ravikumar T, Chamarthy NP
et al (2013) A pure epidural spinal cavernous heman-
gioma – with an innocuous face but a perilous behav-
iour!! J Clin Diagn Res JCDR 7(7):1434–1435.
doi:10.7860/JCDR/2013/6030.3159
84. Zhong W, Huang S, Chen H et al (2012) Pure spinal
epidural cavernous hemangioma. Acta Neurochir
(Wien) 154(4):739–745. doi:10.1007/s00701-012-
1295-3
85. Shin JH, Lee HK, Rhim SC et al (2001) Spinal epi-
dural cavernous hemangioma: MR findings.
J Comput Assist Tomogr 25(2):257–261
86. Aoyagi N, Kojima K, Kasai H (2003) Review of
spinal epidural cavernous hemangioma. Neurol Med
Chir 43(10):471–475; discussion 76
87. Hara K, Hidaka K (2006) Cavernous hemangioma
growing continuously from the spinal epidural space
into the paravertebral muscles: case report. No
Shinkei Geka Neurol Surg 34(5):521–525
M.I. Vargas et al.
88. Hatiboglu MA, Iplikcioglu AC, Ozcan D (2006) Epi-
dural spinal cavernous hemangioma. Neurol Med
Chir 46(9):455–458
89. Kurose K, Kishi H, Sadatoh T (1989) Spinal epidural
cavernous hemangioma. Case report. Neurol Med
Chir 29(6):538–542
90. Minh NH (2005) Cervicothoracic spinal epidural cav-
ernous hemangioma: case report and review of the
literature. Surg Neurol 64(1):83–85. doi:10.1016/j.
surneu.2004.10.030; discussion 85
91. Padolecchia R, Acerbi G, Puglioli M et al (1998)
Epidural spinal cavernous hemangioma. Spine
23(10):1136–1140
92. Richardson RR, Cerullo LJ (1979) Spinal epidural
cavernous hemangioma. Surg Neurol 12(3):266–268
93. Rachinger J, Buslei R, Engelhorn T et al (2006)
Intradural-extramedullary cavernous hemangioma of
the left motor root C7 – case report and update of the
literature. Zentralbl Neurochir 67(3):144–148.
doi:10.1055/s-2006-933362
94. Lefranc F, Baleriaux D, Brotchi J (2007)
Intramedullary cavernomas: personal series of
24 cases. Neurochirurgie 53(2–3 Pt 2):203–207.
doi:10.1016/j.neuchi.2007.02.009
95. Liang JT, Bao YH, Zhang HQ et al (2011) Manage-
ment and prognosis of symptomatic patients with
intramedullary spinal cord cavernoma: clinical article.
J Neurosurg Spine 15(4):447–456. doi:10.3171/
2011.5.SPINE10735
96. Imagama S, Ito Z, Wakao N et al (2011) Differentia-
tion of localization of spinal hemangioblastomas
based on imaging and pathological findings. Eur
Spine J Off Publ Eur Spine Soc Eur Spinal Deform
Soc Eur Sect Cerv Spine Res Soc 20(8):1377–1384.
doi:10.1007/s00586-011-1814-6
97. Browne TR, Adams RD, Roberson GH (1976)
Hemangioblastoma of the spinal cord. Review and
report of five cases. Arch Neurol 33(6):435–441
98. Parsa AT, Lee J, Parney IF et al (2004) Spinal cord and
intradural-extraparenchymal spinal tumors: current
best care practices and strategies. J Neurooncol
69(1–3):291–318
99. Mechtler LL, Nandigam K (2013) Spinal cord
tumors: new views and future directions. Neurol
Clin 31(1):241–268. doi:10.1016/j.ncl.2012.09.011
100. Fox MW, Onofrio BM (1993) The natural history and
management of symptomatic and asymptomatic ver-
tebral hemangiomas. J Neurosurg 78(1):36–45.
doi:10.3171/jns.1993.78.1.0036
101. Jenny B, Radovanovic I, Haenggeli CA et al (2007)
Association of multiple vertebral hemangiomas and
severe paraparesis in a patient with a PTEN
hamartoma tumor syndrome. Case report. J
Neurosurg 107(4 Suppl):307–313. doi:10.3171/
PED-07/10/307
44 Spinal Vascular Imaging: Technique
102. Si-jia G, Meng-wei Z, Xi-ping L et al (2009) The
clinical application studies of CT spinal angiography
with 64-detector row spiral CT in diagnosing
spinal vascular malformations. Eur J Radiol
71(1):22–28
103. Yamaguchi S, Nagayama T, Eguchi K et al (2010)
Accuracy and pitfalls of multidetector-row computed
tomography in detecting spinal dural arteriovenous
fistulas. J Neurosurg Spine 12(3):243–248
1093
104. Thron A (2001) Spinal dural arteriovenous fistulas.
Radiologe 41(11):955–960
105. Krings T (2010) Vascular malformations of the spine
and spinal cord: anatomy, classification, treatment.
Klin Neuroradiol. doi:10.1007/s00062-010-9036-6
106. Kamiyama K, Endo S, Horie Y et al (1985) Neurofi-
bromatosis associated with intra- and extracranial
aneurysms and extracranial vertebral arteriovenous
fistula. No Shinkei Geka Neurol Surg 13(8):875–880
 Spinal Vascular Anatomy
45
Tibor Becske and Eytan Raz

Contents Keywords
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1095
Artery of adamkiewicz • Artery of lazorthes •
Radiculomedullary • Spinal vasculature
Arteries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1096
Veins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1101
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1103 Introduction

The purpose of this chapter is to summarize the

spinal vascular anatomy. After a detailed descrip-
tion of the segmental anatomy, the spinal cord
supply will ensue with description of the radiculo-
medullary arteries, the anterior spinal artery, and
the spinal veins. Part of the text, the concept and
the figures in this chapter are used with permission
from Maksim Shapiro, from the website www.
neuroangio.org.
The neural tube/spinal cord can be conceptual-
ized as a cylindrical mass of tissue. For a short
period of time after closure of the neural tube,
nutritional support can be adequately provided
by simple diffusion. When limits of diffusion are
exceeded, dedicated vascular supply to the spinal
cord and adjacent tissues is established via seg-
mental (metameric) vessels arising from the dor-
sal aorta. The human embryo is in fact subdivided
into 31 somites, each corresponding to a develop-
ing metameric segment, that ultimately gives rise
to all endo-, ecto-, and mesodermal derivatives.
Each somite is supplied by those paired segmental
T. Becske (*) • E. Raz (*) arteries originating from the dorsal aorta. Each
Neurointerventional Radiology Section, Department of segmental vessel supports its endodermal, ecto-
Radiology, NYU Langone Medical Center, New York, NY,
USA dermal, and mesodermal tissue. Longitudinal
e-mail: eytan.raz@gmail.com anastomoses are then established between
# Springer Science+Business Media New York 2016 1095
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_14
1096
segmental vessels, thereby giving rise to new ves-
sels extending craniocaudally along the length of
the cord. A grid-like pattern is hence established
which is beginning to resemble mature
arrangement.
But, as longitudinal vessels develop, hemody-
namic need for having a segmental radiculo-
medullary vessel at each level is diminished.
Most segmental vessels diminish to supply only
the region of segmental nerve root. The cord is
now supplied by the “anterior spinal” artery ven-
trally. A loose network of arteries with two
roughly dominant channels forms the posterior
spinal arteries. Finally, only several segmental
vessels remain to supply the cord. These are the
radiculomedullary vessels (feeding nerve root and
cord) of the adult. Other segmental vessels persist
as purely radicular (supplying only nerve roots).
The successive transversely oriented segmen-
tal arteries anastomose through a number of
paraspinal vessels longitudinally arranged along
the axis of the spine. These longitudinal channels
include the prevertebral artery situated adjacent to
the anterolateral aspect of the vertebral body, the
pre-transverse anastomoses anterior to the trans-
verse process, as well as the post-transverse spinal
anastomotic arcade which extends craniocaudally
along both sides of the spinal processes. In the
lower lumbosacral spine, a homologous arrange-
ment is recognizable – with the median sacral
artery representing the continuation of the aorta
and internal iliac arteries considered the homo-
logues of the paravertebral longitudinal arcades,
respectively – supplying the corresponding lower
lumbar and sacral metamers.
The same segmental arrangement is present in
the upper thoracic and cervical spine: the ascend-
ing cervical artery which corresponds to the
prevertebral artery, the vertebral artery (coursing
in the cervical osseous homologue of the trans-
verse process) which is homologous with the
pre-transverse anastomotic artery, and the deep
cervical artery with the post-transverse
anastomosis.
As a consequence of this vascular disposition,
the contributions to the anterior spinal axis can
T. Becske and E. Raz
originate from any of the above described longi-
tudinal vessels, although most commonly, the
dominant cervical radiculomedullary branch
(artery of Lazorthes) originates from the lower
vertebral artery. The upper thoracic spine is a
transitional zone where segmental arteries are
supplied by the supreme intercostal artery, which
corresponds to the continuation of the
pre-transverse anastomosis. As the cord continues
to enlarge, perforator or sulco-comissural
branches emerge, which unless pathologically
enlarged, they are below limits of angiographic
resolution.
Arteries
The following discussion is visually recapitulated
in Fig. 1, from the website www.neuroangio.org.
Throughout the following description, all upper-
case letters given in parentheses refer to legends in
Fig. 1.
The aorta (A) (Fig. 2) gives rise to the proto-
typical spinal segmental artery, lumbar or thoracic
(B), which after the origin, continues posterior
and laterally along the vertebral body. Examples
of these segmental arteries are shown in Figs. 2, 3,
4, 5, 6, and 7. At the junction between the verte-
bral body and the transverse process, the segmen-
tal artery gives rise to the dorsal spinal trunk, also
called dorsal spinal artery, which quickly bifur-
cates into dorsal (K) and ventral divisions (H).
The segmental artery continues as either an inter-
costal or muscular artery (F). The ventral division
of the dorsal spinal artery (K) traverses the neural
foramen and supplies all adjacent neural, dural,
and osseous structures. It gives off anterior epidu-
ral branches which form a characteristic,
diamond-shaped retrocorporeal arcade
vascularizing the ventral epidural space and serv-
ing as a potential collateral conduit to the contra-
lateral segmental artery (Fig. 8). Branches to the
dorsal epidural space form a less consistently
visualized epidural arcade. The continuation of
the ventral branch becomes associated with the
corresponding nerve root sleeve, where it
45 Spinal Vascular Anatomy
Fig. 1 Schematics of the arterial supply to spinal cord
(Courtesy of Dr. Maksim Shapiro from his website, with
permission, www.neuroangio.org). A – aorta; B – segmen-
tal artery; Ba – intersegmental arterial anastomosis; C –
prevertebral anastomotic network; D – direct vertebral
body feeding arteries; E – dorsal spinal artery; F – inter-
costal/muscular artery; G – pretransverse anastomotic net-
work; H – dorsal division of the dorsal spinal artery; I –
post-transverse anastomotic network; J – muscular
branches of the post-transverse anastomotic network; K –
ventral division of the dorsal spinal artery; Ka – radicular
artery; La – ventral epidural arcade; Lb – dorsal epidural
arcade; M – nerve root sleeve dural branch of the ventral
provides radiculodural branches to supply
regional dura and ultimately penetrates the sleeve
to become the radicular artery, supplying anterior
and posterior nerve roots and at discrete levels
contributing to the vascularization of the spinal
cord as radiculomedullary (P) or radiculopial
arteries (O).
The dorsal division of the dorsal spinal artery
(H) passes posteriorly beneath the ipsilateral
transverse process and along the outer surface of
1097
division dorsal spinal artery; N – dural branch of the
ventral division dorsal spinal artery; O – radiculopial
artery; P – radiculomedullary artery; Q – anterior spinal
artery; R – mesh-like pial arterial network; S, T – posterior
spinal artery; U, V – pial arterial network
(a.k.a. vasocorona) anastomoses between anterior and pos-
terior spinal arterial systems, W – sulcocommissural artery,
X – rami perforantes of the peripheral (centripetal) system,
Y – central (centrifugal) system of sulcal arteries, originat-
ing from pial network of the cord; altogether, the pial
network and rami perforantes (R+Y) are called the
vasocorona or corona vasorum; Z – rami cruciantes
(a.k.a. crux vasculosa, a.k.a. rami anastomotici arcuati)
the lamina, forming a post-transverse longitudinal
anastomosis (I) close to the spinous processes.
The most effective transverse anastomosis
between left- and right-sided vessels is
represented by the retrocorporeal arcade within
the anterior epidural space (Fig. 8).
The arteries supplying neural structures within
the dural sheath can be divided into three groups:
(a) small radicular arteries, below angiographic
resolution, which are branches of the dorsal spinal
1098 T. Becske and E. Raz

Fig. 2 Aortogram demonstrating the segmental branches arising from the aorta. Notice the different orientation of the
segmental branches going from the thoracic to the lumbar region

Fig. 3 This is a left L3 pedicle injection in arterial phase subtracted (a) and unsubtracted (b)
45 Spinal Vascular Anatomy
Fig. 4 This is a left T11 pedicle injection in arterial phase
subtracted (a) and unsubtracted (b). There is a well visible
dominant radiculomedullary contribution to the anterior
artery, which are present at nearly every level, and
which supply the nerve root. At some levels, the
territory of these arteries is not limited to the nerve
root, in which case the vessel is termed radiculo-
medullary, when supplying the anterior spinal
axis, or radiculopial when supplying posterior
spinal axis.
The anterior spinal artery (Q) is a form of
longitudinal anastomosis between multiple trans-
verse radicular arteries. In the adult form, the
anterior spinal artery is a longitudinal channel
located in the anterior median sulcus, with the
caliber varying with regional hemodynamic
requirements, being larger in the cervical
(0.2–0.5 mm) and lumbar regions (0.5–0.8 mm)
1099
spinal artery. The dominant radiculomedullary contribu-
tion to the anterior spinal artery comes more often from the
left side between T9 and T12
and quite smaller in the mid-thoracic area (0.1–04
mm). The anterior spinal artery (Figs. 4 and 9) is
supplied by multiple radiculomedullary arteries,
most commonly from the proximal cervical ver-
tebral artery (artery of Lazorthes) – but can orig-
inate instead from anterior cervical, deep cervical
supreme intercostal arteries. At thoracic and lum-
bar levels, radiculomedullary vessels arise from
the ventral division of the segmental artery
(Figs. 4 and 9). The main thoracolumbar contrib-
utory to the anterior spinal artery, the artery of
Adamkiewicz arises from the lower thoracic or
upper lumbar region, between T9 and L1 most
commonly from the left side. The confluence of
the radiculomedullary artery and anterior spinal
1100
Fig. 5 This is a left T12 pedicle injection in arterial phase
unsubtracted (a) and subtracted (b). Notice the well visible
epidural diamond-shaped arterial plexus (retrocorporeal
arcade). Also partial visualization of the radiculomedullary
artery typically has a classical hairpin-turn shape
(Fig. 9).
The posterior spinal artery axis does not exist
as a continuous channel but a discontinuous chan-
nel beyond several adjacent levels supplied by
multiple radiculopial arteries, the larger visualized
angiographically as hairpin-like vessels, similar to
the one supplying the anterior spinal artery, but
slightly off-midline.
Arterial supply of conus medullaris consists of
a characteristic “basketlike” arrangement,
reflecting the confluence of the anterior spinal
artery with the paired posterior spinal
arteries, via prominent anastomotic channels
called rami cruciantes (Fig. 10). The visualization
of this vascular basket is highly desirable as
part of a complete spinal angiogram
T. Becske and E. Raz
contribution to the anterior spinal artery is again seen,
which is related to anastomotic connections between the
T11 and T12 pedicles
and angiographically identifies the spinal cord
conus.
The intrinsic spinal cord vasculature was
subdivided into centrifugal and centripetal com-
ponents. The anterior spinal artery gives rise to
multiple sulco-comissural vessels, which pene-
trate deep within the anterior median sulcus to
vascularize the central gray matter, below DSA
resolution. The sulco-commissural arteries give
rise to a network of small vessels, which radiate
centrifugally from central cord to the peripheral
spinal cord. The sulco-commissural distribution is
classically described as covering the anterior
two-thirds of the cross-sectional area of the spinal
cord, while the centripetally directed rami
perforantes, originating from the pial vascular
meshwork covering the surface of the spinal
45 Spinal Vascular Anatomy
Fig. 6 This is a left supreme intercostal pedicle injection in arterial phase unsubtracted (a) and subtracted (b)
cord, cover the posterior third. This arrangement
creates a watershed zone within the cord paren-
chyma between centripetal and centrifugal
systems.
Veins
The spinal venous system consists of three differ-
ent components: (a) intrinsic venous network of
the spinal cord parenchyma, (b) extrinsic system
consisting of superficial spinal cord veins and
radicular veins, and (c) extradural venous plexus,
consisting of veins in the epidural plexus, sur-
rounding the vertebral bodies and the paraspinal
musculature.
1101
The intramedullary network consists of two
components, the sulcal and the radial veins. The
sulcal component predominantly drains the terri-
tory of gray matter in a centripetal fashion, with
outflow in the sulco-commissural veins, while the
radial veins drain centrifugally toward the surface
of the cord. Eventually, both systems empty into
the superficial spinal cord veins. The main longi-
tudinal veins are the anterior and posterior spinal
veins which are located in the subpial space. The
anterior midline vein is located behind the artery
in the anterior median sulcus, while the posterior
median spinal vein is located in the dorsal median
fissure. The longitudinal veins are the largest
perimedullary vessels, measuring up to 1.5 mm,
and are the vessels most likely to be seen on the
postcontrast spine MRI of normal patients. The
1102 T. Becske and E. Raz

anterior and posterior spinal veins are connected

by multiple transverse veins surrounding the cord,
as well as transmedullary channels running
through the spinal cord. The superficial venous
blood drains in the epidural venous plexus via the
radicular veins, which have a hairpin shape, sim-
ilar to the shape of the arteries (for this reason it is
not possible to discern between the radicular sup-
ply to an anterior spinal artery and a dominant
radicular vein on a non-dynamic spine MRA
study). The radicular veins represent the weaker
segment of the venous system since there is no
redundancy, while that is not the case for both the
intrinsic spinal cord system and the extrinsic
extradural veins. The radicular veins travel along
the nerve roots across the subarachnoid space and
enter the dural nerve root sleeve at a shallow
angle, an arrangement which acts as a functional
valve, preventing venous reflux into the cord
when there is increased central venous pressure.
The radicular veins drain in the epidural plexus,
which extends all the way continuously from the
base of the skull to the sacrum and via segmental
Fig. 7 This is a median sacral artery pedicle injection in
arterial phase, subtracted view
intervertebral (foraminal) veins is connected with
the azygos and hemiazygos veins.

Fig. 8 This is a left L2 pedicle injection in arterial phase unsubtracted (a) and subtracted (b). Notice the well visible
diamond-shaped retrocorporeal arcade
45 Spinal Vascular Anatomy
Fig. 9 This is a left T7 pedicle injection in arterial phase,
subtracted view. Notice again the radiculomedullary con-
tribution to the anterior spinal artery
References
1. Adamkiewicz AW (1881) Die BlutgefaBe des
mencshlichen Ruckenmarks. I. Die GefaBe der
Ruckenmarks substanz. Sitz. Ber. Akad. Wiss. Wien,
Math. nat. Kl., Berlin, pp 469–502
2. Adamkiewicz AW (1882) Die BlutgefaBe des
mencshlichen Ruckenmarks. II. Die GefaBe der
Ruckenmarks-oberflashe. Sitz. Ber. Akad. Wiss. Wien,
Math. nat. Kl., pp 101–130
3. Djindjian R (1970) Angiography of the spinal cord. In:
Hurth M (ed) Angiography of the spinal cord/by R
Djindjian, with collaboration of M Hurth [and others]
1103
Fig. 10 This image shows the typical angiographic
appearance of the conus medullaris. The basket-like
arrangement reflects the confluence between the anterior
and posterior spinal arteries
Pref by Guy Lazorthes English translation by Irvin I
Kricheff. University Park Press, Baltimore, p 482
4. Lazorthes G, Gouaze A, Djindjian R (1973) Vasculari-
zation et pathologie vasculaire de la moelle epiniere.
Masson, Paris
5. Lasjaunias PL, Berenstein A, Raybaud C (1987) Surgi-
cal neuroangiography. Springer, Berlin/New York
6. Lasjaunias PL, Berenstein A, Brugge KG (2001) Surgical
neuroangiography, 2nd edn. Springer, Berlin/New York
7. Shapiro M (2012) www.neuroangio.org. [cited; Cere-
brospinal neurovascular information source]. Available
from www.neuroangio.org
 Spinal Vascular Malformations and
Treatment 46
Srinivasan Paramasivam

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1106 Spinal vascular malformations are rare, but a
diverse group of neurovascular pathologies
Spinal Vascular Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . 1106
including arteriovenous malformations
Spinal Vascular Malformation Classification . . . 1107 (AVMs), cavernous malformations, and dural
Type I: Dural Arteriovenous Fistula . . . . . . . . . . . . . . . 1107 arteriovenous fistulas (dAVFs) that occur due
Type II: Intramedullary Arteriovenous
Malformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1109 to developmental derangement of the vascular
Type III: Juvenile Arteriovenous Malformation . . . 1113 system. Though classified variously, the com-
Type IV: Intradural Perimedullary Arteriovenous monly followed one is anatomic classification
Fistula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1116 based on the morphology of the malformation
Cavernous Malformation (Cavernomas) . . . . . . . . . . . 1119
to which added are cavernomas and spinal
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1121 vascular tumors. Pathophysiologic mecha-
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1122 nisms leading to clinical symptoms include
intramedullary or subarachnoid hemorrhages,
arterial ischemia, progressive venous conges-
tion resulting in progressive myelopathy,
space-occupying nature of the malformation
and its venous drainage, and circulatory
steal phenomenon. Most spinal vascular
malformations come to clinical attention by
nonspecific symptomatology that may be
acute, subacute, or chronic. The initial clinical
diagnosis is challenging and it is based on
MRI; however, for better understanding and
planning therapeutic strategy spinal angiogram
is necessary. Once diagnosed, most spinal vas-
cular malformations need treatment, as the nat-
ural history suggests progressive stepwise
neurological deterioration. Early diagnosis
S. Paramasivam (*) and timely management in symptomatic
Mount Sinai Roosevelt Hospital, Hyman Newman Institute
patients can result in improvement or stabiliza-
for Neurology and Neurosurgery, Centre for Endovascular
Surgery, New York, NY, USA tion of clinical condition. Treatment should be
e-mail: kpsvasan@hotmail.com performed in specialized centers. Except spinal
# Springer Science+Business Media New York 2016 1105
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_45
1106
cord cavernomas and spinal vascular tumors,
all other vascular disorders are initially man-
aged by endovascular embolization, as it is the
least invasive but technically challenging and
demanding procedure. For cases where embo-
lization is technically not possible or when
they fail embolization treatment, surgical treat-
ment is an available option. In select cases a
combined therapy might be sensible. Early
diagnosis and prompt treatment are essential
to alter the natural course and result in clinical
improvement.
Keywords
Spinal vascular malformation • Arteriovenous
malformation • Dural arteriovenous fistula •
Perimedullary arteriovenous fistula • system [2]. The anterior spinal artery originates
Cavernoma • Juvenile arteriovenous malfor-
mation • AVM • Myelopathy • Paraplegia •
Endovascular embolization • Spinal MRI •
Spinal angiogram • Spinal hemorrhage
Introduction
Spinal vascular malformations are developmental
derangements of the vascular system more specif-
ically interconnections between the arteries and
veins resulting in abnormal flow that may result in
secondary features of vascular enlargement, aneu-
rysm, and varix formation. With the spinal canal
being a closed space, the clinical symptoms may
be due to mass effect from the enlarged vessels,
excessively high flow in the malformation
resulting in vascular steal and rarely cardiac fail-
ure, and/or elevated venous pressure with low
blood flow and back pressure ischemia due to
venous congestion.
Spinal Vascular Anatomy
Spinal vascular anatomy is highly variable but
predictable. Segmental arteries arising from vari-
ous sources that include the vertebral and deep
cervical arteries in the neck, the posterior inter-
costal arteries in the thorax, and the lumbar arter-
ies in the abdomen provide vascular supply to the
S. Paramasivam
spine, including the vertebral bodies, paraspinal
muscles, dura, nerve roots, and spinal cord with
blood. Radicular arteries are the first branches of
the dorsal division of the segmental arteries. At
each level the radiculomeningeal artery arises
from the radicular artery and supplies the dural
and the nerve root. From these radicular arteries
also arise the radiculomedullary and radiculopial
arteries that follow the anterior or posterior nerve
root to reach the anterior or posterior surface of
the cord, where they form the anterior or posterior
spinal arteries [1]. The radiculomedullary feeders
are not present at every level, and their location is
highly variable and is not predictable.
The anterior and posterior spinal arteries con-
stitute a superficial longitudinal anastomosing
from the two vertebral arteries intracranially and
unites in the midline to travel along the anterior
sulcus for the entire length of the spinal cord. The
paired posterolateral spinal arteries originate from
the pre-atlantal part of the vertebral artery or from
the posteroinferior cerebellar artery and run on the
posterior surface of the spinal cord as a network
for the entire length. The posterior and anterior
arterial systems are interconnected at the level of
the cone via the “basket” anastomosis and via
transverse anastomoses by the pial network of
the vasocorona. They are reinforced by the ante-
rior and posterolateral radiculomedullary arteries
at different segmental levels [3]. Most commonly,
the anterior radiculomedullary artery at T9 level is
the largest called arteria radiculomedullaris
magna aka the Adamkiewicz artery. The anterior
radiculomedullary artery follows the spinal nerve
root to ascend to the segmental level on the ante-
rior surface of the spinal cord and branches in a
typical way with the ascending limb continuing in
the direction of the radicular artery, while the
descending limb does a hairpin curve before join-
ing the anterior spinal artery forming a diamond-
shaped defect along the anterior spinal axis
[1]. The substance of the spinal cord is supplied
by the intrinsic network of vessels that include
sulco-commissural arteries arising from the ante-
rior spinal artery and perforating arteries arising
from the circumferential vasocorona derived from
the anterior and paired posterolateral arteries [2].
46 Spinal Vascular Malformations and Treatment
The rich longitudinal and transverse anastomosis
protects the spinal cord from ischemia. Extradural
interconnections between segmental arteries also
provide rich collateral network.
The venous drainage of the spinal cord is via
the radially arranged intrinsic spinal cord veins
and superficial fine network of vessels along
with the longitudinal veins that form the extrinsic
spinal cord veins. These veins more or less follow
the arteries (i.e., the anterior and posterior median
spinal vein) but have many anastomoses that cre-
ate a network with commonly more than one
anterior and posterior vein [3]. They may also
use the roots as a vehicle to reach the epidural
plexus and the extraspinal veins and plexus with a
reflux-impeding mechanism within the dura mater
[4]. It is important to note that the transition of a
median vein into a radicular vein shows the same
hairpin shape as the artery. At the superior cervical
part, they can run through the occipital foramen to
connect to the vertebral plexus or the inferior
dural sinus. Drainage of blood from the spine
occurs through the valveless internal and external
venous vertebral plexus, which is connected to the
azygos and hemiazygos venous systems [5].
Spinal Vascular Malformation
Classification
Several classification schemes have been pro-
posed for spinal vascular malformations
[6–8]. Genetically they are classified as (1) hered-
itary, (2) genetic nonhereditary, and (3) sporadic
[9]. Hereditary lesion like hereditary hemorrhagic
telangiectasia caused by mutation in the vascular
germ cell results in multifocal vascular
malformations including spinal vascular
malformation along with cutaneous capillary
malformations. Genetic nonhereditary lesions
caused by somatic mutations share developmental
metameric links, such as the spinal arteriovenous
metameric syndromes that include Cobb syn-
drome, Klippel-Trenaunay syndrome, and Parkes
Weber syndrome. Sporadic malformations
include most other spinal vascular malformations.
The following four-type system is the clini-
cally relevant and most commonly used, with
1107
cavernous malformations and vascular tumors
added for completeness:
Type I: Dural arteriovenous fistula (dAVF)
Type II: Intramedullary arteriovenous malformation
Type III: Juvenile AVM
Type IV: Intradural perimedullary AVF
Intramedullary cavernous malformations and spi-
nal vascular tumors include hemangioma,
hemangioblastoma, metastatic tumors, aneurys-
mal bone cyst, osteoblastoma, angiosarcoma,
hemangiopericytoma, angiofibroma,
angiolipoma, and hemangioendothelioma.
Spinal vascular malformations are variable in
their epidemiology, clinical presentation and mor-
phology, imaging characteristics, natural history,
and management. In general the term angioarch-
itecture signifies the nature of the feeding arteries
(radiculomeningeal versus radiculopial or
radiculomedullary), the morphology of the spinal
AV shunt proper (nidus versus fistula, intranidal
aneurysms, etc.), and the nature of the draining
veins (medullary, perimedullary, radicular, epidu-
ral, or parachordal). The morphology of a high-
flow vascular malformation and the nature of the
draining veins may change with time as the
shunting causes increasing venous ectasias,
rerouting of blood, changes in the intranidal archi-
tecture, or recruitment of additional arterial
feeders, and this usually is associated with
change in clinical symptoms and imaging
characteristics.
Type I: Dural Arteriovenous Fistula
A spinal dAVF is an abnormal communication
between the radiculomeningeal artery in the
nerve root sleeve and a radicular vein with retro-
grade flow into the perimedullary venous system,
causing venous hypertension and congestion of
the spinal cord. The general belief is that the
glomerulus of Manelfe, a ball-like structure, is
believed to regulate regional venous and CSF
pressures and to be the site of fistula formation
[10]. The fistula typically consists of a single hole,
even when multiple feeders produce a “busy”
1108
appearance suggestive of a nidus. The fistulous
connection may drain centrifugally into the exter-
nal venous paraspinal network, in which case the
patient may be asymptomatic. However, drainage
into a radiculomedullary vein, and the longitudi-
nal venous network responsible for draining the
spinal cord, results in venous congestion of the
spinal cord.
Epidemiology
Spinal dAVFs are the most frequent spinal vascu-
lar malformation, representing approximately
70 % of all spinal vascular malformations
[11]. The exact etiology is unknown and it is
presumed to be an acquired lesion. Spinal
dAVFs have been associated with infection, syrin-
gomyelia, spine trauma, and surgery
[12–14]. Most fistulas affect the thoracolumbar
region with 85 % of lesions below T6 and 100 %
of lesions below T3 [15]. Most fistulas affect
middle-aged men (40–60 year) with a male/
female ratio of 5:1.
Pathophysiology
Spinal dAVFs are located on the dural sleeve of
the spinal nerve root where a radiculomeningeal
artery lies in close proximity to the radicular
vein. The radiculomeningeal artery normally sup-
plies the nerve root and the meninges and may
give rise to a radiculomedullary artery that sup-
plies the spinal cord, and the radicular vein drains
into the veins of the spinal canal. Shunting in
dAVF results in high flow in the valveless spinal
perimedullary veins through the radicular veins
resulting in diminished AV pressure gradient
thereby impeding the drainage of normal spinal
veins leading to venous congestion with
intramedullary edema. The slow flow in the cap-
illaries with stasis in the radiculomedullary arter-
ies results in chronic hypoxia and progressive
myelopathy.
Clinical Presentation
The clinical symptoms resulting from congestive
myelopathy are not specific, often slowly progres-
sive, and even insidious, so patients present with
very long histories of vague but progressive
symptoms. Mean duration of symptoms prior to
S. Paramasivam
diagnosis: 23 months [16]. The site of the fistula
has no relationship with the patient’s symptom-
atology because it is the spinal cord vascular con-
gestion, rather than fistula itself, that produces the
clinical syndrome.
The common symptoms include paraparesis,
followed by back pain that may radiate to the
lower legs, hypesthesia, paresthesias, impotence,
and sphincter disturbances [16, 17]. Patients may
report worsening of symptoms with exertion (neu-
rogenic claudication) or with certain postural
changes [15]. Less commonly, patients may have
acute onset of symptoms or periodic remissions as
the disease progresses. In rare instances, spinal
dAVF can present with subarachnoid hemorrhage.
Without therapy, spinal dAVF invariably results in
irreversible paraplegia or even quadriplegia.
Imaging
MRI
MRI is the screening procedure of choice, and
characteristic findings in spinal dAVF are cord
edema and dilated perimedullary vessels
[16]. T2-weighted (T2-WI) images typically
show poorly marginated centromedullary cord
hyperintensity that extends over multiple seg-
ments [17]. A hypointense rim surrounds the
congested cord and most likely reflects deoxygen-
ated blood within the congested capillaries
[18]. There may be contrast enhancement from
the chronic venous congestion in the post-
gadolinium T1-weighted (TI-WI) images
[19]. With disease progression, the cord will
become atrophic. On T2-WI imaging the dilated
and coiled perimedullary veins appear as flow
voids in most cases. Post-gadolinium T1-WI
images are useful in identifying low-flow dAVF
and differentiating from CSF pulsation artifact by
a typical salt-and-pepper appearance.
For localization of the fistula, noninvasive
diagnostic techniques, such as contrast-enhanced
MR angiography with relative fast acquisition
protocols, have been developed [20, 21]. Apart
from identifying the segmental level of the fistula,
it is also necessary to localize radiculomedullary
arteries especially the Adamkiewicz artery. This
prior knowledge can help to prevent neurologic
46 Spinal Vascular Malformations and Treatment
complications when performing therapeutic pro-
cedures. To depict and separate the submillimeter-
sized spinal cord arteries from the spinal cord
veins, a simultaneously high spatial and temporal
resolution should be achieved. A large
craniocaudal field of view must be used because
the location of the fistula is highly variable
irrespective of the findings previously obtained
by conventional MRI. Contrast-enhanced MR
angiography with large field of view can depict
the spinal cord arteries and veins [21].
Angiography
Digital subtraction angiography (DSA) is the gold
standard test for spinal dAVF [17]. Localization of
the spinal dural AVF can be very difficult. Selec-
tive angiography of the segmental artery harbor-
ing the AVF depicts the early venous filling and
retrograde flow of the opacified radiculo-
medullary veins into the spinal canal. An exten-
sive network of dilated tortuous perimedullary
veins is often visible. When a fistula is found,
the adjacent levels should be imaged as well
because of the possibility of multiple radicular
feeding arteries; most fistulas are located in the
thoracic and lumbar regions. In 10 % of cases, the
sacral arteries are involved [22] and rarely, intra-
cranial dural AVFs may drain inferiorly and
mimic spinal dural AVFs clinically and on MRI.
Neither the location of the pathologic vessels nor
the location of the intramedullary imaging abnor-
malities on MRI seems to be related to level of the
fistula Figs. 1 and 2.
Management
The natural history of untreated spinal dAVFs is
generally thought to be poor with progressive
neurological deficit. The spinal dAVF may be
treated either by surgery or endovascular
obliteration.
In most patients, the fistula may be obliterated
by super selective catheterization of the feeding
radiculomeningeal artery and obliteration of the
fistula using liquid embolic agents like n-butyl
cyanoacrylate (n-BCA) and Onyx. The embolic
agent must pass the nidus and reach and occlude
the proximal segment of the draining vein to pre-
vent subsequent intradural collateral filling of the
1109
fistula (Figs. 1 and 2). At the end of embolization,
the collateral circulation of the dural involved
region must be examined for the presence of nat-
ural arterial collaterals, ipsilateral above and
below the level of the fistula as well contralateral
at the same level of the spinal dAVF. The success
rate of endovascular therapy varies from 25 % to
75 % and recurrence is not uncommon. Up to
80 % of patients have clinical improvement of
various degrees to clinical stabilization following
embolization. Complications of embolization
include inadvertent closure of the spinal radiculo-
medullary artery mostly due to failure to recog-
nize. Inadvertent deposition of liquid embolic
material into the perimedullary veins beyond the
nidal-venous junction can result in thrombosis of
the spinal cord veins. In cases where endovascular
embolization is not possible, placement of a metal
coil in the major feeding artery can facilitate
intraoperative fluoroscopic localization of the
fistula [23].
Surgical ligation is performed by coagulation
and division of the intradural radicular vein that
receives blood from the shunt resulting in visible
change in venous turgor and color of the
arterialized venous plexus. This is more invasive
but a relatively simple and effective intervention
except for sacral fistulas. The key to successful
treatment is precise localization of the fistula.
Complete occlusion of the fistula by either
method stops disease progression, but only
two-thirds of the patients have improvement in
motor symptoms, while about one-third show
improvement of their sensory disturbances. In
most patients, impotence and sphincter distur-
bances do not recover. In patients with no
improvement within 4–6 weeks after treatment,
repeat MRI imaging should be done. MRI
changes may or may not completely reverse; if
persistent flow voids are visualized, it warrants
another angiographic evaluation.
Type II: Intramedullary Arteriovenous
Malformation
Intramedullary arteriovenous malformation
(AVM) consists of malformation within the

Fig. 1 Middle-aged man with subacute onset of progressive paraparesis with bladder
and bowel involvement had hyperintense signal in the MRI T2-WI (arrowhead) with
enlarged perimedullary veins (White Arrow) (a). T4 intercostal artery angiogram reveals
supply to the dural arteriovenous fistula (b–e). Super selective angiogram shows the
exact location of the fistula with multiple feeders (Black Arrow). The fistula is treated by
1110
n-BCA embolization with obliteration of the proximal segment of the vein just distal to
the fistula (Black Arrow in f). Control angiogram reveals total obliteration of the dural
arteriovenous fistula (g). Three months later MRI reveals significant improvement in
signal changes in the cord (arrowhead in h) and total obliteration of the dorsal
perimedullary veins
S. Paramasivam

Fig. 2 Right subclavian angiogram shows the cervical dural arteriovenous fistula
(arrow in a) supplied by a dural branch from thyrocervical trunk. Super selective
angiograms (b–e) reveal the anatomic architecture of the dural fistula. Arrowhead in
(c) shows the perimedullary venous drainage that drains intracranially. Arrow in (d, e)
represents the proximal segment of the vein just distal to the fistula that needs to be
targeted for embolization. Intra-procedural DYNA CT with contrast injection shows (f,
46
Spinal Vascular Malformations and Treatment
g) the anatomic location of the fistula in relation to the bone. The embolization
performed with n-BCA penetration into the proximal portion of the vein just distal to
the fistula (h) (Arrow). Control angiogram reveals total obliteration of the fistula (i).
Posttreatment MRI reveals complete obliteration of the fistula as evidenced by absence
of enlarged perimedullary veins (J)
1111
1112
substance of the spinal cord supplied by the
radiculomedullary and radiculopial arteries and
drained by the pial veins. They include spinal
cord glomerular AVMs with compact or diffuse
nidus and AV fistulas that may be intramedullary
or perimedullary in location.
Epidemiology
Intramedullary AVM is the second most common
spinal vascular lesion, accounting for about 20 %
of spinal vascular lesions. There is slight male
predominance and they commonly come to
clinical attention in the second and third
decade of life [24]. They are most commonly
located in the thoracolumbar region followed by
the cervical region. Syndromes associated
with intramedullary spinal AVMs include neuro-
fibromatosis and the Rendu-Osler-Weber,
Klippel-Trenaunay-Weber, and Parkes Weber
syndromes [25].
Pathophysiology
Glomerular spinal AVMs (nidal AVMs) are more
frequent of spinal cord AVMs. They are charac-
terized by a vascular nidus located intramedullary
and have multiple feeding arteries derived from
both the anterior and posterior spinal arterial sys-
tems. They drain into the dilated spinal cord veins.
They are distributed throughout the spinal cord:
about 30 % in the cervical region but most com-
monly found in the thoracolumbar region com-
prising of 70 % that includes conus medullaris
AVMs [24], a special category of spinal cord
AVMs that are attributed to an abnormality during
neurulation and may be associated with a tethered
cord [26]. Conus medullaris AVMs possess mul-
tiple arterial feeders and are extensive.
Clinical Presentation
Spinal AVMs may present acutely in most cases.
Hemorrhage is the most common presentation and
more so in children than adults [15, 27]. It may be
either subarachnoid or intraparenchymal hemor-
rhage. The typical syndrome of spinal hemorrhage
gives rise to severe pain, which frequently starts in
the back and rapidly spreads to the rest of the back
and nuchal area and to the legs. In severe hemor-
rhage in the cervical region, the blood can spread
S. Paramasivam
to the intracranial cavity and may cause
headaches and disturbances of consciousness.
Rehemorrhage is more common with spinal
AVMs than in the brain. It constitutes about
10 % in the first month and 40 % in the first year
after the event [28]. In adults, the correlation
between angioarchitecture and hemorrhage is
found as suggested by arterial, nidal aneurysm,
pial venous ectasia and congestion, and associated
arteriovenous fistulas.
Progressive symptoms are mostly nonspecific
with features of hypesthesia, paresthesia, weak-
ness, bladder and bowel dysfunction, impotence,
and diffuse back and muscle pain. These symp-
toms can have a stuttering course sudden worsen-
ing followed by some improvements over time.
The causes for deterioration are multiple. Among
the well-accepted causes are intraspinal hemor-
rhage, thrombosis within the AVM, arachnoiditis,
and increased venous pressure. AVMs may
become symptomatic by venous congestion
alone, in the absence of intraparenchymal or sub-
arachnoid hemorrhage [29].
Location of the AVM does have influence on
the clinical presentation: cervical lesions typically
present as progressive sensory deficits, whereas
thoracic AVMs most commonly present as hem-
orrhage (hematomyelia or subarachnoid hemor-
rhage) and often show progressive myelopathy.
Lesions of the conus and filum terminale
manifest as progressive myelopathy or acute
nonhemorrhagic paraplegia.
Aggravating factors including maneuvers that
increase intra-abdominal or intrathoracic pressure
such as the Valsalva maneuver and bending often
increase medullary venous pressure. Pregnancy is
known to aggravate symptomatology by increas-
ing the intra-abdominal pressure, increased intra-
vascular volume, hormonal changes, and added
stress on the venous circulation during delivery.
Trauma as a precipitating factor for symptoms is
not well supported.
Apart from neurologic symptoms, spinal
deformities, and complications of the spinal
cord, dysfunctions including urinary tract infec-
tions, respiratory infections, and decubitus ulcer-
ations are taken into consideration in determining
the morbidity and final outcome of these patients.
46 Spinal Vascular Malformations and Treatment
Imaging
MRI
On MRI the glomerular spinal AVMs show focal
dilatation of the cord in the region of the lesion
with conglomerate dilated, perimedullary, and
intramedullary vessels. There may be an area of
low signal intensity around the nidus on T1-WI
and T2-WI corresponding to hemosiderin deposi-
tion. The enlarged vessels appear as flow voids on
T2-WI imaging and as mixed hyperintense to
hypointense tubular structures on T1-weighted
(T1-WI) imaging (depending on their flow veloc-
ity and direction). The edema due to venous con-
gestion appears as cord swelling with
intramedullary hyperintensity on T2-WI
[29]. Intraparenchymal hemorrhages demonstrate
variable signal intensity on T1-WI and T2-WI
based on the age of hemorrhage.
MRI plays a significant role in defining the
location of the AVM in relation to the spinal
cord and dura and can also distinguish between
low-flow and high-flow forms of AVMs. As the
small arterial feeders and draining veins escape
detection on non-contrast MRI, contrast-
enhanced fast MR angiography is capable of
detecting the main arterial feeder of
glomerular-type AVMs. Contrast-enhanced MRI
is useful to detect subtle venous dilatations
[30, 31].
Angiography
Catheter angiography is the gold standard evalu-
ation for spinal cord AVMs [24]. A complete
angiogram characterizes all feeding vessels,
nidal configuration, flow characteristics,
intranidal aneurysms, and venous drainage with
or without venous ectasias. It is essential for treat-
ment plan. Selective catheterization and injection
of numerous arteries are essential to fully charac-
terize a spinal AVM. The feeding vessels may
arise from respective or distant radiculomedullary
arteries supplying the anterior or posterior spinal
arteries. The sources as far afield as the occipital,
ascending pharyngeal, vertebral, ascending and
deep cervical, supreme intercostal, intercostal,
lumbar, and lateral and median sacral arteries
[28] (Figs. 3 and 4).
1113
Management
The natural history of spinal AVMs is unknown.
In general, symptomatic spinal AVMs presenting
with hemorrhage or progressive neurologic deficit
warrant treatment to ameliorate symptoms and
improve the outcome. Due to the eloquence of
spinal cord and complex nature of spinal AVM
architecture, any modality of treatment procedure
has a relatively higher complication risk. Manage-
ment strategies have to be individualized based on
the available expertise.
The therapy of choice for all spinal cord AVMs
is endovascular embolization. Embolization is
performed after careful analysis of the AVM by
selective spinal angiography and by using an
embolic agent appropriate to the specific
angioarchitecture. Proximal arterial occlusion
will not obliterate the fistula, because of collateral
recanalization from the radiculopial network.
Obliteration of proximal venous segment should
be the target of embolization. In most cases partial
embolization is achieved with complete oblitera-
tion rates ranging from 25 % to 30 % [28, 32]. Par-
tial obliteration may be an adjunct to surgery; in
general, partial embolization improves the out-
come, with reduced rehemorrhage risk and
improved prognosis in most patients. Transient
and permanent complication rates following
embolization range from 10.6 % to 14 %
[28, 32] (Figs. 3 and 4).
Surgical management requires appropriate
case selection. Surgically accessible lesions with
compact nidus can be completely resected with a
cure rate of up to 94 % of cases [33].
Radiosurgery may be helpful in selected
patients as preliminary reports are encouraging.
However, it carries a risk of collateral acute and
chronic radiation-induced injury to the surround-
ing spinal cord [34].
Type III: Juvenile Arteriovenous
Malformation
Type III lesions aka spinal arteriovenous
metameric syndrome (SAMS) are complex
AVMs involving structures derived from two or
more embryonic layers. These AVMs involve the

Fig. 3 Intramedullary AVM with perimedullary venous drainage cranially (a–c) has
gliotic changes in the spinal cord. DSA reveals supply from the left vertebral artery
through anterior and posterolateral spinal arteries (d) and dominant supply from the left
radiculomedullary artery arising from the thyrocervical trunk supplying the anterior
spinal artery (e, f). The AVM has a dominant cranial venous drainage into the
1114
perimedullary venous system (arrow in g). Selective partial embolization of the AVM
was performed and control angiogram reveals significant reduction in the supply to the
malformation from the thyrocervical trunk with preservation of the anterior spinal axis
(h). There is residual supply to the malformation from the left vertebral artery (i)
S. Paramasivam
46 Spinal Vascular Malformations and Treatment
Fig. 4 Young patient with cervical intramedullary AVM
with fistula presented with cervical subarachnoid hemor-
rhage due to perimedullary large venous ectasia. DSA
examination shows anterior spinal artery from the right
vertebral artery supplying part of the AVM (a, b). The
dominant supply to the AVM is from the posterolateral
spinal artery arising from the left vertebral artery (c, d).
spinal cord, bony structures, paraspinal
musculature, subcutaneous tissue, and skin in
the same segment. The portion of the nidus
involving the spinal cord typically has neural tis-
sue within its interstices. Cobb syndrome is by
characterized skin, bone, and spinal cord involve-
ment aka cutaneous vertebral medullary
angiomatosis.
1115
Note the venous ectasia (arrow in c). Super selective
angiogram reveals the fistulous component of the AVM
with venous ectasia (e, f). Embolization of the AVM
performed with n-BCA, which is deposited in the AVM
and ectatic venous sac (g, h). Post-embolization control
angiogram reveals complete obliteration of the AVM (i–l)
Epidemiology
Type III lesions constitute about 6 % of all spinal
vascular malformations. They are equally located
in the cervical, thoracic, and lumbar or conus
medullaris regions. There is a 2:1 male to female
ratio. They come to clinical attention in childhood
and early adulthood. They may be multiple in as
high as 19 % of patients. They may be associated
1116
with Klippel-Trenaunay and Parkes Weber
syndromes [28].
Pathophysiology
Type III lesions are AVMs that involve derivatives
of all embryonic layers and consist of extensive
vascular malformation embedded within the tis-
sue. The segmented mesodermal origin of the
endothelial cells establishes the topographic link
between these segmentally related targets. The
spinal cord AVM is metameric if the myelomere
involved corresponds to that of the nerve
involved. The extraspinal and limb lesions are
usually high-flow arteriovenous fistulas supplied
by the segmental arteries like the intercostal or
lumbar arteries and corresponding segmental
arteries. The cutaneous lesions are fistulas of cap-
illary or arterial type with high-flow
characteristics.
Clinical Features
The clinical presentation may be due to spinal cord,
paraspinal, limb, or cutaneous lesions. The typical
presentation is in children or young adults with pain
and/or myelopathy [28]. Spinal cord involvement
may cause progressive myeloradiculopathy due to
venous congestion, compression due to enlarged
medullary veins, and spinal hemorrhage. The
paraspinal vascular malformations can present with
steal syndrome due to high-flow AVF. The cutane-
ous vascular malformations are readily apparent and
often present with a bruit. The comorbid conditions
such as spinal deformities such as kyphosis and
scoliosis along with complications common to spi-
nal cord dysfunction such as urinary tract infections,
respiratory infections, and decubitus ulcerations
must be taken into consideration in determining
the morbidity and final outcome of these patients.
Imaging
Multimodality imaging is the key in complete
analysis of these types of malformations. MRI is
essential to look for the anatomic extent of the
disease, the cord edema, and the intradural venous
engorgement. Catheter angiography assessment
can be difficult as only a part of the lesion can be
seen with injection of contrast into any particular
artery (Fig. 5).
S. Paramasivam
Management
These lesions are extremely difficult to treat. The
management of each individual case is tailored
based on the presentation. Endovascular emboli-
zation can be targeted at spinal cord lesions if
symptomatic as done for nonmetameric lesions.
Extraspinal lesions are treated by embolization if
needed; the key is to understand the pathophysi-
ology of the clinical presentation and preserve the
spinal cord vascular supply arising from the
radiculomedullary arteries. Although resection of
the lesion by combining staged embolization
followed by surgery has been reported, it is asso-
ciated with fatal complications and such an
approach is not generally aimed for [35, 36].
Type IV: Intradural Perimedullary
Arteriovenous Fistula
Type IV lesions are intradural perimedullary AV
fistulas located on the pial surface of the spinal
cord, usually on the anterior or lateral surface.
They consist of a fistula between a spinal cord
artery or arteries and the coronal venous plexus,
and there is often ectasia of the vein at the artery-
to-vein transition site. They are a spectrum of pial
AV malformations ranging from the most com-
mon single giant fistulae supplied by anterior or
posterolateral spinal arteries to larger, multiple-
vessel fistulae supplied by the anterior and pos-
terolateral spinal arteries.
Epidemiology
Type IV lesion accounts are detected in relatively
young patients (20–25 years) without sex predom-
inance and account for 20 % of all spinal vascular
lesions [15, 37]. An association with split cord
malformation and with Rendu-Osler-Weber and
Cobb syndrome is reported [38].
Pathophysiology
Type IV lesions occur most commonly at the
thoracolumbar junction, particularly at the
conus, and, to a lesser extent, in the upper cervical
region [25]. Fistulous AVMs are characterized by
direct AV shunts located superficially on the spi-
nal cord. There is usually no intramedullary
46 Spinal Vascular Malformations and Treatment 1117

Fig. 5 (continued)
1118
Fig. 5 Type III juvenile metameric AVM involving pre-
dominantly the mesodermal element. The child presented
with progressive weakness of all four limbs with predom-
inant lower motor neuron-type weakness in the upper
limbs with sensory disturbances. MRI revealed a meta-
meric AVM with venous congestion of the cord with
edema (b, c). The cervical spine X-ray reveals bony
destruction due to AVM (a). The DSA examination reveals
supply to the AVM in the vertebra and the soft tissue
component and no intervening nidus. Fistulous
spinal AVMs fed by the ventral spinal arterial
system are subpial while fistulous spinal AVMs
fed by the dorsal spinal arterial system are sub-
arachnoid in location. Fistulous spinal AVMs may
be in micro- and macrofistulous forms.
Microfistulous AVMs (mAVFs) comprise
about 75 % and have low shunt volume and are
mostly seen in adults, fed by the radiculo-
medullary arteries and drained by the superficial
perimedullary veins.
Macrofistulous AVMs (MAVFs) are typically
found in children due to high-flow nature comes
to clinical attention much earlier in life. There is
remodeling of the arteries and veins including
S. Paramasivam
around with feeders arising from vertebral arteries (d–g),
dorsal (arrow) and ascending cervical artery (arrowhead)
(h, i), and external carotid artery branches (j). Partial
embolization with Onyx and n-BCA (k, l) to reduce the
venous congestion was performed and posterior surgical
decompression to relieve pressure on the cord was
performed (o). Residual flow to the malformation is visu-
alized in the angiogram (m, n). Patient’s condition
stabilized
development of large variceal pouches due to
high shunt volume. They may be associated with
hereditary hemorrhagic telangiectasia (HHT).
Clinical Features
The clinical presentation depends on the micro- or
macrofistulous forms and the location. The
mAVFs present during adulthood; the thoracic
mAVFs present with hemorrhage (two-third
intraparenchymal and one-third subarachnoid);
cervical, conus, and cauda equina mAVFs present
with slowly progressive, asymmetric myeloradi-
culopathy of the conus and cauda equina [37]. The
MAVFs present earlier in life and the common
presentation is progressive myeloradiculopathy.
46 Spinal Vascular Malformations and Treatment
Imaging
MRI characteristically demonstrates large flow
voids [38] that comprise the enlarged arterial feeder
and the dilated perimedullary venous channels and
associated cord changes. However, low-flow
microfistulous forms may not be apparent on regu-
lar MRI [39], while contrast-enhanced fast MR
angiography is capable of detecting the feeding
arteries. Apparent diffusion coefficient (ADC)
values may be reduced, indicating edema, and it
may normalize after treatment [40]. Catheter angi-
ography is the gold standard and should be done if a
type IV lesion is suspected but not seen on MRI.
Angiography is necessary to make a clear diagnosis
and understand the architecture of a perimedullary
fistula and plan therapy (Figs. 6 and 7).
Management
Type IV fistulous lesions are rare, and estimations
of the natural history of untreated lesions suggest
progression from myelopathy to paraplegia within
5–7 years, and a high incidence of repeated hemor-
rhage in patients presenting with hemorrhage [15,
25, 37]. Prompt diagnosis and treatment of
perimedullary fistulae is mandatory with the objec-
tive of treatment being occlusion of the fistula. Most
lesions can be effectively obliterated; endovascular
embolization targeted at the arteriovenous transition
leads to complete exclusion of the lesion with relief
of the venous congestion and clinical improvement
(Figs. 6 and 7). In cases where the embolization
appears to be associated with higher risk or is
deemed to be technically infeasible because of
small feeders or tortuous long feeding arteries
such as in filum terminale lesions, surgery should
be considered. The biggest challenge with surgery
is the localization of the fistulous location, which is
usually embedded, in the congested veins.
The clinical outcome appears to be somewhat
variable, with partial neurological improvement or
no change occurring in the majority of patients, and
worsening in a small minority of patients [37, 41].
Cavernous Malformation (Cavernomas)
Spinal cord cavernomas are intramedullary vascu-
lar malformations composed of dilated thin-
1119
walled capillaries and sinusoids with no interven-
ing parenchymal tissue, without the identifiable
feeding artery or specific draining vein.
Epidemiology
Cavernomas are distributed along the entire
neuraxis and represent 5–12 % of all spinal vas-
cular lesions [42]. Women are affected more than
men with a ration of 2:1. They usually come to
clinical attention in the middle age with a peak in
the fourth decade but may be found in all age
groups. Familial occurrence, coexistent meta-
meric vascular nevi, and concurrent intracranial
lesion are described in as many as 40 % of patients
[43]. As in the brain, they are frequently associ-
ated with a venous angioma. The thoracic spine is
more often affected than the cervical spine.
Pathophysiology
Spinal cavernomas may cause symptoms acutely
by intramedullary or subarachnoid hemorrhage or
chronically with progressive myelopathy
resulting from lesion expansion by capillary dila-
tation or repetitive small hemorrhages in or
around the lesion or due to toxic effects of degra-
dation products of the intramedullary blood.
Clinical Presentation
Spinal cavernomas may cause acute, recurrent,
and/or progressive myelopathy. The acute clinical
presentation is usually acute pain followed by
sensorimotor deficit indicating new hemorrhage
within or around the lesion. Slowly progressive
sensorimotor deficit and bladder and bowel dis-
turbance likely reflect repeated episodes of minor
hemorrhage or toxic effects of blood products on
the surrounding spinal cord tissue. The course of
symptoms is highly variable and the natural his-
tory is far from clear. Assessments of the annual
rate of symptomatic rehemorrhage rates range
from 0 % to 66 % [44].
Pathology
Grossly, spinal cavernomas are discrete, well-
circumscribed, lobulated, red-purple lesions.
Microscopically, they are identical to those in the
brain and are composed of dilated, thin-walled
capillaries that have a simple endothelial lining
1120 S. Paramasivam

Fig. 6 Child with

paraplegia revealed a spinal
pial arteriovenous fistula
with dilated perimedullary
vein (a). DSA examination
revealed an enlarged
radicular artery and anterior
spinal artery that supply the
fistula on the pial surface
(b) with enlarged
perimedullary veins
draining into the veins
intracranially (c). Super
selective catheterization of
the fistula (d) with coiling
of the venous sac (e) to
reduce the flow in the fistula
followed by embolization
of the fistula using n-BCA
(f) resulted in successful
obliteration of the fistula

with variably thin fibrous adventitia and may have
calcifications. In the spinal cord around the
cavernoma, residua of previous hemorrhage may
be present, including scarring, collections of
hemosiderin-laden macrophages, variable degrees
of gliosis, and edema.
Imaging
CT is of limited value as in most other spinal
pathologies; only acute hemorrhagic lesions and
calcified lesions are detected and others remain
indetectable. On MRI, cavernomas appear as
well-defined, circumscribed lesions of varying
size. On T2-WI they characteristically display a
hypointense rim due to susceptibility artifacts
from hemosiderin deposits [45].
The inhomogeneous central core represents
hemorrhage in different stages of evolution giving
a “popcorn” appearance to the center [42]. The
size of the cavernoma and its precise relation to
the surface of the cord is best visualized on
T1-WI, and T2-WI tend to overestimate the size
of the cavernoma, because of the hemosiderin
ring. If multiple, cavernomas are readily picked
46 Spinal Vascular Malformations and Treatment
Fig. 7 Young boy with progressive pial arteriovenous
fistula (a–c) was treated surgically initially with proximal
occlusion of the radicular artery at T4 level (clip in d). He
had recurrence of the symptoms, as the fistula was not
treated. Repeat angiogram reveals the reconstitution of
the anterior spinal artery from above and below along the
up with MRI. Spinal cord cavernomas may also
mimic spinal cord tumors such as glioma,
ependymoma, or anaplastic astrocytoma, espe-
cially if they are large (Fig. 8).
The current noninvasive imaging techniques
cannot distinguish acute spinal cord hemorrhage
due to solitary cavernoma or a micro-AVM. In
such a situation, in spite of the low yield, spinal
angiography is essential to pick up AVM, as
cavernomas are angiographically occult.
Management
Asymptomatic cavernomas are usually followed
by serial imaging studies and not treated. The
natural history is highly variable and the consen-
sus is to treat the symptomatic cavernomas.
1121
axis to supply the fistula (e, f). Super selective catheteriza-
tion up to the fistulous point and coil embolization of the
fistula were performed with preservation of the anterior
spinal axis (g, h). Follow-up angiogram 6 months later
reveals persistent obliteration of the fistula with preserva-
tion of the anterior spinal axis (i–k)
Surgical resection is the treatment of choice and
can be performed safely using meticulous micro-
surgical technique. Careful localization,
myelotomy, and meticulous dissection are key to
good functional outcome. Endovascular treat-
ments of cavernomas are inappropriate.
Summary
Spinal vascular malformations are a group of var-
ied disorders due to developmental derangement
of the vascular system that may present acutely
due to hemorrhage or subacutely due to venous
hypertension or vascular steal with nonspecific
neurologic symptomatology making diagnosis
1122
Fig. 8 Cervical spinal cord
cavernoma that came to
clinical attention due to
hemorrhage. T2-WI
displays a hypointense rim
due to susceptibility
artifacts from hemosiderin
deposits. The central core
represents hemorrhage in
different stages of evolution
(a–c)
challenging for neurologists, neurosurgeons, and
neuroradiologists. Spinal vascular malformations
are rare but treatable causes of progressive spinal
cord symptoms. The pathophysiology of many
spinal vascular malformations is determined by
their venous drainage. Elevated intravenous pres-
sure impedes blood flow, leading to stasis, pro-
gressive venous congestion, congestive venous
edema, and cord ischemia.
When spinal vascular diseases are suspected,
MRI is the first diagnostic modality to identify the
lesion and rule out potential differential diagnoses
like acute cord compression, tumor, degenerative
diseases of the spine, myelitis, etc. Initial diagno-
sis depends on demonstration of pathologic flow
voids, cord edema, or spinal hemorrhage and pro-
vides an initial assessment of which malformation
to expect. Contrast-enhanced fast-spin echo MR
angiography of the spinal cord constitutes a help-
ful adjunct to noninvasive diagnostics because it
may demonstrate pathologic AV shunts and help
to guide selective spinal arteriography.
When neurologic symptoms and MRI suggest
a vascular malformation, spinal angiography is
the next and gold standard test to define the type
of vascular malformation and decide on the
appropriate therapy. Most spinal vascular
malformations that come to clinical attention
S. Paramasivam
need treatment as the natural history is suggestive
of progressive stepwise neurological deteriora-
tion. Early diagnosis and timely management in
symptomatic patients may result in improvement
or stabilization of clinical conditions. Treatment
should be performed in specialized centers.
Except spinal cord cavernomas, all other vascular
disorders of the spine are initially managed by
endovascular embolization as it is the least inva-
sive but technically demanding. For cases that fail
embolization, surgical treatment is an available
option. In select cases a combined therapy might
be sensible.
References
1. Thron A (1988) Vascular anatomy of the spinal cord:
neuroradiological investigations and clinical syn-
dromes. Springer, Berlin
2. Lasjaunias PL, Berenstein A, ter Brugge K (2004)
Surgical neuroangiography. Clinical vascular anatomy
and variations, vol 1, 2nd edn. Springer, Berlin
3. Grunwald I, Thron A, Reith W (2001) Spinal angiog-
raphy: vascular anatomy, technique and indications.
Radiologe 41:961–967 [in German]
4. Thron A, Otto J, Schroeder JM (1990) Functional
anatomy of the dural segment of spinal cord draining
veins: a histological and microangiographic study. In:
du Boulay G (ed) Symposium neuroradiologicum.
Springer, London, p 323
46 Spinal Vascular Malformations and Treatment
5. Mull M, Thron A (2006) Spinal infarcts. In: von
Kummer R, Back T (eds) Magnetic resonance imaging
in ischemic stroke. Springer, Berlin, pp 251–267
6. Borden JA, Wu JK, Shucart WA (1995) A proposed
classification for spinal and cranial dural arteriove-
nous fistulous malformations and implications for
treatment. J Neurosurg 82:166–179. doi:10.3171/
jns.1995.82.2.0166
7. Spetzler RF, Detwiler PW, Riina HA, Porter RW
(2002) Modified classification of spinal cord vascular
lesions. J Neurosurg 96:145–156
8. Zozulya YP, Slin’ko EI, Al Q II (2006) Spinal arterio-
venous malformations: new classification and surgical
treatment. Neurosurg Focus 20:E7
9. Krings T et al (2007) Imaging in spinal vascular dis-
ease. Neuroimaging Clin N Am 17:57–72.
doi:10.1016/j.nic.2007.01.001
10. Manelfe C, Lazorthes G, Roulleau J (1972) Arteries of
the human spinal dura mater. Acta Radiol Diagn
13:829–841
11. Gilbertson JR, Miller GM, Goldman MS, Marsh WR
(1995) Spinal dural arteriovenous fistulas: MR and
myelographic findings. AJNR Am J Neuroradiol
16:2049–2057
12. Finsterer J, Bavinzski G, Ungersbock K (2000) Spinal
dural arteriovenous fistula associated with syringomy-
elia. J Neuroradiol 27:211–214
13. Asakuno K, Kim P, Kawamoto T, Ogino M (2002)
Dural arteriovenous fistula and progressive conus
medullaris syndrome as complications of lumbar
discectomy. Case report. J Neurosurg 97:375–379
14. Vankan Y, Demaerel P, Heye S, Van Calenbergh F, van
Loon J, Maleux G, Wilms G (2004) Dural arteriove-
nous fistula as a late complication of upper cervical
spine fracture. Case report. J Neurosurg 100:382–384
15. Rosenblum B, Oldfield EH, Doppman JL, Di Chiro G
(1987) Spinal arteriovenous malformations: a compar-
ison of dural arteriovenous fistulas and intradural
AVM’s in 81 patients. J Neurosurg 67:795–802.
doi:10.3171/jns.1987.67.6.0795
16. Atkinson JL et al (2001) Clinical and radiographic
features of dural arteriovenous fistula, a treatable
cause of myelopathy. Mayo Clin Proc 76:1120–1130.
doi:10.4065/76.11.1120
17. Koch C, Kucinski T, Eckert B, Rother J, Zeumer H
(2003) Spinal dural arteriovenous fistula: clinical and
radiological findings in 54 patients. Rofo
175:1071–1078. doi:10.1055/s-2003-40925
18. Hurst RW, Grossman RI (2000) Peripheral spinal cord
hypointensity on T2-weighted MR images: a reliable
imaging sign of venous hypertensive myelopathy.
AJNR Am J Neuroradiol 21:781–786
19. De Marco JK, Dillon WP, Halback VV, Tsuruda JS
(1990) Dural arteriovenous fistulas: evaluation with
MR imaging. Radiology 175:193–199. doi:10.1148/
radiology.175.1.2315480
20. Huffmann BC, Gilsbach JM, Thron A (1995) Spinal
dural arteriovenous fistulas: a plea for neurosurgical
treatment. Acta Neurochir 135:44–51
1123
21. Lee TT, Gromelski EB, Bowen BC, Green BA (1998)
Diagnostic and surgical management of spinal dural
arteriovenous fistulas. Neurosurgery 43:242–246, dis-
cussion 246–247
22. Partington MD, Rufenacht DA, Marsh WR, Piepgras
DG (1992) Cranial and sacral dural arteriovenous fis-
tulas as a cause of myelopathy. J Neurosurg
76:615–622. doi:10.3171/jns.1992.76.4.0615
23. Britz GW, Lazar D, Eskridge J, Winn HR (2004)
Accurate intraoperative localization of spinal
dural arteriovenous fistulae with embolization coil:
technical note. Neurosurgery 55:252–254, discussion
254–255
24. Niimi Y, Berenstein A (1999) Endovascular treatment
of spinal vascular malformations. Neurosurg Clin N
Am 10:47–71
25. Djindjian M, Djindjian R, Hurth M, Rey A, Houdart R
(1977) Spinal cord arteriovenous malformations and
the Klippel-Trenaunay-Weber syndrome. Surg Neurol
8:229–237
26. Hurst RW, Bagley LJ, Marcotte P, Schut L, Flamm ES
(1999) Spinal cord arteriovenous fistulas involving
the conus medullaris: presentation, management,
and embryologic considerations. Surg Neurol
52:95–99
27. Tobin WD, Layton DD (1976) The diagnosis and nat-
ural history of spinal cord arteriovenous
malformations. Mayo Clin Proc 51:637–646
28. Berenstein A, Lasjaunias P, TerBrugge KG (2004)
Surgical neuroangiography: clinical and endovascular
treatment aspects in adults, vol 2.2, 2nd edn. Springer-
Verlag, Germany, pp 737–872
29. Kataoka H, Miyamoto S, Nagata I, Ueba T, Hashimoto
N (2001) Venous congestion is a major cause of neu-
rological deterioration in spinal arteriovenous
malformations. Neurosurgery 48:1224–1229, discus-
sion 1229–1230
30. Doppman JL, Di Chiro G, Dwyer AJ, Frank JL,
Oldfield EH (1987) Magnetic resonance imaging of
spinal arteriovenous malformations. J Neurosurg
66:830–834. doi:10.3171/jns.1987.66.6.0830
31. Thron A, Mull M, Reith W (2001) Spinal arteriovenous
malformations. Radiologe 41:949–954
32. Rodesch G, Lasjaunias P, Berenstein A (1993) Embo-
lization of arteriovenous malformations of the spinal
cord. In: Valavanis A (ed) Interventional neuroradiol-
ogy. Springer, Berlin, p 135
33. Connolly ES Jr, Zubay GP, McCormick PC, Stein BM
(1998) The posterior approach to a series of glomus
(type II) intramedullary spinal cord arteriovenous
malformations. Neurosurgery 42:774–785, discussion
785–776
34. Sinclair J, Chang SD, Gibbs IC, Adler JR Jr
(2006) Multisession CyberKnife radiosurgery for
intramedullary spinal cord arteriovenous malformations.
Neurosurgery 58:1081–1089. doi:10.1227/01.
NEU.0000215891.25153.BA, discussion 1081–1089
35. Menku A, Akdemir H, Durak AC, Oktem IS (2005)
Successful surgical excision of juvenile-type spinal
1124
arteriovenous malformation in two stages following
partial embolization. Minim Invasive Neurosurg
48:57–62. doi:10.1055/s-2004-830184
36. Spetzler RF, Zabramski JM, Flom RA (1989) Manage-
ment of juvenile spinal AVM’s by embolization and
operative excision. Case report. J Neurosurg
70:628–632. doi:10.3171/jns.1989.70.4.0628
37. Mourier KL, Gobin YP, George B, Lot G, Merland JJ
(1993) Intradural perimedullary arteriovenous fistulae:
results of surgical and endovascular treatment in a
series of 35 cases. Neurosurgery 32:885–891, discus-
sion 891
38. Halbach VV, Higashida RT, Dowd CF, Fraser KW,
Edwards MS, Barnwell SL (1993) Treatment of giant
intradural (perimedullary) arteriovenous fistulas. Neu-
rosurgery 33:972–979, discussion 979–980
39. Dillon WP, Norman D, Newton TH, Bolla K, Mark A
(1989) Intradural spinal cord lesions: Gd-DTPA-
enhanced MR imaging. Radiology 170:229–237.
doi:10.1148/radiology.170.1.2909101
40. Inoue T, Takahashi T, Shimizu H, Matsumoto Y,
Takahashi A, Tominaga T (2006) Congestive
myelopathy due to cervical perimedullary arteriove-
nous fistula evaluated by apparent diffusion
S. Paramasivam
coefficient values – case report. Neurol Med Chir
46:559–562
41. Hida K, Iwasaki Y, Goto K, Miyasaka K, Abe H (1999)
Results of the surgical treatment of perimedullary arte-
riovenous fistulas with special reference to emboliza-
tion. J Neurosurg 90:198–205
42. Deutsch H, Jallo GI, Faktorovich A, Epstein F (2000)
Spinal intramedullary cavernoma: clinical presentation
and surgical outcome. J Neurosurg 93:65–70
43. Cohen-Gadol AA, Jacob JT, Edwards DA, Krauss WE
(2006) Coexistence of intracranial and spinal cavern-
ous malformations: a study of prevalence and natural
history. J Neurosurg 104:376–381. doi:10.3171/
jns.2006.104.3.376
44. Sandalcioglu IE, Wiedemayer H, Gasser T, Asgari S,
Engelhorn T, Stolke D (2003) Intramedullary spinal
cord cavernous malformations: clinical features and
risk of hemorrhage. Neurosurg Rev 26:253–256.
doi:10.1007/s10143-003-0260-2
45. Weinzierl MR, Krings T, Korinth MC, Reinges MH,
Gilsbach JM (2004) MRI and intraoperative findings
in cavernous haemangiomas of the spinal cord.
Neuroradiology 46:65–71. doi:10.1007/s00234-003-
1072-3
 Spinal Cord Infarction and Differential
Diagnosis 47
Srikanth R. Boddu, Alessandro Cianfoni, Kyung-Wha Kim,
Mohammad Amin Banihashemi, Emanuele Pravatà,
Y. Pierre Gobin, and Athos Patsalides

Contents Ischemic Spinal Cord Infarction . . . . . . . . . . . . . . . . . . . 1132

Hemorrhagic Spinal Cord Infarction . . . . . . . . . . . . . . . 1138
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1127 Spinal Venous Infarction . . . . . . . . . . . . . . . . . . . . . . . . . . . 1138
Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1127 Pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139
History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1127 Concepts of Primary and Secondary injury . . . . . . . . 1140
Spinal Venous Infarction . . . . . . . . . . . . . . . . . . . . . . . . . . . 1142
Arterial Anatomy of the Spine . . . . . . . . . . . . . . . . . . . 1127 Imaging Correlation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1142
Segmental Arteries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1128
Radicular Arteries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1128 Clinical Presentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1143
Ischemic Spinal Cord Infarction . . . . . . . . . . . . . . . . . . . 1143
Arterial Supply of Spinal Cord . . . . . . . . . . . . . . . . . . 1128 Hemorrhagic Spinal Cord Infarction . . . . . . . . . . . . . . . 1145
Superficial Arterial System of the Spinal Cord . . . . 1128 Venous Hypertension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1145
Intrinsic Arterial System . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1131
Imaging Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1145
Watershed Territories of the Spinal Cord . . . . . . . 1131 Evolution of Spine Imaging . . . . . . . . . . . . . . . . . . . . . . . . 1145
Longitudinal Anastomosis . . . . . . . . . . . . . . . . . . . . . . . . . . 1131 Standard Spine Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1145
Conventional MRI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1145
Etiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1132
Diffusion-Weighted Imaging (DWI)
of the Spinal Cord . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1146
Typical MRI Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1147
Associated Vertebral Body Infarction . . . . . . . . . . . . . . 1149
S.R. Boddu (*) • M.A. Banihashemi • Holistic Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1149
Y.P. Gobin • A. Patsalides Advanced Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1149
Department of Neurological Surgery, Weill Cornell Follow-Up Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1149
Medical Center, New York, NY, USA
e-mail: srb9017@med.cornell.edu;
mob2009@med.cornell.edu; yug2001@med.cornell.edu;
atp9002@med.cornell.edu
A. Cianfoni
Department of Radiology, Medical University of South
Carolina, Charleston, SC, USA
e-mail: cianfoni@musc.edu
K.-W. Kim
New York Presbyterian Hospital, Weill Cornell Medical
College of Cornell University, New York, NY, USA
e-mail: kyk2003@nyp.org
E. Pravatà
Neuroradiology Department, Neurocenter of Italian
Switzerland, Lugano, Switzerland
e-mail: emanuele.pravata@gmail.com

# Springer Science+Business Media New York 2016 1125

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_30
1126 S.R. Boddu et al.

Differential Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1150 Abstract

Multiple Sclerosis (MS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1150 Spinal cord infarction is a rare disease and
Idiopathic Transverse Myelitis (ITM) . . . . . . . . . . . . . . 1151
Acute Disseminated Encephalomyelitis
constitutes one of the acute spinal emergencies.
(ADEM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1152 In comparison to its cerebral counterpart, the
Spinal Cord Arteriovenous Shunts . . . . . . . . . . . . . . . . . 1153 spinal cord infarction has extremely low inci-
Spinal Cord Neoplasms . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1156 dence, possibly related to the abundance of the
Role of Angiography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1158 arterial anastomosis and low incidence of the
Challenges in Spinal Angiography . . . . . . . . . . . . . . . . . 1158 atherosclerosis in the spinal arteries. Since the
Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1169 first spinal cord infarction reported in early
Spinal Cord Ischemia Secondary to Embolism . . . . 1170 nineteenth century, there has been remarkable
Spinal Cord Ischemia Secondary progress in the understanding of this disease
to Hypoperfusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1170
Treatment of Underlying Cause . . . . . . . . . . . . . . . . . . . . 1171 entity. However, the fact that there is no
Role of High-Dose Steroids . . . . . . . . . . . . . . . . . . . . . . . . 1171 established standard of care treatment as of
Surgical Decompression . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1171 today highlights the complexity and challeng-
Prognosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1172 ing nature of this disease. In this chapter the
Short-Term Outcome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1172 arterial anatomy of the spinal cord,
Long-Term Outcome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1172 etiopathogenesis, and clinical presentation of
Future Direction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1172 the cord infarction with specific emphasis on
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1173
the imaging findings and differential diagnosis
are discussed. The role of catheter angiography
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1173
was specifically addressed along with the
merits and pitfalls of the noninvasive angiog-
raphy. The chapter concludes a discussion on
the current and potential future treatment
strategies.

Keywords
Watershed territory • Thrombolysis • Myelitis •
Infarction • Digital subtraction angiography •
Embolization • Signal-to-noise ratio • Diffu-
sion-weighted imaging • Fast spin echo • Gra-
dient echo • Diffusion tensor imaging • Fiber
tractography • Echo-planar imaging • Apparent
diffusion coefficient • Anterior spinal artery •
Posterior spinal artery • Segmental artery •
Radicular artery • Artery of Adamkiewicz •
Multiple sclerosis • Neuromyelitis optica • Idi-
opathic transverse myelitis • Acute dissemi-
nated encephalomyelitis • Acute
polyneuropathy • Subacute combined degener-
ation • Neurosarcoidosis • Neuro-Behcet’s dis-
ease • AIDS-associated myelopathy •
Radiation myelitis
47 Spinal Cord Infarction and Differential Diagnosis
Introduction
Spinal cord infarction is a diverse entity broadly
categorized into vascular syndromes based on
etiology and clinical presentation. Acute spinal
cord ischemia constitutes 5–8 % of all acute
myelopathies and only 1–2 % of all vascular neu-
rological pathologies [1]. Still spinal cord infarc-
tion (SCI) represents one of the most common
causes of acute non-compressive myelopathies
[2]. Despite extensive diagnostic efforts,
35–60 % [1, 3] of acute spontaneous spinal cord
syndromes remain etiologically ambiguous. Thor-
ough understanding of the etiology, pathophysi-
ology, limitations of the conventional imaging,
and poor prognosis from delayed treatment is
vital in the care of SCI patients. Recent advances
in spinal cord imaging such as diffusion-weighted
imaging (DWI), lessons learnt from the extensive
evidence of the brain stroke thrombolysis, and
cutting-edge neurointerventional techniques may
define an effective early treatment of spinal cord
ischemia in the near future.
Epidemiology
Spinal cord infarction (SCI) is a rare entity with
limited epidemiological data. In a largest autopsy
series of 3,737 consecutive postmortem studies
during 1909–1958, only 11 cases of SCI are
reported, of which nine are ischemic and two
hemorrhagic in nature [4]. The absence of com-
mon etiological factors such as atherosclerosis
and hypertensive vascular disease in these
11 patients is debatable for its completeness and
its representation of the current population. This
low number may be due to the lack of routine
examination of the spinal cord in autopsies, as
19 cases of SCI in a 6-year interval were reported
in the early twentieth century [5]. In another series
of autopsies of asymptomatic patients from the
1970s, an incidental rate of 3 % was reported
[6]. Currently the incidence of spinal cord infarc-
tion is estimated to account for 1.2 % of all stroke
admissions in the United States [7]. Based on
available data on stroke admissions in the United
1127
States, this would amount to 5,000–8,000 cases of
SCI a year [8]. However, none of these series
include post-aortic surgery patients, which is cur-
rently the most common etiology [7, 9, 10]
for SCI.
History
Autopsy reports consistent with spinal cord
infarction, referred to as spinal softening, had
been published as early as 1831 [11]. At that
time spinal cord infarction cases were categorized
as acute myelitis. The earliest clinical description
of this condition, consistent with spinal cord syn-
drome, was described in literature in 1842
[12]. Prior to 1866 it was widely believed that
the underlying etiology of such case was inflam-
mation, as identified by autopsy reports that
showed a “softening” of the spinal cord. Bastian
et al. first described the term “softening” for brain
tissue in 1866, and later in 1882 it was suggested
by the same author that softening of the spinal
cord was caused by thrombosis of spinal vessels
rather than inflammation of the cord, although this
etiology was not widely accepted for more than a
decade after its description [5].
Arterial Anatomy of the Spine
A detailed knowledge of the vascular anatomy of
the spine and spinal cord is a prerequisite for
understanding the pathophysiology of spinal
cord infarction. A thorough understanding on vas-
cular anatomy helps to better understand how
lifesaving treatments of thoracoabdominal aortic
aneurysm and intra-arterial embolization proce-
dures can turn into a risk factor for the spinal
cord infarction. Also this is vital for planning of
safe surgical and endovascular interventions. Dig-
ital subtraction angiography is the gold standard
for spinal vascular imaging. Increasingly, mag-
netic resonance angiography (MRA) with time-
resolved imaging has gained widespread accep-
tance in the localization of the spinal vascular
malformations and promising results in the
1128
detection of large vascular structures, such as the
artery of Adamkiewicz.
Segmental Arteries
The arterial supply to the spinal column,
paraspinal muscles, dura, nerve roots, and spinal
cord derives from the segmental arteries. The seg-
mental arteries in the thoracic and upper lumbar
spine originate in pairs from the posterior aspect
of the descending aorta adjacent to the spinal
column. The segmental arteries usually include
one pair of supreme or superior intercostal arteries
(above T3), nine pairs of posterior intercostal
arteries (T3–T11), one pair of subcostal arteries
(T12), and four pairs of lumbar arteries (L1–L4)
[13]. The superior/supreme intercostal artery
arises from aorta or costocervical trunk or verte-
bral artery. Segmental arteries in the lower lumbar
and sacral region originate from branches of the
internal iliac artery (mainly the iliolumbar and
lateral sacral arteries) and the median sacral artery
(branch of the aorta at the level of the bifurcation),
providing arterial supply to the L5 vertebra and
the sacrum. The distribution of the segmental
arteries is summarized in Table 1.
It is important to remember that the segmental
arteries are named by the level they supply and not
Table 1 Summary of distribution of segmental arteries
Cervical spine
Vertebral arteries
Deep and ascending cervical arteries
Thoracic and upper lumbar spine
Subclavian arteries via deep cervical artery
T1 and T2 segmental/intercostal arteries
Aorta
Thoracic segmental arteries (intercostal)
One pair of superior/supreme intercostal artery
(above T3)
One pair at each vertebral level from T3–T12
Lumbar segmental arteries
One pair at each vertebral level from L1–L4
Lower lumbar spine and sacrum
Internal iliac artery (iliolumbar and lateral sacral
arteries)
Median sacral artery (branch of the aorta)
S.R. Boddu et al.
by the level from which they originate. Segmental
arteries supply all of the dorsolateral tissues of a
single metamere, except the spinal cord. There are
extensive anastomoses between segmental arter-
ies with important connections both above and
below a given level, as well as contralaterally.
The segmental arteries divide into three major
trunks: (i) lateral or ventral (posterior intercostal
or lumbar artery), (ii) middle or dorsal (muscular
and cutaneous branches), and (iii) medial or
spinal [13].
Radicular Arteries
Radicular artery along with the anterior and
posterior spinal arteries branches out from the
medial (spinal) trunk, after entering the spinal
canal at intervertebral foramen. Further at each
nerve root level, the radicular artery ramifies
[14] into:
1. Radiculoradial artery – supplies both the dura
and the nerve roots.
2. Radiculomeningeal artery – supplies only
the dura.
3. Radiculomedullary artery – supplies the spinal
cord. Unlike other two branches, radiculo-
medullary branches are present only at some
nerve root levels.
Arterial Supply of Spinal Cord
1. Superficial or external arterial system
2. Deep or intrinsic arterial system
Superficial Arterial System
of the Spinal Cord
At the surface of the cord, two arterial systems can
be described (Fig. 1):
1. Longitudinal arterial trunks: extend along the
long axis of the spinal cord and are constituted
by one anterior spinal artery and two posterior
47 Spinal Cord Infarction and Differential Diagnosis
Fig. 1 Arterial supply of the spinal cord: (1) anterior
spinal artery, (2) posterior spinal artery, (3) spinal trunk
of segmental artery, (4) anterior radiculomedullary artery,
(5) posterior radiculomedullary artery, (6) central (sulcal)
arteries, (7) vasocorona
(or posterolateral) spinal arteries, which
receive flow from the anterior and posterior
radiculomedullary arteries, respectively
2. Pial plexus: covers the periphery of the spinal
cord
Longitudinal Arterial Trunks
Anterior Spinal Artery
The anterior spinal artery (ASA) is typically
formed at the level of the foramen magnum by
the confluence of descending branches of the
intracranial segments of the vertebral arteries
(above the posterior inferior cerebellar artery
(PICA)). The descending branches from both ver-
tebral arteries may join at the C2–C4 level [15]
and travel along the anterior sulcus of the spinal
cord to the conus medullaris. ASA has variable
caliber (diameter 0.2–0.8 mm); it is thinnest in the
thoracic cord and thickest in the region of the
conus medullaris [15]. The ASA supplies the
anterior two-thirds of the spinal cord tissue
(including the anterior horns and the
spinothalamic and corticospinal tracts) by central
and pial branches.
The ASA is usually continuous, and during its
long course, it receives additional arterial supply
1129
via the anterior radiculomedullary arteries in order
to maintain adequate blood flow to the entire
spinal cord. As a result, the ASA should not be
thought of as a single straight artery but rather as a
consecutive series of anastomotic vascular
loops [16].
ASA receives major contribution from the
radiculomedullary arteries at three regions:
cervicothoracic, mid-thoracic, and thoracolumbar,
with watershed areas at the border of each region,
especially in the upper thoracic region. The most
important anterior radiculomedullary arteries and
most easily recognized in angiography are the
artery of cervical enlargement in the cervical and
upper thoracic region and artery radiculome-
dullaris magna, also known as the artery of
Adamkiewicz (AKA) in the lower thoracic and
lumbar region (Fig. 2).
The mid-thoracic spinal cord is more vulnera-
ble to ischemic compromise as there is
minimal collateral supply to the spinal cord
superior to the junction of the AKA with the
ASA. The ASA continues caudally past the
conus as an insignificant branch to the filum
terminale; the more important, functional contin-
uations of the ASA are the “rami cruciantes” that
surround the conus and provide robust “basket”
anastomoses with the posterior spinal
arteries [17].
Posterior Spinal Arteries
The two posterior spinal arteries (diameter <0.5
mm) also originate at the level of the foramen
magnum by branches of the ipsilateral
vertebral or posterior inferior cerebellar arteries
and travel along the posterolateral surface of the
spinal cord, receiving supply at various levels
from the posterior radiculomedullary arteries
(Fig. 3). 11–16 feeders contribute to the posterior
spinal artery (PSA) at various levels through the
spine. The largest posterior radiculomedullary
artery often enters below the level of the
AKA [15].
Throughout its course, each PSA gives off
branches that supply the posterior third of the
spinal cord, including the posterior columns, dor-
sal gray matter, and superficial dorsal aspect of the
lateral columns of the spinal cord.
1130
Fig. 2 Anterior spinal artery: DSA demonstrating the
midline anterior spinal artery (ASA) with contribution
from the dominant radiculomedullary branch in the
Radiculomedullary Arteries
At several levels, the radicular artery gives branches
that follow the anterior and/or posterior nerve roots
to supply the spinal cord, called the radiculo-
medullary arteries. The average number of anterior
radiculomedullary arteries is 6 (range 2–14) [18,
19], whereas the number of posterior radiculo-
medullary arteries varies from 11 to 16 [15].
Radiculomedullary Artery of Cervical
Enlargement
There are 2–3 anterior radiculomedullary arteries
in the cervical region. The most important supply
to the cervical ASA is usually located between the
C4–C8 levels of the spinal cord and is termed the
artery of cervical enlargement [20], often a branch
of the deep cervical artery, and usually accom-
panies the C6 nerve root.
Artery of Adamkiewicz (AKA)
The most important anterior radiculomedullary
artery and the one most easily recognized in
S.R. Boddu et al.
thoracolumbar segment, artery of Adamkiewicz with typ-
ical hairpin appearance (a, b). Note the vertebral artery
contribution to ASA in the cervical cord segment (c, d)
angiography is the artery radiculomedullaris
magna, also known as the artery of Adamkiewicz
(AKA). It has a diameter of 0.5–1.0 mm [15] and
almost always arises in the thoracolumbar region,
between T8 and L2 in 75 % of cases [16,
21–23]. It has, however, been identified as supe-
riorly as T5 and as inferiorly as L4 [19]. In 80 % of
cases, it is found in the left side. The AKA forms
the classic “hairpin” loop when it reaches the ASA
and gives off a thin ascending branch and a larger
descending branch.
Pial Plexus
Besides the direct connections between the ASA
and PSAs at the conus medullaris, there is an
extensive arterial network on the entire surface
of the spinal cord formed by effective anastomo-
ses between the transverse and oblique branches
from the ASAs and PSAs. This superficial rich
anastomotic network is termed the pial plexus
(vasocorona) and has been visualized on
microangiograms of the spinal cord [15]. This
47 Spinal Cord Infarction and Differential Diagnosis
Fig. 3 Posterior spinal artery: DSA demonstrating the
posterior spinal artery (PSA) in the lower thoracic cord.
Note the paramedian location of the PSA compared to ASA
and the sharper angle of the hairpin curve
network is responsible for vascularizing the
periphery of the spinal cord.
Intrinsic Arterial System
The spinal cord parenchyma is supplied by the
intrinsic arterial system, which is subdivided into
a central (centrifugal) system and a peripheral
(centripetal or vasocorona) system [13]. The cen-
tral system is comprised of the central arteries
(sulcal or sulcocommisural arteries; diameter
0.06–0.40 mm), which originate from the ASAs
and travel into the anterior median fissure, pene-
trate into the cord either on the left or right, and
branch centrifugally, mainly within the gray mat-
ter [19]. The peripheral system (vasocorona;
diameter 0.1–0.2 mm) [19] consists of small per-
forators (rami perforantes) that originate from the
1131
pial plexus at the periphery of the spinal cord and
course into the white matter centripetally.
Watershed Territories
of the Spinal Cord
The concept of watershed areas has been applied
to the areas of relative poor vascularity in the rich
spinal cord vascular network. As detailed in the
anatomy section, the spinal cord has rich vascu-
larity contributed from the longitudinal and cir-
cumferential networks. The blood supply of the
cord thus depends on the integrity of half a dozen
radicular arteries so far as the anterior three-
quarters of the cord is concerned and on a slightly
greater number of posterior radicular arteries for
the nourishment of the posterior third.
Given the relative paucity of these networks in
certain regions, there exists “watershed terri-
tories” in both the networks.
Longitudinal Anastomosis
Although radicular arteries accompany the spinal
nerves at each level, the radiculomedullary
branches supplying the spinal cord are present
only at few levels. The average number of anterior
radiculomedullary arteries is 6 (range 2–14) [18,
19], whereas the number of posterior radiculo-
medullary arteries varies from 11 to 16 [15]. The
superior cord segment involving cervical and
upper thoracic cord (up to T2/T3) is usually sup-
plied by vertebral arteries and 2–3 anterior
radiculomedullary arteries, including the artery
of cervical enlargement [20], often a branch of
the deep cervical artery, and usually accompanies
the C6 nerve root. Similarly 2–3 anterior radiculo-
medullary arteries, including the artery of
Adamkiewicz (AKA), usually supply the inferior
cord segment involving the lower thoracic and
lumbar cord. The intermediate cord segment
involving the mid-thoracic cord (T4–T8) is a
poorly vascularized segment with paucity of the
radicular arteries, although occasional radicular
artery is present at T7 level. In addition to the
1132
overlapping areas between the upper, mid, and
lower cord segments, the entire mid-thoracic
cord due to its paucity of the radiculomedullary
arteries is considered as an anatomical “watershed
territory” of the spinal cord.
Circumferential Anastomosis
Within the intrinsic arterial system, there is anas-
tomosis between the central (centrifugal) system
and the peripheral (centripetal or vasocorona) sys-
tem [13]. The sulcal artery is a branch of the
anterior spinal artery, which, centrally oriented,
runs through the anterior median sulcus, enters
the cord, and supplies most of the gray matter
and the adjacent white matter. The peripheral sys-
tem (vasocorona) [19] consists of small perfora-
tors (rami perforantes) that originate from the pial
plexus at the periphery of the spinal cord and
course into the white matter centripetally. Located
in between, there exists a border zone that is
supplied by both the central and peripheral
arteries.
Ischemic Vulnerability of the Spinal Cord
Historically, the literature has supported the
notion of a spinal cord “watershed zone” of ische-
mic vulnerability centered at the mid-thoracic
level (T4–T8 level) [24, 25]. However, Duggal
et al. showed that >95 % of cardiac arrest and
hypotensive patients with associated ischemic
myelopathy had predominant involvement of the
lumbosacral level followed by cervical level with
relative sparing of thoracic levels. None of the
examined patients in their study had neuronal
necrosis limited to the thoracic level only. Authors
concluded that, although the thoracic spinal cord
may be the anatomic watershed zone with respect
to regional blood supply, the lumbosacral cord
neurons appear to be more susceptible to
ischemia.
Neuronal necrosis restricted to the lumbosacral
level of the spinal cord was first reported by Gilles
and Nag [26] in 6 neonates suffering transient
cardiopulmonary arrest. These findings were
later confirmed in a larger series by Azzarelli
and Roessmann [27] in their material from
16 patients, including 11 adults with ischemic
S.R. Boddu et al.
myelopathy, which demonstrated lesions located
throughout the spinal cord in nine cases and
restricted only to the lumbosacral cord in four
patients.
High metabolic demands and the large number
of neurons in the gray matter make it more vul-
nerable to ischemia than the white matter
[24]. The spinal cord blood flow analysis in pri-
mates showed a higher blood supply to the gray
matter (57.6 +/ 2.3 ml/100 g/min), due to the
much higher metabolic rate compared to the white
matter (10.3 +/ 0.2 ml/100 g/min) [24]. Similarly
primate studies’ total flows in cervical, upper tho-
racic, and lumbar areas were 14.9 + 1.4, 10.4 
0.8, and 19.7  1.2 ml/min/100 g, respectively
[24]. The lowest total flow is observed in the
thoracic cord; in part, this difference has been
attributed to the relative decrease in the gray mat-
ter at this level.
Etiology
Spinal cord infarction (SCI) can be categorized
into three groups (Table 2):
1. Ischemic infarction
2. Hemorrhagic infarction
3. Venous infarction
Ischemic Spinal Cord Infarction
This is the most common type of spinal cord
infarction. The common etiological factors in
this group are included in two distinct groups:
hypoperfusion and embolism. The individual
risk factors are discussed in detail below:
Hypoperfusion Related
Aortic Surgery
Paraplegia as a complication of aortic surgery in
humans was first reported in 1956 by Adams
et al. [36]. Surgical repair of thoracic and
thoracoabdominal aortic aneurysms and aortic
dissections (Fig. 4) are the most common etiology
47 Spinal Cord Infarction and Differential Diagnosis 1133

Table 2 Etiology of spinal cord infarction for SCI [7, 9, 10]. Although this complication is
Ischemic infarction reported both after open and endovascular repair
Hypoperfusion of ruptured aneurysm, its incidence is lower by
Aortic surgery using endovascular techniques, partly related to
Aortic dissection the patient selection bias [37].
Systemic hypotension Variable incidence of the SCI after
Thrombosis of abdominal aortic aneurysm endovascular stent graft has been reported up to
Atherosclerosis of segmental arteries 10 % [38], but it could be as low as 0.6 % [39] and
Vertebral artery dissection as high as 28 % in high-risk cases [40]. In a
Mechanical injury (surfer’s myelopathy) multicenter comparison trial, the reported inci-
Abdominal/pelvic surgeries dence of SCI after endovascular repair of
Spinal surgery or intervention
descending aortic aneurysms has a risk of 3 %
Embolism
versus 14 % for open surgery [41]. SCI incidence
Thromboembolism
is much lower for endovascular repair of the
Cardiac arrhythmias
abdominal, rather than thoracic aorta [42]. In the
Valvular heart disease
Therapeutic embolization (spinal, renal, and
setting of trauma and open repair, the rate of spinal
bronchial) cord ischemia is about 7 % when cross-clamping
Aortic catheterization is used alone. This can be decreased to 3 % by
Hypercoagulable conditions [28, 68] using shunts for distal perfusion and cardiopul-
Fibrocartilaginous embolism monary bypass [37].
Septic (infection) embolus In a large case series, the incidence of paraple-
Hemorrhagic infarction gia as a result of thoracoabdominal aneurysm
Subarachnoid hemorrhage repair was reported as high as 21 %, whereas the
Arterial dissection aneurysmal repair confined to only the abdominal
Ruptured spinal aneurysm aorta was 1 % [43]. Spinal cord ischemia has been
Dural arteriovenous fistula (DAVF) reported with iliac aneurysm repair as well
Perimedullary arteriovenous malformation (AVM)
[44]. Hypoperfusion in the abdominal aorta with
Anticoagulation
prolonged clamping [45], declamping syndrome
Bleeding diathesis
after declamping of the infrarenal aorta and spinal
Epidural/subdural hemorrhage
cord tamponade [46], and increased pressure of
Spontaneous rupture of spinal DAVF, perimedullary
AVM [29] the cerebrospinal fluid (CSF) during clamping
Intramedullary hemorrhage – hematomyelia [47] are the main risk factors in the pathogenesis
Spinal angiomas [29] of the spinal cord ischemia following the aortic
Ruptured spinal AVM surgery.
Cavernoma Spinal cord ischemia may also happen in pro-
Spinal tumors cedures involving the ligation of radicular arter-
Venous infarction ies, such as spinal tumor resection, scoliosis
Spinal DAVF, perimedullary AVM correction, or pneumonectomy [48].
Cord compression from extramedullary lesions
Tumor (e.g., pancreatic) embolus [30] Atherosclerosis
Decompression sickness (air embolus) The link between atherosclerosis and spinal cord
Miscellaneous
ischemia has been considered since the late nine-
Radiation injury [31–33]
teenth century [49]. Atherosclerosis is minimal in
Vasculitis [31–33]
the spinal arterial network when compared to the
Epidural anesthesia [7, 9, 25]
rest of the body. The anterior spinal artery is the
Substance abuse (e.g., cocaine) [34]
Drug induced (e.g., sildenafil citrate) [35]
most affected, while other arteries in this system
may undergo fibrosis and thickening instead.
1134
Fig. 4 Spinal cord infarction following aortic surgery: a
69-year-old male with Marfan’s syndrome, status post-
graft repair of the ascending aorta (a) following type A
dissection, presented with paraplegia following surgery.
Note the bilateral iliac artery dissections. (b) MRI –
There is both ascending and descending flows
within both the anterior and posterior spinal arte-
rial systems. This allows for collateral circulation
in the event of occlusion [50]. While moderate to
severe aortic atherosclerosis was observed in
86 % in an autopsy series of SCIs, no meaningful
correlation has been observed between degree of
aortic atherosclerosis, hypertension, or coronary
artery disease and occurrence of lumbosacral
ischemia [51]. In a case series of 1,000 autopsies,
28 patients who had atherosclerotic emboli from
the abdominal aorta to visceral organs were iden-
tified, and the spinal cord was dissected. In
12 cases atheromatous emboli were evident in
the spinal arteries, most commonly in the anterior
spinal artery. However, only one had a spinal cord
infarction, showing that apparently good collat-
eral circulation and lack of general atherosclerosis
prevents infarction, unless hypoperfusion leads to
decompensation [52].
S.R. Boddu et al.
T2-WI hyperintensity in the central gray matter T7/T8
level (c, d) and extends distally to the conus (inferiorly
involves both the central and peripheral aspects of the
entire distal cord). SCI in the anterior and posterior spinal
arteries with cord swelling noted inferiorly (c, e)
Aortic Dissection
Spinal cord ischemia as a possible cause of weak-
ness in the limbs as a result of impaired perfusion
of the spinal cord aortic dissection had been
known since at least the late nineteenth century
[53]. Dissecting thoracoabdominal aortic aneu-
rysms, especially those with a left-sided false
lumen, are associated with a higher incidence of
spinal cord ischemia, because in 85 % of the
population, the anterior spinal artery arises from
the left lower intercostal vessels [46]. Based on
the Stanford classification of the aortic dissection:
Stanford type A aortic dissection – dissection
involving the ascending aorta, regardless of
the site of primary intimal tear [54] (Fig. 5).
Neurological symptoms can be present in 29 %
upon presentation. Spinal cord ischemia can be
present preoperatively in about 1–9 %. In a
study showing a preoperative incidence rate
47 Spinal Cord Infarction and Differential Diagnosis
Fig. 5 SCI following aortic dissection: a 71-year-old
male: status post-dissection of the ascending aorta (a, b).
Subacute spinal cord infarction involving the central gray
matter extends from T11 to the conus medullaris. In
of 1 %, this rate increased to 4 %
postoperatively [55].
Stanford type B aortic dissection – consists of all
aortic dissections except type A. Spinal cord
ischemia is present in about 3 % of this
type [56].
Systemic Hypotension
Based on the autopsy series of 145 fatal cases of
cardiac arrest or severe systemic hypotension, SCI
was documented in 46 % of cases [51]. Although
widely understood that the mid-thoracic is the
watershed area, and presumably most susceptible
to ischemic injury, the thoracic area was spared,
and the lumbar area was the most commonly
1135
addition to the T2-WI hyperintensity (c, d), notice the
post-contrast enhancement in the area of infarction (e, f)
suggestive of subacute nature (>5 days) of this infarction
involved area in 95 % of cases (Fig. 6). Ischemic
damage limited to only the thoracic region was
seen in only 7.6 % of cases and never in isolation,
whereas lumbosacral involvement was the most
common affected area, with nearly 70 % of cases
of spinal cord ischemia showing only involve-
ment of this region. This is potentially due to
higher metabolic demand and a large number of
neurons in the lumbosacral region [51]. Similar
predisposition of lumbar spinal cord for SCI is
also documented in the neonates [57].
Surfer’s Myelopathy
This is a rare acute nontraumatic myelopathy,
affecting first time/inexperienced surfers.
1136
Fig. 6 Anterior spinal artery territory infarct: a 6-year-old
girl with gastroenteritis presented with sudden onset of
lower extremity weakness, thoracic pain, loss of rectal
tone, and mid-thoracic sensory loss. MRI shows diffuse
Thompson et al. first described this entity in 2004
[58]. The exact etiology is unknown. Prolonged
spinal hyperextension while in prone posture and
repetitive flexion and extension are the main pre-
cipitating factors resulting in spinal cord arterial
insufficiency and watershed ischemia [58]. Pro-
posed vascular mechanisms of injury include
avulsion of perforating vessels, vasospasm of the
artery of Adamkiewicz, and transient ischemia in
areas of borderline perfusion as a result of tension
on the spinal cord with hyperextension [59].
Other possible mechanisms include inferior
vena cava obstruction and fibrocartilaginous
embolism. The former may occur secondary to
compression by the liver while lying prone over
the surfboard aggravated by concomitant
prolonged Valsalva maneuver during paddling,
resulting in increased retrograde venous pressure
in the epidural venous plexuses leading to infarc-
tion [59]. Potential risk factors that may predis-
pose surfers to ischemic cord injury include body
habitus (thin with underdeveloped musculature),
dehydration (related to travel and prolonged hours
at the beach), and long-distance travel (associated
S.R. Boddu et al.
T2 hyperintense signal in the anterior spinal artery territory
from T6 to the conus (a–c) in anterior spinal artery syn-
drome. Subsequent DSA shows patency of AKA (d) and
PSA (e)
with hypercoagulable state and deep venous
thrombosis) [58].
The thoracic spinal cord is most commonly
affected in the surfer’s myelopathy [60]. The vari-
able levels of thoracic spinal cord involvement may
be explained by the variable origins of the artery of
Adamkiewicz, arising from T5–T8 in 15 %,
T9–T12 in 75 %, and L1–L2 in 10 % of the popu-
lation [60]. The outcome is variable and does not
correlate with the extent or intensity of the signal
abnormality on the MRI [61]. Most patients had a
complete or near-complete recovery; residual com-
plete paraplegia has been reported [58, 59, 61].
Embolism Related
Thromboembolic Infarction
The main embolic source for the SCI is cardiac,
well described in the literature with diverse
pathology such as atrial fibrillation, atrial myx-
oma [62], mitral valve disease [31], and patent
foramen ovale [63]. Spinal angiogram may dem-
onstrate the abrupt termination of the anterior or
posterior spinal arteries (Fig. 7).
47 Spinal Cord Infarction and Differential Diagnosis
Fig. 7 Embolic spinal cord infarction: a 66-year-old
female presented to ED with acute onset bilateral lower
extremity motor and sensory deficits. MRI: T2-WI
hyperintensity extends from T11 to the conus medullaris
involving the central gray matter with cord swelling (a–c).
Fibrocartilaginous Embolism
Fibrocartilaginous embolism (FCE) was first
described as a cause of spinal cord infarction in
1961 [64]. A history of heavy lifting or trauma to
the spine may be a clue to this etiology, but this
may not always be the case. Nearly 70 % of cases
are young women, in contrast to thrombotic or
other embolic etiology, predominantly affecting
men. Three characteristics distinguish FCE from
other thrombotic or embolic causes of SCI
[65]. First, their presenting symptom is that of
the sudden severe pain in the region of the spine
that is affected, whereas in thrombotic and other
embolic causes, this is hardly ever the case. Sec-
ond, after the onset of pain, there is a period with
no neurological symptoms. Over the course of
15 min to 48 h, symptoms of neurological deficit
progress, indicating the evolution of a spinal
stroke, whereas in thrombotic and other embolic
etiologies, numbness and weakness of the legs
begin at onset and at the same time as pain, if
present. Finally, these cases have a very poor
1137
Note the abrupt termination of the AKA (arrow) on subse-
quent DSA suggesting an embolic occlusion (d) (Courtesy
Dr. Kim Nelson and Tibor Becske, NYU Medical Center,
New York, NY)
outcome, while those of thrombotic or other
embolic causes often improve over time.
It is thought to occur when, due to increased
intra-disk pressure, fragments of the intervertebral
disk are pushed into the disk vasculature or the
bone marrow and venous sinuses of vertebral
bodies. Supporting this is the finding of fibrocar-
tilaginous emboli in the bone marrow and vascu-
lar supply of the affected regions in autopsies [65,
66]. Seventy percent involve the upper cervical
cord, with some extending to the lower medulla
oblongata or the lower thoracic, which explains
the high fatality of these cases. The presence of
cord swelling and collapsed intervertebral disk at
the corresponding level indicates this
etiology [65].
Infection
Syphilis was diagnosed in a majority of cases of
spinal cord ischemia in a report from 1902
[5]. This was prior to the introduction of penicillin
as a cure. Later case series show the most common
1138
identified etiology to be of aortic origin, with
many as a complication of surgery on the aorta
[9]. Bacterial meningitis [67] and mucormycosis
[68] are reported in the literature as an etiology
for SCI.
Iatrogenic Embolization
The spinal cord ischemia has been reported as a
complication related to therapeutic embolization.
This is particularly reported with bronchia artery
embolization for hemoptysis between 1.4 % and
6.5 % [69]. Many other iatrogenic procedures are
reported to cause spinal cord infarction including
but not limited to renal artery embolization [70],
intercostal embolization in hemoptysis [71], inter-
costal artery embolization in traumatic hemor-
rhage [72], endoscopic variceal sclerotherapy
[73], ligation [74], regional anesthesia, and pain
injections [75].
Spinal Angiography and Embolization
Angiography of the thoracic or abdominal
branches of the aorta rarely results in spinal cord
ischemia. Spinal angiography is more challenging
in patients with atherosclerotic aorta and
thoracoabdominal aneurysm. Spinal cord infarc-
tion as a complication of angiography was
documented as early as 1949 and reported on in
the decade thereafter [76]. However, with
increased experience and better techniques and
equipment in this field, complications have been
lowered. In a series of 134 consecutive spine
angiograms in 63 patients published in 1988,
three neurological complications were reported
(incidence 2.2 %), all of which resolved in a
matter of 1 day to 1 week [76]. In more recent
series on spinal angiography and embolization
from high-volume centers [77] including authors’
experience with 220 spinal angiograms and
180 embolizations [78], no neurological compli-
cations were reported.
Hemorrhagic Spinal Cord Infarction
Spinal subarachnoid hemorrhage (SAH) accounts
for 1 % of all SAH. However, this is the most
common form of hemorrhagic spinal cord
S.R. Boddu et al.
infarction (Fig. 8). Cervical spinal cord ischemia
has been described in association with spinal
subarachnoid hemorrhage in patients
following dissection of the vertebrobasilar system
or ruptured vascular malformations of the
spinal cord [77]. Disruption of the blood supply
due to the involvement of radiculomedullary
arteries is thought to be the underlying
mechanism in the acute setting [78]. Micro- and
macrocirculatory failure secondary to subarach-
noid hemorrhage may result in delayed ischemic
deficits. Vasospasm has shown to be the most
likely mechanism causing secondary injury in
experimental animal models of spinal cord
injury [79].
Although different lesions predispose to hem-
orrhage in different compartments (Table 2), their
strict localization is not possible. The hemorrhage
tends to extend across multiple compartments
depending on the amount and site of hemorrhage
and may result in subdural and epidural hematoma
in addition to the subarachnoid hemorrhage.
Prolonged mass effect from the extra-axial hema-
toma in addition to vascular disruption related to
involvement of the radiculomedullary feeders, cir-
culatory failure, and vasospasm might result in
spinal cord ischemia in these scenarios [78]. Con-
ditions such as anticoagulation and bleeding
diathesis may cause hemorrhage in either
compartments.
Spinal Venous Infarction
This less common etiology is usually associated
with spinal dural arteriovenous malformations
(Fig. 9) secondary to venous congestion [80] and
is discussed in multiple subsequent sections of
this chapter. It may also be a complication of
deep sea diving, in which gaseous occlusion of
the spinal venous plexus has been postulated as a
mechanism of injury in spinal decompression
myelopathy [9]. Early recompression reverses
symptoms in about three-fourths of patients [81,
82]. Kim et al. [83] stated that the hemorrhagic,
nonhemorrhagic, and embolic venous infarction
could not be differentiated with certainty on clin-
ical grounds. Hemorrhagic venous infarctions
47 Spinal Cord Infarction and Differential Diagnosis
Fig. 8 Spinal cord infarction from SAH: a 54-year-old
female presented with acute subarachnoid hemorrhage
(SAH) and quadriplegia. CT – posterior fossa predominant
SAH with intraventricular extension and hydrocephalus (a,
b). MRI – posterior decompression of the cervical spine
with residual diffuse T2-WI hyperintense cord signal
involving the central gray matter with edema and cord
were associated with marked pain at the time of
onset and progressed rapidly; survival time was
relatively short. Nonhemorrhagic infarctions were
less often painful and evolved slowly. Survival
time was longer than in hemorrhagic infarction.
Embolic venous infarction was associated with
pain of abrupt onset. As the initial presentation
does not correlate with severity of the disease,
treatment with hyperbaric chamber should be ini-
tiated promptly in decompression sickness, even
in those with mild symptoms [82]. Also, the
potential association of malignancy and venous
spinal cord infarction is well recognized, particu-
larly in the case of pancreatic neoplasm [30].
1139
swelling (c, d). Also notice the tracking of the extra-axial
hemorrhage from posterior fossa along the anterior border
of the spinal cord with dorsal displacement (c). DSA dem-
onstrates a small anterior spinal artery aneurysm (arrow),
as a culprit of SAH, pre- (e, f) and post- (g) successful coil
embolization projections
Pathophysiology
Understanding of the biomolecular changes that
occur during and after spinal cord infarction (SCI)
is essential for all those involved in the care of
these patients. Extensive research over the past
30–40 years has focused on elucidating acute
spinal cord injury including both primary and
secondary mechanisms, slowly unraveling the
complex pathophysiologic processes such as
roles of calcium, free radicals, sodium, excitatory
amino acids, vascular mediators, and apoptosis.
These concepts of primary and secondary
1140
Fig. 9 Spinal dural arteriovenous fistula (DAVF): a
42-year-old male with progressive paraplegia. MRI – dif-
fuse serpiginous flow voids over the cord surface with
associated T2-WI hyperintense cord edema and swelling
(a). Contrast administration reinforces the visualization of
mechanisms and their duality in acute spinal cord
injury have since also been embraced in the under-
standing of the pathophysiology of spinal and
cerebral ischemia, subarachnoid hemorrhage,
and head trauma [84].
Concepts of Primary and Secondary
injury
Fehlings proposed a two-step process involving
primary and secondary mechanisms in acute spi-
nal cord injury [85]. Allen first postulated the
concept of a secondary mechanism in 1911
based on canine studies [86, 87].
Primary Mechanism
The primary mechanism involves the initial pre-
cipitating event such as atherosclerosis, cardiac
S.R. Boddu et al.
blood vessels over cord surface (b). Reconstructions from
the contrast-enhanced time-resolved MRA demonstrate the
spinal DAVF (c). This is confirmed on the subsequent
spinal DSA (d), supplied by the right T10 radiculo-
medullary artery
arrest, thromboembolism, hypovolemic shock,
and hemorrhage with associated compression.
Similar to the brain, spinal cord vessels
autoregulate according to SBP and PaCO2, but
the spinal cord perfusion is more directly affected
by changes in systemic blood pressure. The
“watershed region” of the spinal cord in the
mid-thoracic region (T4–T8) is anatomically
more susceptible for the ischemic injury. In case
of spinal arteriovenous malformations, the
shunting of the arterial blood away from the spinal
cord (arterial steal phenomenon) or venous hyper-
tension may result in hypoperfusion. The hypoxia
and hypoperfusion resulting from the primary
mechanisms trigger the secondary mechanism.
Secondary Mechanisms
Over the past 40 years, various theories of sec-
ondary mechanisms of SCI have undergone a
47 Spinal Cord Infarction and Differential Diagnosis
process of maturation. In the 1970s, Demopoulos
et al. [88] proposed the free radical hypothesis, as
crucial to the injury process. Ten years later the
focus shifted onto the role of calcium, opiate
receptors, and lipid peroxidation. Research in the
twenty-first century has suggested apoptosis,
intracellular protein synthesis inhibition, and
glutaminergic mechanisms, among a myriad of
pathophysiologic pathways that mediate second-
ary injury mechanisms. There is considerable evi-
dence that the primary mechanical injury initiates
a cascade of secondary injury mechanisms [85,
89–93].
Although there are many mediators of second-
ary injury, vascular mechanisms, free radicals,
and more recently apoptosis have received much
attention.
Free Radicals
Free radicals are highly reactive molecules that
possess an extra electron in the outer orbit. Free
radicals most commonly form from molecular
oxygen. There is good evidence for the early
occurrence and pathophysiologic importance of
oxygen free radical formation with cell membrane
lipid peroxidation in central nervous system
injury [88]. Severe uncontrolled production of
the hydroxyl radicals (HO) from dissociation of
hydrogen peroxide (H2O2) can cause lipid perox-
idation, impairment of phospholipid-dependent
enzymes, disruption of ionic gradients, and ulti-
mately the lysis of cell membrane [94, 95]. Lipid
peroxidation may also play a role in postictal
hypoperfusion after initial precipitating event
and may result in SCI [94].
Vascular Mechanisms
The changes that occur in spinal cord blood flow
after an acute spinal cord injury can be divided
into systemic and local.
Local effects: Following acute cord injury, a
major reduction in blood flow occurs at the site of
ictus, i.e., embolus or hemorrhagic compression,
possibly precipitated by vasospasm or vasoactive
amines [96, 97]. Based on animal studies,
autoregulation is intact during the initial
60–90 min after cord injury but is then lost coin-
cident with the onset of ischemia. It has been
1141
suggested that the ischemic response to cord
injury is mediated both by the loss of
autoregulation and by relative constriction of the
resistance vessels [98]. Ischemia may play a role
in the formation of local cord edema, although
whether edema formation is injurious in itself or
an epiphenomenon is still unclear [84].
Because of differences in relative vascularity,
the central gray matter, along with adjacent white
matter, is more severely affected by acute cord
injury than peripheral white matter [99]. Primate
studies have shown that the gray matter has higher
blood supply than the white matter due to the
much higher metabolic rate. Also, the flow was
more variable with lower values in the dorsal
horns and higher values in the central gray and
anterior horns, compared to the more uniform
flow in the white matter [100]. The blood flow to
the corticospinal tracts was considerably higher
than the rest of the white matter, suggestive of
higher metabolic activity in these tracts and
increased predisposition to the ischemia [101].
The intrinsic arterial blood supply to the spinal
cord is directly proportional to the cross-sectional
area of the gray matter, which is most abundant in
the thoracolumbar segment [102]. Therefore, in
normal conditions, the blood flow to the spinal
cord is highest in the thoracolumbar segment,
which has the highest concentration of central
arteries owing to the relative abundance of gray
matter. Consequently, the same segment is the
most vulnerable to hypoperfusion, particularly
the gray matter that corresponds to the lumbar
enlargement of the spinal cord. Arterial infarcts
of the posterior spinal artery territory are rare
compared with the anterior spinal artery territory.
The scarcity of this condition is most likely due to
extensive anastomosis of the posterior spinal
artery [103] and relative resistance of the posterior
column to the ischemia due to its inherent lowest
blood flow of all the tracts in the spinal cord [101].
White matter perfusion typically decreases
within 5 min of an acute cord injury and begins
to return to normal within 15 min. It remains
thereafter near-normal pre-injury values during
the first 24 h. In contradistinction, in the central
gray matter, numerous microhemorrhages typi-
cally occur, as early as 5 min after an acute cord
1142
injury. Perfusion is relatively absent 1 h after an
injury, and this state remains so for at least the first
24 h. This vascular standstill has been confirmed
using microangiography [97, 104], fluorescent
tracer studies [105], and the operating microscope
[106] and plays a major role in the devastating
outcomes from acute cord injury. Alterations in
endothelial cell function causing an increase in
vascular permeability and edema formation have
been well documented [107–109].
Systemic effects: After an acute SCI, an imme-
diate but transient increase in mean arterial blood
pressure may occur followed by profound hypo-
tension and bradycardia [110]. The reason for this
transient hypertension is unknown but may be
mediated by both the thoracic sympathetic gangli-
ons and the adrenal glands [111]. Acute cord
injury is one of the causes of neurogenic shock
[112], typically being related to the magnitude and
severity of the cord injury.
Of all purported mechanisms of secondary
injury, the vascular hypothesis has considerable
weight, with biochemical, angiographic, histo-
pathologic, and clinical support for its key role
in damage after acute SCI.
Apoptosis
Apoptosis also called programmed cell death
occurs in a wide variety of disease states in
eukaryotic cells. Apoptosis is recognized as
occurring in utero as a form of neuronal cell
death during embryonic development [113] and
is also now thought to play a role in many post-
developmental disorders of the central nervous
system, including ischemia [114]. Apoptosis is
an active process that is characterized by cell
shrinkage, chromatin aggregation, and nuclear
pyknosis [115]. In the spinal cord, apoptosis was
first identified in 1995 as occurring in rats [116]
and, more recently, in the human spinal cord
[117]. This is a tightly regulated process with a
sequence of activation steps that require energy
and specific macromolecular synthesis as de novo
gene transcription [115]. A family of cysteine
proteins, the caspases, is thought to play an impor-
tant role in apoptosis, especially caspase-3
[84]. Oligodendrocytes are the major cell type in
compressive cord injury that undergo apoptosis
S.R. Boddu et al.
[118], seen in areas of wallerian degeneration and
detectable between 24 h and 3 weeks post
injury [116].
Spinal Venous Infarction
Spinal cord veins are valveless. However, an anti-
reflux mechanism has been described at the dural
exit of radicular veins. It consists of an increased
smooth muscle fibers and a Z-shaped course of
the radicular vein when it crosses the dura to
drain spinal cord venous blood in the epidural
veins [14]. In case of spinal dural arteriovenous
fistulas, the arterialized vein drains the
fistula blood flow toward the spinal cord, and the
increased pressure is transmitted to the
intrinsic cord veins in retrograde fashion. This
results in decreased intramedullary arteriovenous
gradient and hypoxia, loss of autoregulation, and
disruption of blood-cord barrier and, if left
untreated, results in subacute necrotic
myelopathy.
Imaging Correlation
Ischemia in the distribution of the anterior spinal
artery rapidly initiates vasogenic and cytotoxic
edema in the gray matter with subsequent increase
in mobile water. This is manifested by a focal
increase in proton density and a prolongation of
the T2 relaxation time, which appears as distinct
increase of signal intensity on the long TR pulse
sequences.
The ischemia/signal abnormalities start in the
anterior horns of the gray matter and, with increas-
ing severity, spread posteriorly to involve the pos-
terior horns. Ultimately, the ischemic changes and
corresponding MR signal abnormalities extend
laterally to the posterolateral funiculi of the spinal
cord, including the crossed corticospinal tracts. In
severe cases, the whole cross section of the spinal
cord is infarcted. There seems to be a correlation
between the distribution of the MR signal abnor-
malities and the severity of the clinical symptom-
atology seen in spinal cord ischemia [46]
(Table 3).
47 Spinal Cord Infarction and Differential Diagnosis
Table 3 Summary of secondary injury in acute spinal
cord infarction
Vascular changes [92, 97, 119]
Ischemia
Impaired autoregulation
Neurogenic shock, hemorrhage
Microcirculatory derangements
Vasospasm
Thrombosis
Ionic derangements [93, 120]
Increased intracellular calcium
Increased extracellular potassium
Increased sodium permeability
Neurotransmitter accumulation
Catecholamines [121] (e.g., norepinephrine,
dopamine)
Excitotoxic amino acids [122, 123] (e.g., glutamate)
Arachidonic acid release [124]
Free radical [124] and eicosanoid production [125,
126]
Lipid peroxidation [125, 126]
Endogenous opioids and cytokines [127, 128]
Edema [129]
Decreased adenosine triphosphate production [130]
Apoptosis [131–133]
Clinical Presentation
Acute spinal cord infarction syndrome (ASCIS)
can result in various clinical manifestations
based on the vasculature affected and etiology,
but has common symptoms. Bilateral
weakness, sensory loss, back pain at onset, and
urinary retention were the most common
symptoms at the time of presentation in a
prospectively collected series of 28 patients
[134]. The most commonly affected location is
the thoracolumbar level, and paraparesis com-
prises of 20–25 % of all spinal cord infarction
cases [135]. Pain and sensory changes occur
first, followed by weakness within minutes to
hours, which peaks in 12 h [135]. Almost
invariably spinal cord infarction presents with
pain, at the level of the lesion, one retrospective
study noting pain in 72.7 % of 23 patients
[136]. Transient ischemic attacks (TIA) rarely
precede spinal cord infarct, comprising of 10 %
1143
of spinal cord infarcts [135]. TIA manifest as
weakness of the lower limbs precipitated by phys-
ical activity or as “drop attacks” for cervical cord
ischemia.
Ischemic Spinal Cord Infarction
Based on international standards for neurological
and functional classification of spinal cord
injury proposed by the American Spinal Injury
Association (ASIA) and International Medical
Society of Paraplegia (IMSOP), ASCIS can be
categorized into four groups [137]. The clinical
syndromes are:
1. Anterior spinal artery syndrome (ASAS)
2. Posterior spinal artery syndrome (PSAS)
3. Brown-Sequard syndrome (BSS)
4. Complete spinal cord transaction (CSCT)
Anterior Spinal Artery Syndrome (ASAS)
This is the most common presentation of ASCIS,
affecting the anterior two-thirds of the spinal
cord that is supplied by one anterior spinal
artery (Fig. 10). Cervical anterior cord infarction
manifests with diffuse or radicular neck pain
followed by quadriparesis, associated with anal
or bladder sphincter dysfunction and sensory
loss of pain, pinprick, and temperature modalities.
Proprioception and vibration sensations, which
are usually spared in anterior cord syndrome,
can be affected in anterior cervical lesion due to
involvement of medial lemniscus. Later on, flac-
cidity and areflexia give way to spasticity and
hyperreflexia. Patients present with acute
interscapular pain, followed by paraparesis or
paraplegia, urinary and rectal incontinence or
retention, and a sensory loss to pain and temper-
ature most common at the T4 level. Lumbosacral
region is affected by lesion in the radicular artery
of Adamkiewicz. Extensive lesions can be seen
with flaccid paraplegia and autonomic dysfunc-
tion, including respiratory compromise with cer-
vical lesions. Rarely, one segmental branch of
anterior spinal cord artery can result in
monoparesis or hemiparesis, instead of
paraparesis [138].
1144
Fig. 10 Anterior spinal artery (ASA) territory infarction.
T2-WI hyperintense signal abnormality involving the ante-
rior two-thirds of the spinal cord including both gray matter
Fig. 11 Posterior spinal artery (PSA) territory infarction.
Schematic representation of the posterior cerebral artery
(PCA) territory infarction on the right. Note the
Posterior Spinal Artery Syndrome (PSAS)
Posterior cord infarction is rare due to the
rich anastomotic network (Fig. 11). Posterior
column and dorsal horns are affected,
manifested as total anesthesia and areflexia at the
level of lesion. Affected patients have deficit in
vibratory and proprioception deficits below the
lesion.
Brown-Sequard Syndrome (BSS)
Central sulcal artery occlusion can cause Brown-
Sequard syndrome, which typically presents with
ipsilateral weakness and posterior column sensory
S.R. Boddu et al.
and white matter secondary to occlusion of the ASA (a).
Pictorial representation of the ASA territory infarction (b)
corresponding MRI on the left with subtle focal T2-WI
hyperintensity in the right PSA territory of the spinal cord
deficit (vibration and proprioception) below the
lesion and contralateral decrease in pain and tem-
perature sensation a few segments below the
lesion.
Central Spinal Cord Syndrome (CSCS)
Central cord syndrome may theoretically be a man-
ifestation of spinal cord ischemia, but commonly
result from cervical spondylosis, trauma, syringo-
myelia, or other tumors. Location is the cervical or
upper thoracic regions, involving the central por-
tions of the spinal cord. The hallmark sign is char-
acterized by weakness in the arms, with relative
47 Spinal Cord Infarction and Differential Diagnosis
sparing of the legs, due to the latter’s corresponding
corticospinal tract in lateral part of spinal cord.
Affected patients also have varying degree of sen-
sory loss below the lesion and urinary
retention [139].
Hemorrhagic Spinal Cord Infarction
Hemorrhagic infarction presents with sudden
onset of pain, usually at the level of hemorrhage.
Meningeal irritation is a typical finding. Patients
may have associated headache, cranial neuropa-
thies, and decreased level of consciousness
depending on extent and distribution of the hem-
orrhage. The evaluation for spinal SAH typically
follows negative intracranial workup.
Venous Hypertension
This is the most common presentation in SDAVF.
Patients present with progressive lower extremity
weakness (>50 %), sensory deficits (30 %)
including paresthesia, sensory loss and hyperes-
thesia, gait disturbance, back pain (24 %), bowel
and bladder disturbance, and impotence. Once
motor or sensory deficits appear, severe disability
ensues over 6 months to 2 years.
Imaging Findings
Magnetic resonance imaging (MRI) is the primary
imaging modality of choice for the diagnosis of
suspected spinal cord infarction due to its superior
spatial and contrast resolution for differentiation of
the intraspinal soft tissues and detection of intrinsic
cord pathology [140] (Fig. 12). Also MRI is most
sensitive in the detection of the marrow changes of
the adjacent vertebral bodies [140]. Perhaps the
most important role of MRI in both ischemic and
hemorrhagic SCIs is to rule out alternative diagno-
sis. MRI not only confirms the evidence of spinal
cord infarction but also may provide information as
to the underlying etiology. The extent of signal
abnormalities correlates well with the severity of
the clinical findings and the potential for recovery
1145
[46]. Following acute presentation, MRI should be
performed on emergency basis to confirm the diag-
nosis, delineate pattern of SCI, and rule out differ-
ential diagnosis such as cord compression from
traumatic, degenerative, or neoplastic processes
and intramedullary lesions such as acute inflamma-
tion, hemorrhage, and congestive ischemia due to
vascular malformations.
Evolution of Spine Imaging
Low signal-to-noise ratio (SNR) and prolonged
imaging times while scanning the entire spine are
the main challenges of the spine imaging. Introduc-
tion of the phased-array coil technology [141] and
increased field strength has increased the baseline
SNR, a major improvement in spine imaging,
which has inherently limited SNR. Similarly intro-
duction of the parallel imaging [142, 143] and
improvement of multielement radiofrequency
(RF) coils have not only enhanced the SNR of
high-resolution MR imaging over a large
craniocaudal extent but also dramatically shortened
the imaging times.
Standard Spine Protocol
Regular structural MR imaging sequences for
diagnosis of the cord infarction include
multiplanar T1- and T2-WI fast spin-echo (FSE),
gradient-echo (GRE), and contrast-enhanced
(CE) acquisitions. Diffusion-weighted imaging
(DWI) has increasingly shown as a promising
technique in the evaluation of the acute spinal
cord syndrome [144, 145] and used as a part of
the standard protocol in the evaluation of acute
non-compressive myelopathy.
Conventional MRI
The sensitivity of conventional MRI is limited,
particularly in the first several hours. The reported
sensitivity of T2-WI cord signal abnormality in
SCI ranges between 45 % and 73 % depending in
part on timing of the scan [1, 9, 10,
1146
Fig. 12 Normal MR imaging of the spinal cord. Homo-
geneous cord signal, no intrinsic T2 hyperintensity of the
cord or cord swelling or serpiginous flow voids on the cord
surface (a, c, d). No vascular enhancement either intrinsi-
cally within the cord or vascular enhancement over the
cord following contrast administration (b). The focal
146, 147]. Hence the initial scan may either con-
firm the diagnosis or be a normal study and serve as
a baseline for future follow-up imaging. Isolated
finding of T2 signal abnormality, even with
restricted diffusion, is not specific for ischemia
and can be seen in transverse myelitis and other
intrinsic cord pathologies. These findings, confined
to a vascular territory, are more specific for SCI.
Diffusion-Weighted Imaging (DWI)
of the Spinal Cord
Historical Aspects
DWI of the spinal cord is technically challenging,
but is currently feasible. Hajnal et al. are the first
to publish the DW images of the human cervical
S.R. Boddu et al.
T2-WI hyperintensity and post-contrast enhancement
(arrows) in the middle of the posterior vertebral border
represent a vertebral vasculature, a normal finding. Note
the exquisite spatial and soft tissue contrast resolution
offered by the MRI in spine imaging
spinal cord in 1991 [148]. By using a standard EPI
sequence and a spinal coil, Stepper and Lovblad
[149] demonstrated areas of hyperintensity on
DWI (ischemic region) corresponding to a
decrease of the local ADC (75 % of normal
values) when compared with non-affected spinal
cord. The time course of the changes in the ADC
values corresponds to what has been reported in
the brain [150].
Technical Challenges
Several technical difficulties for the application of
the most frequently available DWI technique, i.e.,
single-shot echo-planar imaging (EPI), to spinal
cord imaging include motion artifacts caused by
physiological movement of the spinal cord and
47 Spinal Cord Infarction and Differential Diagnosis
surrounding structures (e.g., cerebrospinal fluid
pulsatile flow, swallowing, respiration, and car-
diac motion), susceptibility artifacts caused by
the presence of bone and cerebrospinal fluid inter-
faces, and the low signal-to-noise ratio caused by
the small pixel dimensions required for appropri-
ate visualization of the spinal cord
[151]. Improved imaging techniques such as nav-
igated interleaved EPI or parallel imaging-
enhanced EPI provide superior resolution and
adequate resistance to motion artifacts compared
to the conventional EPI [140]. More recently, line
scan diffusion imaging (LSDI) has been reported
as new advancement in the DWI of the spinal
cord [151].
Clinical Role
Increasingly, DWI has shown to be highly sensi-
tive to the early detection of ischemic lesions in
spinal cord and thus can improve the diagnostic
capabilities especially in the acute stage
[152–154]. DWI is also a very useful
technique in differentiating diagnosis of subacute
infarction. After all, hyperintensities on T2-WI
images are not specific for ischemia and can also
be seen in inflammatory diseases and tumors
[155]. DWI will have increasingly important role
for ischemic diseases of the spinal cord not only in
early diagnosis but also in differentiating
diagnosis [155].
Typical MRI Findings
The typical MRI findings in spinal cord infarction
include isolated pencillike area of T2
hyperintensity involving the centromedullary
region with sparing of the anterior rim, often
encompassing more than two vertebral segments
on sagittal images [2, 156], and bilateral
hyperintensities that are mostly confined to the
anterior horn area, leading to the typical “snake
eyes or owl’s eyes” configuration on the axial
acquisitions [156, 157]. Although the T2
hyperintensities are seen in 90 % of patients with
SCI, these are nonspecific [156]. Further in the
acute phase, only 50 % of the patients show a
demarcation of T2 hyperintensity of the spine
1147
during the first 24 h which limits the sensitivity
of this finding [80].
A finding of restricted diffusion on diffusion-
weighted images (DWI-MRI) reported to be sig-
nificantly more sensitive than standard T2
images, appearing as early as 3–8 h after the
ictus, but the limited evidence precludes a
numerical estimate of its sensitivity [144, 151,
152, 158]. Studies have shown that restricted
diffusion can be seen in SCI as early as 3–8 h
following an ictus [152–154]. Moderate
swelling of the cord is noted in the acute phase
with onset of T1-WI hypointensity in the cord
from days 5–7 [80]. Contrast enhancement is
noted in the subacute stage, typically after
5 days, and persists for up to 3 weeks after onset
[159]. Similar to the cerebral edema, venous
hypertension from spinal DAVF without cord
infarction gives pure extracellular edema, thus
increasing T2-WI signal intensity but with normal
findings on DWI.
Stages of Spinal Cord Infarction
The abnormal signal intensities in SCI typically
start in the territory of the anterior spinal artery
and spread to the adjacent central white matter.
The patterns of MR signal abnormality with
increasing signal changes reflect a temporal
sequence of ischemic changes that can often be
recognized as distinct, separate stages. However,
the MR appearance is frequently a mixture of
patterns, particularly in the patients with com-
pleted spinal cord stroke.
Depending on the extent of involvement on
MRI, spinal cord ischemia can be categorized
into four patterns [46]:
Anterior Horn Ischemia
Results in abnormal T2-WI hyperintense signal
involving the anterior horns of the gray matter
giving an “owl’s eye” appearance (Fig. 13a).
Central Gray Matter Ischemia
T2-WI hyperintense signal abnormality involving
both anterior and posterior horns of the gray mat-
ter (Fig. 13b).
1148
Fig. 13 (a) Anterior horn infarction. Schematic represen-
tation of the anterior horn infarction on the right. Note
the corresponding MRI on the left with subtle focal
T2-WI hyperintensity in the anterior horns of the spinal
cord. (b) Central gray matter infarction. Schematic repre-
sentation of the central gray matter infarction on the right.
Note the corresponding MRI on the left demonstrating
focal T2-WI hyperintensity in the central gray matter of
the spinal cord. (c) Central gray and white matter infarc-
tion. Schematic representation of the central gray and white
Gray and White Matter Ischemia
T2-WI hyperintense signal abnormality involving
the entire gray matter and the adjacent white mat-
ter (Fig. 13c).
S.R. Boddu et al.
matter infarction on the right. Note the corresponding MRI
on the left depicting the diffuse T2-WI hyperintensity
involving the central gray and white matter of the spinal
cord. (d) Cross-sectional infarction. Schematic representa-
tion of the cross-sectional infarction on the right. Note the
corresponding MRI on the left with diffuse focal T2-WI
hyperintensity involving the cross-sectional area of the
spinal cord. Note the thin rim of preserved normal cord
parenchyma secondary to the perfusion from peripheral
pial plexus
Cross Section Ischemia
T2-WI hyperintense signal abnormality involving
the entire cross section of the spinal cord involv-
ing both gray matter and white matter (Fig. 13d).
47 Spinal Cord Infarction and Differential Diagnosis
Associated Vertebral Body Infarction
A finding of vertebral body infarction on MRI is a
useful indicator of vascular nature of the
acute spinal cord syndrome [160–163]. However,
this finding is not specific, seen in only 4–35 % of
patients with spinal cord infarction [1, 146,
164]. SCI may not always be accompanied by
vertebral body infarction, because arterial
occlusion may be located distal to those
vessels supplying the vertebral body or
because of the good collateral blood supply of
the vertebra [160]. Cheng et al. [164]
showed that concomitant spinal cord and
vertebral body infarction is highly associated
with aortic pathology, which highlights the
common blood supply segmental arteries
supplying the metamere and radiculomedullary
branches of the radicular artery (branch of spinal
trunk of the segmental artery) feeding the spinal
cord [13].
Useful Clues on MRI to the Underlying
Etiology
1. Intervertebral disk disease at the level of
infarction: possible fibrocartilaginous
embolism
2. Spondylotic disease at the level of
infarction: possible compressive myelopathy
[164]
3. Multiple serpiginous flow voids on the surface
of the spinal cord: spinal cord vascular
malformations [165]
Holistic Approach
In a patient with suspected SCI, the MRI signal
abnormalities alone are nonspecific. Care should
be taken to interpret these findings in conjunction
with history, clinical findings, and CSF analysis.
Although CSF analysis in spinal cord infarction
is typically normal, but there can be pleocytosis
(rarely more than 100 WBC) and an
elevated protein (usually less than 119 mg/dl)
[7, 9, 10, 166].
1149
Advanced Techniques
Advanced techniques such as diffusion tensor
imaging (DTI) and fiber tractography (FT) are
increasingly investigated in the cord infarction.
These techniques have technical challenges limit-
ing the image quality, related to the highly mag-
netically inhomogeneous material surrounding
the spinal canal, small size of the spinal structures,
long craniocaudal extent of the spinal canal, pul-
sation effects of cerebrospinal fluid (CSF) and
blood, and motion degradation from respiration,
swallowing, and patient movement. In patients
with cervical cord injury, fractional anisotropy
(FA) of DTI and FA showed better correlation
with motor function and clinical findings such as
neurological impairments compared to the con-
ventional MRI [167]. DTI imaging of the spinal
cord is currently only a research tool, but prelim-
inary studies have shown that it holds consider-
able promise in predicting the severity of spinal
cord injury [168].
Follow-Up Imaging
Unlike cerebral infarct, onset of T2 hyperintense
signal abnormality in the spinal cord infarction is
variable from 8 to 72 h, and studies reported
normal MRI findings at 48 h following ictus
[157, 169]. High index of clinical suspicion war-
rants follow-up imaging. However, up to 14 % of
patients with suspected SCI have normal follow-
up MRI scans [10, 166].
Repeated or follow-up DWI may also help in
differentiating diagnosis. There seems to be more
sensibility on DW images than on T2-WI images
during signal changes within the course of dis-
ease, and DWI may be useful in monitoring the
progress of the disease [155]. Serial MR studies
with contrast showed sequential changes of
enhancement similar to those seen in cerebral
infarcts [169]. Understanding of characteristic
MR findings and evolutional changes at different
stages of SCIs and follow-up using serial MRI
1150
with contrast are vital in the early diagnosis
and more effective in the management of the
patients.
Differential Diagnosis
The differential diagnosis of spinal cord infarction
is broad and holds multiple entities under the roof
of “acute nontraumatic myelopathy.” These can
be categorized into (a) inflammation,
(b) infection, (c) compressive myelopathy,
(d) venous congestion related to vascular
malformations such as spinal dural arteriovenous
fistula (SDAVF), and (e) tumor. Distinction of SCI
from “SCI mimics” is challenging especially in
subacute stage. After all, hyperintensities on
T2-WI images are not specific for ischemia and
can also be seen in inflammatory diseases and
tumors [155].
Sudden onset of symptoms may support the
diagnosis of medullary ischemia. In addition,
due to the characteristics of blood supply in
the spinal cord, infarction lesions are confined to
arterial territories and usually lie in the anterior
or posterior part of the spinal cord, whereas
inflammatory or neoplastic medullary lesions
often occupy or affect the whole medulla.
Detection of synchronous lesions in the brain
may isolate some of the differential diagnosis
(Table 1).
DW MRI will have increasingly important role
for ischemic diseases of the spinal cord not only in
early diagnosis but also in differentiating diagno-
sis [155]. Marked hyperintensity on DWI may
provide another indicator leading to the specific
diagnosis of spinal medullary ischemia. Mild
hyperintensity was likely to be seen on DW
images of myelitis or tumors, suggesting the pos-
sibility of separating ischemic from inflammatory
cord lesions and intramedullary tumors by the
different range of ADC values [153,
155]. Repeated or follow-up DWI may also help
in differentiating diagnosis. There seems to be
more sensibility on DW images than on T2-WI
images during signal changes within the course of
disease, and DWI may be useful in monitoring the
progress of the disease [155].
S.R. Boddu et al.
The list of differential diagnosis is summarized
in the table. Main differential diagnoses are
discussed below.
Multiple Sclerosis (MS)
Multiple sclerosis is a chronic inflammatory
demyelinating disease of the central nervous sys-
tem (CNS). The frequent involvement of spinal
cord in MS is shown up to 99 % in autopsy series
[170]. 70–80 % of the patients with MS have
T2-WI cord signal abnormalities [171]. MS
plaques are best demonstrated on the T2-WI
acquisitions, with signal characteristics of T2-WI
hyperintensity and T1-WI iso-hypointensity
[172], similar to cord infarction. Cord swelling is
a usual feature of the relapsing-remitting form of
MS [173, 174], and similarly cord atrophy with
axonal loss is a recognized form of MS
[175]. Post-contrast enhancement of the demye-
lination lesions is described due to the associated
breakdown of the blood-brain barrier, although
significantly lower than their brain
counterpart [176].
However, the MS lesions differ from SCI in
pattern of cord involvement and extent of CNS
distribution. MS plaques present as well-
circumscribed foci of T2-WI hyperintensity with
asymmetric involvement of the cord. They are
characteristically peripherally located, are less
than two vertebral segments in length, and occupy
less than half the cross-sectional area of the cord.
On sagittal acquisitions, they are located centrally,
anteriorly, and dorsally. On axial images, plaques
are located in the lateral segments and have wedge
shape with the basis at the cord surface or round
shape if there is no contact with the cord
surface [172].
Unlike SCI, MS predominantly involves cer-
vical cord (62 %) and usually demonstrates
relapsing-remitting disease course [172]. Simulta-
neous imaging of the brain to detect synchronous
lesions helps to differentiate MS from cord infarc-
tion. Up to 92 % of MS patients with cord lesions
have associated brain lesions [177] (Fig. 14).
Detection of synchronous lesions in brain and
spinal cord is useful in differentiating the MS
47 Spinal Cord Infarction and Differential Diagnosis
Fig. 14 Multiple sclerosis: a 41-year-old female
presenting with relapsing-remitting course of neurological
deficits. MRI – multiple T2-WI hyperintense demyelin-
ation plaques involving the cervical (a, b) and thoracic
(c, d) spinal cord. MS plaques are peripherally located
and occupy less than half the cross-sectional area of the
from other inflammatory and cerebrovascular dis-
eases with an accuracy of 93–95 % [178–180].
Idiopathic Transverse Myelitis (ITM)
Transverse myelitis (TM) is predominantly
immune system mediated [181]. This may exist
as an isolated idiopathic entity or as a part of the
systemic process with reported incidence in
patients with vasculitis, neurosarcoidosis, multi-
ple sclerosis, and systemic lupus erythematosus
[182] and following vaccination [183, 184] and
acute gastrointestinal and respiratory
illness [181].
1151
cord (b, d, f). The thoracic cord lesion shows post-contrast
enhancement (e, f) suggestive of active demyelination
plaque. Simultaneous brain MRI shows multiple intracra-
nial white matter lesions with distribution of multiple scle-
rosis involving the corpus callosum and peritrigonal
distribution (g–i)
ITM has reported a prevalence of 15.6 % in a
retrospective study involving 288 patients with
transverse myelitis [185]. ITM predominantly
affects the middle-aged adults. The term “trans-
verse” is self-explanatory describing the position
of the inflammation, that is, across the width of the
spinal cord. Thoracic cord is most commonly
involved [182]. The criteria proposed by the
Transverse Myelitis Consortium Working Group
for idiopathic transverse myelitis [186] include:
1. Bilateral sensory, motor, or autonomic spinal
cord dysfunction
2. Defined sensory level with bilateral signs and
symptoms
1152 S.R. Boddu et al.

Fig. 15 Idiopathic
transverse myelitis: a
23-year-old female with
abrupt onset paraplegia.
Diffuse T2-WI
hyperintense cord signal
abnormality with edema
and cord swelling (a–c).
Note the signal abnormality
does not follow a specific
arterial territory and
involves more than
two-thirds of the cross-
sectional area of the spinal
cord (b, c). No associated
intracranial lesions are
demonstrated on the MRI
brain performed
simultaneously

3. Proof of spinal cord inflammation on MRI or
CSF
4. Duration between symptoms’ onset to maxi-
mum intensity: few hours to 21 days
5. Exclusion of extra-axial compressive etiology
Typical MR findings of ITM [182] include
focal centrally located T2-WI hyperintensity, usu-
ally occupying more than two-thirds of the spinal
cord cross-sectional area (88 %), signal abnormal-
ity that extends for more than three to four verte-
bral segments in length (53 %), variable presence
of cord swelling or expansion (47 %), and patchy
or diffuse post-contrast enhancement (47–53 %)
[182, 185]. The variability in the duration of clin-
ical onset, extent, and pattern of cord signal abnor-
mality and given up to 40 % of ITM cases can have
normal MR imaging [187] make it quite challeng-
ing to differentiate ITM from cord infarction
(Fig. 15). A detailed history of associated systemic
conditions, through clinical examination and
close monitoring of the clinical condition in com-
bination with CSF analysis, may be helpful in
differentiating these two entities. A history of
recent vaccination or viral illness may suggest
ITM, but these are often absent. Symptoms isolated
to the anterior spinal artery territory suggest infarc-
tion over inflammation, but this is not definitive.
Progression of symptoms over at least a few hours,
symptoms and MRI lesions that extend beyond a
vascular territory, gadolinium enhancement on the
initial presentation, and cerebrospinal fluid find-
ings of pleocytosis or elevated IgG levels favor
transverse myelitis over infarction [188].
Acute Disseminated Encephalomyelitis
(ADEM)
Acute disseminated encephalomyelitis (ADEM)
is a non-vasculitic inflammatory demyelinating
disorder of the central nervous system, most likely
triggered by the autoimmune response to the mye-
lin basic protein [172]. This usually affects the
prepubertal children, in contrary to the SCI. The
thoracic cord is most commonly affected. The
47 Spinal Cord Infarction and Differential Diagnosis
Fig. 16 Acute disseminated encephalomyelitis (ADEM):
an 8-year-old boy with history of recent viral illness
presented with weakness in all four limbs and flaccid
reflexes. MRI: T2-WI hyperintensity involving the cervical
and thoracic cord with no specific arterial territory
onset of ADEM usually occurs in the wake of a
clearly identifiable febrile prodromal illness or
immunization and in association with prominent
constitutional signs and encephalopathy of vari-
able degree [189], feature that are uncommon in
cord infarction. MRI findings of the spinal cord
are nonspecific with large T2-WI hyperintense
and T1-WI hypointense lesions extended over
long segment of the spinal cord with associated
cord expansion [190] (Fig. 16). Spinal cord
involvement is noted in 71 % of patients, and all
patients with spinal involvement have cerebral
involvement and features of myelopathy
[191]. Thus, simultaneous cerebral imaging to
detect brain lesions helps to differentiate this
entity from cord ischemia. The typical signs of
ADEM described in literature are involvement of
the basal ganglia, thalamus, brain stem, and
1153
distribution (a–d). Presence of intracranial lesions in the
brachium pontis (e), thalamus (f), deep white matter (g),
and centrum semiovale (h) on brain MRI performed simul-
taneously. Child was diagnosed with ADEM
cortical lesions [192, 193]. The associated brain
lesions explain the frequent clinical manifesta-
tions of brain stem signs and ataxia in ADEM
[194] along with routine motor and sensory defi-
cits seen in the context of acute myelopathy.
Rarely, ADEM lesions may contain areas of
hemorrhage, suggestive of herpes simplex virus
2 (HSV-2) encephalomyelitis [193, 195], which
can mimic the hemorrhagic cord infarction.
Spinal Cord Arteriovenous Shunts
Spinal cord arteriovenous shunts (SCAVS) are a
rare heterogeneous group of vascular
malformations presenting with similar MR imag-
ing findings mimicking the ischemic and/or hem-
orrhagic spinal cord infarction.
1154
Fig. 17 Spinal epidural arteriovenous fistula (DAVF): a
77-year-old male with chronic back problems and prior
spinal surgery presented with new onset right lower
extremity weakness following L4/L5 injection. MRI –
multiple serpiginous flow voids over the cord surface
with associated T2-WI hyperintense cord edema and
Spinal dural arteriovenous fistula (SDAVF) is a
close differential diagnosis for the spinal cord
infarction (Fig. 17). SDAVF constitutes 70 % of
all spinal vascular malformations (SVMs), com-
monly affecting middle-aged (40–60 years)
males. It usually presents as slowly progressive
neurological deficits; however, acute onset of the
disease and progressive deterioration interrupted
by remissions have been described [196, 197]. If
left untreated, it results in irreversible paraplegia
or tetraplegia. The increased spinal venous hyper-
tension decreases the arteriovenous pressure gra-
dient resulting in impairment of the normal spinal
venous drainage and resulting in venous conges-
tion with intramedullary edema [198, 199] and
end stage with spinal cord infarction.
On MRI diffuse edema from congestion results
in centromedullary T2-WI hyperintensity of the
cord involving multiple segments and mimics the
cord infarction [197]. Both entities have similar
findings of cord swelling and contrast
S.R. Boddu et al.
swelling (a–c) suggestive of spinal vascular malformation.
Also note the extensive degenerative changes in the lumbar
spine and prior surgery (a). This patient had extensive
search for SVM on DSA that was localized to the left
internal iliac artery (d) on the second spinal DSA
enhancement in subacute stage and cord atrophy
in chronic stage [200]. However, additional
T2-WI findings of dilated perimedullary vessels
manifested as tortuous flow voids [197], and a
hypointense rim related to deoxygenated blood
within the dilated capillary vessels surrounding
the congestive cord edema [201] is characteristic
of SDAVF and helps to differentiate the two enti-
ties. If the shunt volume is small, the
perimedullary vessels might only be seen after
contrast enhancement. The dilated perimedullary
vessels are seen as filling defects in the continuous
contrast column in myelography (Fig. 18). It is
important to remember that neither the location of
the pathologic flow voids nor the intramedullary
imaging findings correspond to the localization of
the fistula [80].
The arteriovenous malformations (AVMs)
carry increased risk of intraparenchymal and
subarachnoid hemorrhage (Fig. 19) in addition
to venous congestion, compared to
47 Spinal Cord Infarction and Differential Diagnosis 1155

Fig. 18 Myelographic
demonstration of spinal
DAVF: a 48-year-old male
with known
spondyloarthropathy
evaluated with
myelography for lower limb
symptoms. Sagittal (a) and
coronal (b) myelography
reformats demonstrating
multiple serpiginous filling
defects over the cord
surface, suggestive of
underlying spinal vascular
malformation. Subsequent
DSA (c) confirmed
spinal DAVF

Fig. 19 Spinal intramedullary arteriovenous malforma-
tion (AVM): a 31-year-old male presented with sudden
onset quadriparesis. MRI: heterogeneous, predominantly
T2-WI hyperintense signal (a, b) with patchy post-contrast
enhancement (c, d), which can be seen in subacute spinal
cord infarction. Spinal DSA: intramedullary AVM at C5,
feeders from the ASA, right vertebral artery (e, f), left
ascending cervical artery from thyrocervical trunk (g).
Note the right cervical vertebral artery dissection with
small pseudoaneurysm (e, f)
1156
Fig. 20 Ependymoma: a 64-year-old male with progres-
sive weakness of the bilateral lower limbs. MRI – hetero-
geneous predominantly T2 hyperintense intramedullary
SDAVF [80]. The AVMs along with cavernous
malformation can mimic hemorrhagic cord infarc-
tion. The presence of intramedullary hemorrhage
makes MRI findings more complex with heteroge-
neous T1-WI and T2-WI signal abnormalities and
associated T2-WI peripheral hypointense rim from
the blood products [202, 203]. However, the pres-
ence of intra- and perimedullary flow voids in
AVMs [80] and typical intramedullary mulberry-
like appearance, representing hemorrhage in differ-
ent stages [204], is a helpful clue to differentiate
these diagnoses from isolated spinal cord infarction.
Spinal Cord Neoplasms
Primary and secondary tumors of the spinal cord
result in extensive intramedullary T2-WI
hyperintensity with associated cord expansion,
S.R. Boddu et al.
mass lesion of the conus medullaris with cord edema and
swelling (a, c). Note the patchy post-contrast enhancement
(b, d)
mass effect, and discrete borders mimicking the
cord infarction. Ependymomas, astrocytomas,
and hemangioblastomas are the three most fre-
quently seen primary intramedullary tumors in
the adult population [205]. Majority of these
tumors are usually seen in the cervicothoracic
region with some variability in the histologic
type [206, 207]. The myxopapillary subtype,
which makes up nearly 30 % of all spinal
ependymomas, is found almost exclusively in
the cauda equina and filum terminale [208]
(Fig. 20). The mean size of ependymomas
corresponds to height of 3–4 vertebral bodies
and 5–6 vertebral bodies for astrocytomas
[205]. Ependymomas are more centrally located
compared to the eccentric location of the astrocy-
tomas and hemangioblastomas. Diffuse avid post-
contrast enhancement with smooth margins is
noted in ependymomas and lesser extent of
47 Spinal Cord Infarction and Differential Diagnosis
Fig. 21 Hemangioblastoma: a 46-year-old male with progressive worsening of lower limb weakness. MRI – large
intramedullary mass lesion, predominantly cystic (a–c) with enhancing mural nodule (arrow) on post-contrast images (d)
enhancement and nodular tumor margins in astro-
cytomas [206, 209]. Hemangioblastomas are
either solitary or multiple (von Hippel-Lindau
disease) [210] with an intensely enhancing mural
nodule and a large cystic component [206]
(Fig. 21). Hemangioblastoma larger than 24 mm
in size is frequently associated with flow voids on
MRI [207]. These tumors are slow growing and
often present with features of osseous remodeling,
such as canal expansion and scalloping of the
posterior vertebral bodies [211], which are not
seen in acute scenario of cord infarction. In addi-
tion, slower clinical onset, diffuse or nodular post-
contrast enhancement, and presence of
intramedullary cystic components favor neoplasm
over infarction.
Intramedullary spinal cord metastases (ISCM)
are rare with reported frequency of 0.9–1.2 %
based on the autopsy series from cancer patients
[212, 213]. Lung cancer is the most common
source followed by metastasis from the breast,
kidney, melanoma, and lymphoma [214]. The
1157
cervical cord is the most frequent site for ISCM,
although they have reported at all levels of the
spinal cord [214]. ISCM are associated with dis-
proportionate vasogenic edema involving the long
segments of the spinal cord resulting in T2-WI
hyperintensity and cord expansion, which may
mimic cord infarction [215]. However, the metas-
tasis usually shows ringlike or homogeneous
intense enhancement following contrast adminis-
tration [205]. Moreover, most patients have
known malignancy and systemic metastasis with
frequent concomitant brain metastasis or
leptomeningeal carcinomatosis at the time of
ISCM diagnosis [216].
Rarely, an acute ASA infarction of the bilateral
anterior horns may be mimicked by a polio viral
infection. Such appearance is related to the
peculiar viral tropism underlying this
condition [217]. Vaccine-associated paralytic
poliomyelitis cases have been also reported occa-
sionally in the pediatric population [218] (Tables 4
and 5).
1158
Table 4 List of differential diagnoses for spinal cord
infarction
A. Inflammation/demyelination
1. Multiple sclerosisa
2. Devic’s neuromyelitis opticaa
3. Idiopathic transverse myelitis
4. Acute disseminated encephalomyelitisa
5. Acute polyneuropathy (Guillain-Barré syndrome)
6. Subacute combined degeneration
7. Neurosarcoidosisa
8. Neuro-Behcet’s diseasea (Fig. 22)
9. AIDS-associated myelopathya
10. Radiation myelitis
B. Infection
1. Bacterial myelitis
2. Viral myelitis
3. Fungal myelitis
4. Tuberculous myelitis
5. Toxoplasmosis
C. Compressive myelopathy
1. Epidural/subdural hematoma (Fig. 23)
2. Epidural/subdural abscess (Fig. 24)
3. Acute disk herniation with cord compression
4. Epidural/dural based mass lesions and metastasis
(Figs. 25, 26, and 27)
D. Spinal cord arteriovenous shunts (SCAVS)
(Figs. 28, 29, and 30)
E. Neoplastic
1. Primary intramedullary tumors
2. Intramedullary spinal cord metastasis
3. Leptomeningeal carcinomatosisa
a
Simultaneous brain imaging may provide useful diagnos-
tic information in differentiating these entities from spinal
cord infarction
Role of Angiography
The principal role of catheter angiography in
suspected spinal cord infarction (SCI) is to con-
firm the differential diagnosis of spinal vascular
malformations (SVMs) and evaluate the patency
of the anterior and posterior spinal arteries and
localization of the dominant radiculomedullary
artery (artery of Adamkiewicz). Accurate locali-
zation of the artery of Adamkiewicz is crucial in
the diagnosis and treatment planning of spinal
vascular lesions [225] and thoracoabdominal aor-
tic aneurysm repair [226] and for planning the
preoperative embolization of hypervascular spinal
S.R. Boddu et al.
tumors [227] (Figs. 31 and 32) to minimize the
risk of SCI. Certainly spinal digital subtraction
angiography (SpDSA) is the ultimate imaging
technique for diagnosing, localizing, and classify-
ing spinal vascular lesions [80].
The congestive myelopathy from spinal dural
arteriovenous fistulas or intracranial dural arterio-
venous fistulas with venous reflux into the brain
stem and spinal cord can mimic spinal cord infarc-
tion (Fig. 33). It is proven beyond doubt that the
extent and site of the cord signal abnormality or
surface flow void do not correlate with the loca-
tion of the SDAVF [80]. The fact that the SDAVF
constitutes 70 % of all SVMs [80] highlights the
limitation of the conventional MRI in accurate
localization of the spinal vascular malformations.
On the other hand, with intrinsic cord pathology
such as intramedullary vascular malformation,
cavernoma, or intrinsic cord tumor, conventional
MRI can accurately localize the site of pathology;
however, it fails to offer additional crucial infor-
mation regarding the vascular supply of these
lesions that is vital in their treatment
planning [172].
Challenges in Spinal Angiography
The main challenges in the spinal angiography in
general irrespective of the imaging modality are
the following:
1. Differentiation of submillimeter to millimeter
normal spinal cord arteries from veins (ASA,
0.2–0.8 mm; AKA, 0.5–1.0 mm; PSA, <0.5
mm) [15] and their relation to the adjacent
structures such as the spinal cord and verte-
brae, requiring high spatial resolution.
2. Characterization of SVMs, identification of
multiple submillimeter feeding arteries, and
separation of the inflowing arterial feeders
from the outflowing draining veins that
demand high temporal resolution. Although
veins have relatively large caliber (0.4–1.5
mm) [15], the great anterior radiculomedullary
vein (GARV), which is the largest vein
draining the anterior thoracolumbar spinal
cord, is easily mistaken for the AKA due to
47 Spinal Cord Infarction and Differential Diagnosis
Table 5 Summary of main differential diagnosis for SCI and useful clues
Differential diagnosis
1. Multiple sclerosis (MS)
2. Spinal cord neoplasm
3. Idiopathic transverse myelitis
(ITS)
4. Spinal cord arteriovenous
shunts (SCAVS)
5. Guillain-Barré syndrome [219]
(GBS)
6. Acute disseminated
encephalomyelitis (ADEM)
7. Subacute combined
degeneration (SACD) [220]
8. AIDS-associated myelopathy
[221, 222]
9. Radiation myelitis [211, 223]
1159
Highlights
Peripheral in location
Less than two vertebral segments in length
Less than half cross-sectional area of the cord
>90 % incidence of associated intracranial lesions
Relapsing and remitting clinical course
Invariable cord expansion present
Diffuse or nodular contrast enhancement
Extensive peritumoral edema
Associated cystic changes
Slower clinical onset
Remodeling of adjacent osseous structures
Lesion centrally located
3–4 segments in length
Occupying >2/3 of cord cross-sectional area
No associated intracranial lesions
Onset not quite as sudden
Prominent enlarged serpentine pial veins on cord surface with cord swelling and
edema
May rarely show only cord expansion and edema on T2-WI, without prominent
vessels
Hypointense rim related to deoxygenated blood within the dilated capillary
vessels surrounding the congestive cord edema
Neither the location of the pathologic flow voids nor the intramedullary imaging
findings correspond to the localization of the fistula
Gradual progression
Back pain, progressive, fairly symmetric muscle weakness accompanied by
absent or depressed deep tendon reflexes
Ascending symptoms and development of upper motor neuron signs over time
No definite sensory level
Usually affect prepubertal children
Commonly involves thoracic cord
Preceded by a prodromal febrile illness or immunization
Synchronous brain lesions: basal ganglia, thalamus, brain stem, and cortical
lesions
Vitamin B12 deficiency. Gradual onset
Associated pernicious anemia or neuropsychiatric symptoms
Most commonly affects posterior columns, followed by anterolateral and anterior
tracts
Bilateral hyperintense T2-WI hyperintense areas in posterior columns of the
cervical and thoracic cord
Typically no post-contrast enhancement
Clinical and imaging improvement after B12 supplementation
Vacuolar myelopathy. Gradual onset
Edematous swelling of the myelin in the absence of demyelination or
inflammation
Begins in the mid-low thoracic cord and extends rostral with disease progression
Bilateral symmetrical T2-WI hyperintensity in dorsal and lateral columns
extending over multiple segments
Typically no post-contrast enhancement
Acute: transient myelopathy or Lhermitte’s sign or acute paraplegia
Chronic: lower motor neuron disease of extremities and chronic progressive
myelitis
Prior radiation to the spinal cord with no active compression
(continued)
1160 S.R. Boddu et al.

Table 5 (continued)
Differential diagnosis
10. Compressive myelopathy
[224]
Highlights
3–40 months latent period
Acute: long segment T2-WI hyperintensity, cord swelling, and post-contrast
enhancement
Chronic: cord atrophy and cystic
Postoperative period, known malignancy, or degenerative spine disease
Hematoma or abscess in the epidural or subdural location
Usually has a more subacute presentation with evolution over time; this
presentation may be obscured in the perioperative period
Postoperative setting; the differential diagnosis of paraplegia includes an epidural
hematoma, especially if the patient has had either a lumbar drain or epidural
anesthesia

Fig. 22 Neuro-Behcet’s disease: a 36-year-old male with
history of oral ulcerations presented with progressive
weakness of all limbs, lower limbs more than upper
limbs. MRI: T2-WI hyperintense lesion in the cervical
cord involving two-thirds of the cross-sectional area and
cord swelling (a, b). Simultaneous MRI brain shows
predominantly unilateral hyperintense lesions (c–e) with
increased diffusivity (f) involving the posterior limb of the
internal capsule, thalamus, cerebral peduncle, and pons
along the course of the corticospinal tracts, typical features
seen in CNS involvement of Behcet’s disease
47 Spinal Cord Infarction and Differential Diagnosis
Fig. 23 Epidural hematoma: a 62-year-old female with
history of thrombocytopenia presenting with back pain and
increasing leg weakness. MRI – anterior epidural hema-
toma in the lumbar region (arrows) with T2-WI iso-/
its comparable spatial course and location
[228] (Fig. 34).
3. Coverage of a large area of potential fistulous
communication from the skull base, perhaps
including the posterior fossa and caudally to
the level of internal iliac arteries which requires
a large field of view (FOV), which itself chal-
lenges the high spatial resolution.
Spinal Digital Subtraction Angiography
Spinal digital subtraction angiography (SpDSA)
with its high spatial and temporal resolution is the
gold standard for the study of spinal vascular
anatomy, detection of SVMs, localization of the
AKA, and differentiation of embolic spinal cord
infarction from a non-embolic entity. The sensi-
tivity of spinal angiography in detecting the AKA
is very high and approaches 100 % [23,
229–232]. The three-dimensional rotational spi-
nal angiography (3D RSA) performed as a sup-
plement to DSA was reported to better delineate
the complex, extremely tortuous vascular
1161
hyperintense (a) and T1-WI hyperintense signal (b) com-
pared to adjacent cord. Faint post-contrast enhancement (c)
suggests subacute nature of this hematoma
architecture of the malformation, superior demon-
stration of the nidal and venous aneurysms,
enhance distinction of pathologic vessels in rela-
tion to the spinal cord (perimedullary versus
intramedullary) [233], and offer a 3D relationship
of the vasculature to the bone [234]. However,
DSA is time-consuming which challenges patient
compliance and has small risk of complications
inherent to its invasive nature. Performing spinal
DSA under general anesthesia helps to obtain
better quality images by dedicated timing of
apnea and overcomes the limitations related to
patient compliance. So a noninvasive spinal vas-
cular imaging method such as spinal CTA or
MRA that can substitute for or guide DSA to
shorten acquisition time and reduce the complica-
tions is highly desirable in clinical practice.
Spinal Multidetector CT Angiography
The advent of multidetector spiral CT has enabled
very short acquisition time and increased scan
range and high spatial and temporal resolutions.
1162 S.R. Boddu et al.

Fig. 24 Epidural abscess: a 56-year-old female with pro- enhancement and dural thickening on post-contrast imag-
gressive worsening of back pain and lower limb weakness. ing (b, c). Note the underlying diskitis in the lower cervical
MRI: T2-WI hyperintense anterior epidural collection in region at C6/C7 and C7/T1 with associated large
the upper thoracic cord (a) with peripheral rim prevertebral abscess (d)

Fig. 25 Meningioma: an
84-year-old female with
chronic left lower limb
weakness. MRI – focal
dural-based mass lesion
dorsally over the conus
medullaris. The lesion is
isointense to the cord on
T2-WI sequence with
associated cord
compression and intrinsic
cord signal abnormality (a).
The lesion shows mild post-
contrast enhancement (b)
47 Spinal Cord Infarction and Differential Diagnosis
Fig. 26 Nerve sheath tumor: a 34-year-old male with
worsening left lower limb weakness. MRI – heterogeneous
predominantly cystic extra-axial lesion over the left
anterolateral aspect of the conus medullaris with associated
All these features increased the feasibility of CT
spinal angiography. MDCT angiography takes
only seconds, has submillimeter spatial resolu-
tion, and can even be performed in dynamic fash-
ion [235]. Takase et al. reported that MDCT
(four-detector row) could clearly visualize the
normal medullary arteries and veins [236]. Lai
et al. using a 16-detector-row MDCT reported
good correlation between CT angiography and
DSA for SDAVFs [237, 238]. Si-jia et al. using a
64-detector-row MDCT for CT angiography
published similar results. Authors have admitted
the role of CTA as a quite valuable screening
exam before DSA with guiding effect on DSA,
thus reducing the amount of time, contrast, and
radiation involved in DSA [237–239].
CTA offers superior spatial and temporal reso-
lution over MRA. The wider scan range offered
by CTA would be a lot more time-consuming for
MRA [240].
1163
cord compression and intrinsic T2-WI hyperintense signal
(arrow) abnormality (a, b). Note the post-contrast
enhancement (c, d) in the solid component of the tumor
(arrows)
Spinal Magnetic Resonance Angiography
In recent years, CE-MRA has come up with many
innovative developments, including the 3D MRA
and time-resolved MRA (TR-MRA). Conven-
tional MR angiography techniques, including
contrast-enhanced time-of-flight (TOF) and
phase-contrast angiography (PCA), offered suffi-
cient spatial resolution but were not able to depict
the normal arteries of the spinal cord [241]. With
contrast-enhanced TOF MR angiography, only
normal (large) veins in the spinal cord were
depicted to some extent [242]. For PCA, no
reports on the feasibility of depicting normal ves-
sels were published.
Contrast-enhanced 3D MR angiographic
approaches appear successful for the localization
of the AKA at 1.5T field strength. Current
contrast-enhanced MR angiography techniques
claim success rates for the detection of the AKA
ranging from 69 % to 100 % [22, 226,
1164
Fig. 27 Vertebral metastasis with epidural extension: a
64-year-old female presenting with worsening quadriple-
gia in the lower limbs. MRI – vertebral metastasis involv-
ing C4–C6 and T5 with pathological compression fractures
of C5 and T5, epidural soft tissue, cord compression
Fig. 28 Cavernoma. Intramedullary vascular lesion with
lobulated T2-WI hyperintense signal (popcorn pattern),
rim of hypointensity related to hemosiderin, adjacent
S.R. Boddu et al.
resulting in cord edema, and swelling on sagittal T2-WI
(a) and post-contrast images (b–d). Note the large pleural
effusion in the right hemithorax (*) and underlying primary
lung tumor in the left hemithorax (#)
cord edema, and faint post-contrast enhancement. Cervical
(a–c) and lower thoracic (d–f) cavernomas from two dif-
ferent patients are included
47 Spinal Cord Infarction and Differential Diagnosis
Fig. 29 Posterior spinal artery aneurysm: a 39-year-old
male presented with headache, back pain, and diplopia.
MRI brain showed subarachnoid hemorrhage at the con-
vexity sulci (a) with no intracranial aneurysm. MRI spine
243, 244]. The approach uses a strong bolus, of
which mainly the first passage is exploited for
imaging during a period of 20–40 s [226]. The
technique relies on intravenous contrast adminis-
tration and spoiled gradient-echo pulse sequences
with ultrashort TE and TR for most effective
signal intensity from T1 shortening of the gado-
linium and suppression of the non-enhanced back-
ground tissue signal intensity, respectively. The
main advantage of the strong bolus technique is
that it may provide separation of intradural arter-
ies and veins. However, as the technique uses
1165
showed T1 hyperintense focal hematoma in the distribu-
tion of left PSA (b, c). Further evaluation with DSA
showed a tiny left PSA aneurysm as a culprit for the local
hematoma and diffuse SAH
centric k-sampling technique of a bolus-shaped
signal time course, the contrast between arteries
and veins increases with vessel caliber [244] and
is, therefore, better for larger (typically extradural)
than for smaller (intradural) vessels.
Ali et al. [245] used the newer technology of
dynamic multiphase time-resolved MRA in
11 patients with suspected SVMs. The authors
correctly diagnosed six vascular lesions and
were able to localize within one vertebral level
in 5 of the 6 cases. More recently Saindane
et al. [246] also used CE-TR-MRA for the initial
1166
Fig. 30 Spinal epidural arteriovenous fistula: a 70-year-
old female with progressive paraplegia being evaluated for
spinal canal stenosis demonstrates edema and swelling of
the conus medullaris with multiple flow voids over the
spinal cord surface on T2-WI MRI (a). Further evaluation
with DSA confirms the sacral epidural arteriovenous fistula
at L5/S1 level with dominant arterial feeders from the left
evaluation of suspected SDAVFs, with high con-
cordance with DSA. Authors showed that in cases
where the CE-TR-MRA is able to localize an
arterial feeder, the number of catheterized vessels
and total contrast dose can be decreased during
DSA, with attention to the suspected 1 segmen-
tal levels. However, if the CE-TR-MRA study
does not confidently identify an arterial feeder or
is of suboptimal quality, a full spinal DSA is
recommended.
Spinal DSA Versus CTA and MRA
It is not unreasonable to question, what is the
justification to use DSA over these noninvasive
techniques? With regard to accuracy and breadth
S.R. Boddu et al.
sacral branches of the iliosacral arteries (b) and single
epidural draining vein (d). Patient had successful Onyx
embolization of the bilateral arterial feeders with venous
penetration resulting in complete obliteration of the fistula
(c). Mild improvement of the paraplegia is noted immedi-
ately following the procedure
of diagnostic information offered, conventional
angiography is unmatched. Although CT angiog-
raphy and MR angiography provide excellent ana-
tomic definition of intracranial vascular lesions
comparable to that of DSA, these modalities
have not been well developed for the evaluation
of spinal vascular lesions. Indeed, the size of
feeding vessels to these malformations is usually
beyond the resolution of these modalities. The
spatial resolution (0.2 mm) and temporal resolu-
tion (0.25 s) of catheter cerebral angiography
remain unparalleled. CTA may be approaching
this for some applications, with potentially
0.4-mm spatial resolution and 0.5-s temporal res-
olution, whereas MRA remains a little further
47 Spinal Cord Infarction and Differential Diagnosis
Fig. 31 Filum terminale paraganglioma: a 74-year-old
female evaluated for new onset bilateral lower limb weak-
ness. MRI – focal extradural mass lesion in the filum
terminale with T2-WI hyperintensity (a), T1-WI isointense
(b) signal to the cord and avid post-contrast enhancement
away with regard to 1-mm spatial resolution and
2–3-s temporal resolution on original acquisition.
Thus, conventional angiography remains the clear
winner in these measures of performance [247].
This variation of spatial and temporal resolu-
tion reflects DSA comparable sensitivity of CTA
and MRA for the feeding arteries and fistulas of
SDAVF [239], against other SVMs. This is
because the feeding arteries of SDAVF are the
radiculomedullary arteries arising from intercostal
arteries, the diameter of which is relatively larger.
However, their major limitation is in the detection
of all the feeding arteries in SVMs (low spatial
resolution) and most importantly the arterialized
draining vein [239] (low temporal resolution),
which is the primary treatment target. Further the
inferior resolution of CTA and MRA limits pre-
cise distinction between the intramedullary AVM
and perimedullary AVF, which is crucial in the
treatment planning.
Contrast resolution is another important dis-
criminator of DSA and noninvasive angiography
1167
(c). Note multiple flow voids over the cord surface (b) that
enhance following contrast administration (c). Time-
resolved MRA (d) confirms the vascular nature of this
pathologically confirmed paraganglioma
techniques. Selective arterial injections in DSA
offer superior contrast resolution (2,000–7,000
HU) compared to contrast administration from
the peripheral venous access in CTA or MRA
[233]. These factors combined with cross-
sectional imaging technique-specific limitations,
such as artifacts, translate into clinically meaning-
ful differences in accuracy between these nonin-
vasive techniques (CTA and MRA) and
conventional angiography [247].
The risk of ionizing radiation and injection of
iodinated contrast material is shared by both CTA
and DSA, where MRA takes a clear edge. Both
CTA and MRA require tedious post-processing of
source images, requiring patience and
experience [228].
MRA is limited by long acquisition time, lim-
ited scanning range, and lower spatial resolution
compared with MDCT [239]. The small scan
range explains the limitation of arterial feeder
detection for SDAVF [240, 246, 248, 249] and
repetition of MRA in published series [240],
1168
Fig. 32 Hypervascular vertebral body metastasis: a
78-year-old female with history of renal cell carcinoma
and back pain presented with sudden onset paraplegia
from vertebral metastasis. Patient underwent surgical
decompression (surgical clips in a, d) that was precluded
by the extensive blood loss. Subsequent preoperative
embolization of this hypervascular metastasis shows
highlighting the possibility of the feeder source
being outside the limited field of view. In general,
however, the imaging quality of MR-based
modalities is easily affected by motion degrada-
tion secondary to respirations, especially in the
thoracolumbar region, and the spatial resolution
does not yet approach the sensitivity or anatomi-
cal detail provided by standard DSA [249]. Of
course, the major merit of MRA is not using
ionization radiation, so the research of MRA has
rich prospect. As MRI continues to improve, it is
likely that in the near future it will play a major
role in the diagnosis of vascular lesions of the
spinal cord.
S.R. Boddu et al.
obvious tumor blush (b, e), which was successfully oblit-
erated by the particle embolization (c, f) minimizing the
intraoperative blood loss. Remember, the torrential
intraoperative blood loss from these hypervascular tumors
without preoperative embolization may precipitate
hypovolemia and spinal cord infarction
Spinal DSA (SpDSA) is a highly skilled pro-
cedure and is technically challenging. Unfortu-
nately, the safety of this procedure has a poor
historical reputation, and there are few current
reports on complication rates. The assumptions
that SpDSA poses a high risk of neurologic com-
plications, requires general anesthesia, and
exposes patients to dangerous doses of radiation
and contrast remain prevalent, despite being
largely based on historical data [250,
251]. These perceptions are maintained, in part,
by the scarcity of contemporary literature on this
subject [76, 231]. Neuroradiologists relatively
less experienced in catheter angiography will
47 Spinal Cord Infarction and Differential Diagnosis
Fig. 33 Intracranial DAVF with vertebral venous drain-
age (type 5): a 72-year-old male with worsening quadri-
plegia. MRI – diffuse edema and swelling involving the
craniocervical junction and cervical cord with multiple
flow voids predominantly over the dorsal cord surface
further skew the imaging in the direction of CTA
and MRA, irrespective of the accuracy of these
noninvasive techniques relative to catheter angi-
ography [252]. More recent evidence based on
302 spinal angiograms showed no intra- or post-
procedure neurological complications and neph-
rotoxicity, which reflects the effect of modern
advances in endovascular technology [253] and
emphasizes that SpDSA in the light of modern
technological advances performed by experienced
practitioners in an appropriate patient population
is relatively a safe procedure.
Summary
In summary, conventional MRI is the initial
screening investigation in the evaluation of spinal
vascular malformation, a common differential
diagnosis for SCI. The main purpose of spinal
CTA and MRA is to display the feeding arteries
and fistulas of vascular malformations and
1169
(a, b) suggestive of underlying vascular shunt. DSA con-
firms the posterior fossa DAVF with arterial supply from
the meningohypophyseal trunk and venous drainage into
the vertebral venous plexus (c–e) resulting in venous con-
gestion and cord edema
provide guidance for DSA. Despite the improve-
ments in technology discussed above, neither
cross-sectional imaging modality, CTA or MRA,
currently approaches the necessary degree of sen-
sitivity and spatial resolution of standard DSA to
rule out spinal vascular lesions. Therefore, DSA
remains the “gold standard” in the diagnosis of
these often-elusive spinal vascular lesions
[234]. 3D RSA is a useful adjunct to conventional
DSA and helps to provide the 3D relationship of
the vasculature to the bone to better distinguish
between arterial and venous anatomy.
Treatment
Ischemia if left untreated becomes progressively
worse over the first few hours [96]. The primary
mechanism of injury reaches plateau by 24 h,
beyond which the injury from the secondary
1170
Fig. 34 Great anterior radiculomedullary vein (GARV).
In spite of the relatively large caliber, the great anterior
radiculomedullary vein (GARV), which is the largest vein
draining the anterior thoracolumbar spinal cord, is easily
mistaken for the artery of Adamkiewicz (AKA) due to its
comparable spatial course and location. This highlights the
importance of high temporal resolution and real-time imag-
ing of DSA in differentiating the small spinal arteries and
veins
mechanism dominates the picture [92]. This high-
lights the importance of acute phase treatment in
the first 24 h in reversing the primary mechanism
and minimizing the damage from the secondary
mechanism. In addition to the supportive mea-
sures, the current treatment strategies for acute
spinal cord ischemia are broadly categorized
based on the primary mechanism of injury, i.e.,
embolism versus hypotension/hypoperfusion.
The primary aim is to treat the primary cause as
soon as possible and restoration of normal spinal
cord circulation. Early resuscitation and support-
ive measures help to minimize the spinal cord
damage from the secondary pathophysiological
mechanisms and may improve the long-term
outcome.
S.R. Boddu et al.
Spinal Cord Ischemia Secondary
to Embolism
Thrombolysis for acute spinal cord ischemia is
still evolving and has only been reported in liter-
ature a few times. Intravenous thrombolysis with
recombinant tissue plasminogen activator was
reported in one patient. This case was quickly
diagnosed as it happened as a complication of an
aortography evaluating a fusiform abdominal
aneurysm and therefore was within a 3-h treat-
ment window. The patient made a remarkable
recovery and was discharged 72 h after with
almost no deficits [254]. Three other cases were
reported, which were treated by introducing strep-
tokinase and steroids selectively via endovascular
therapy. Three injections were given in 1-week
intervals. All patients recovered with no to mini-
mal neurological (one with mild spasticity of the
lower extremities and another with minimal sen-
sory impairment) [255]. The main obstacle to such
interventions is timeliness, as there are contrain-
dications to using thrombolytic therapy [256] and
excluding these, as well as diagnosing spinal cord
ischemia, may take beyond the optimal therapeu-
tic window.
Spinal Cord Ischemia Secondary
to Hypoperfusion
The hypotension/hypoperfusion can be a primary
mechanism caused by myocardial infarction and
internal hemorrhage or may be part of secondary
mechanism of cord damage. The precipitating
event should be addressed immediately. Specific
management guidelines have been developed to
limit or reverse neurological deficits in spinal cord
ischemia after aortic surgery and thoracic
endovascular aortic repair (TEVAR) [257]. There
is a lesser chance of reversing this pathology when
the patient awakes from anesthesia with neurolog-
ical deficit, rather than the findings developing
after a period of normal neurological function
postoperatively. Measures such as increasing sys-
tolic blood pressure to 95 mmHg and decreasing
CSF pressure to 10 mmHg can result in symptom
relief and lead to recovery in a majority of affected
47 Spinal Cord Infarction and Differential Diagnosis
patients with delayed onset symptoms of SCI after
aortic surgery [258]. Alternatively preemptive
measures can include intraoperative neurophysio-
logic monitoring, which can identify SCI, while
the patient is still under general anesthesia.
Lumbar CSF drain placement should be
considered prior to starting the vasopressors in
the event of suspected SCI to maintain CSF pres-
sure between 8 and 12 mmHg. These measures
have been shown to mitigate and, at times, reverse
the neurological sequelae in both intra- and post-
operative scenarios. Other measures include
epidural cooling, sequential sacrifice of intercostal
vessels before aneurysm excision, and the use
of distal aortic perfusion during these
procedures [257].
Treatment of Underlying Cause
The improvement of the neurological function
following the removal of hematomyelia was
documented as early as 1911 in canine studies
[86]. Authors speculated that there was a putative
“biochemical factor” that was present in the hem-
orrhagic necrotic material at the epicenter of
hematomyelia and that may be instigating ongo-
ing damage [87]. Hemorrhage may promote
ischemia [259] or induce thrombosis via platelet
aggregation which in turn may precipitate ische-
mia [132]. Hence the prompt decompression of
the intrinsic spinal hematoma is vital in minimiz-
ing the subsequent injury. The precipitating fac-
tors for hematomyelia such as SVMs and spinal
tumors should be promptly treated to prevent
recurrent hemorrhage.
For treatment of venous hypertension, com-
plete obliteration of SVMs should be performed
with surgery or endovascular embolization. Using
permanent liquid embolic material (glue, Onyx),
the results of embolization and surgery are similar
in safety and efficacy [260]. The critical factors
for good outcome in endovascular treatment
includes the identification of the fistula preopera-
tively, avoidance of temporary embolic material
(PVA particles), and occlusion of draining vein. If
inadvertent embolization of ASA is a concern,
surgery is definitely the best choice [260].
1171
Meticulous search for spinal artery aneurysms
should be carried out in the evaluation of spinal
SAH. Extrinsic compression on the spinal cord
triggering ischemic changes such as epidural
hematoma, abscess, extramedullary tumors, and
metastasis needs timely surgical decompression
and/or radiotherapy. In the event of epidural
abscess, thorough scrutiny should be made for
diskitis/osteomyelitis as well as the source of sys-
temic sepsis. Finally, always check for the syn-
chronous pathologies such as cervical disk and
intramedullary SVM.
Role of High-Dose Steroids
Following acute spinal trauma, administration of
high-dose methylprednisolone within 3 h is more
beneficial than within a 3–8-h window in patients
with complete and incomplete spinal cord inju-
ries. High-dose steroids may improve spinal cord
blood flow and microvascular perfusion [110,
261] as well as clinical neurologic recovery after
experimental spinal cord injury [262]. They may
also provide some cytoprotection through the
inhibition of lipid peroxidation, facilitate spinal
cord impulse generation, and inhibit
prostaglandin-induced vasoconstriction [263].
Because lipid peroxidation begins within the first
5 min after an acute cord injury, administration of
high-dose steroid should occur as close to the time
of injury as possible for maximal efficacy. Given
the similarity of pathophysiological concepts in
secondary mechanism of injury between acute
spinal cord injury and cord infarction, there may
be a beneficial role for high-dose steroids in cord
infarction.
Surgical Decompression
The timing of decompression of the neural ele-
ments and, in particular, the efficacy of early
decompression (within 24 h) in improving neuro-
logic recovery are still a matter of debate [264,
265]. A meta-analysis of studies of early decom-
pression from 1966 through 2000 [266] showed
that surgery performed within 24 h produced a
1172
significant improvement in neurologic recovery
compared with late surgery but concluded that
the evidence was not strong and that early surgery
could be considered only as a practice option.
Starting from this framework, a recent prospective
multicentric study [267] demonstrated that the
odds of achieving a 2 AIS grade improvement
are 2.8 times higher in patients undergoing early
surgical decompression (within 24 h). However, a
recent meta-analysis [268] reported a lack of sta-
tistical robustness of the articles examined; there-
fore, the relationship between early surgery and
better neurological outcome is still to be
demonstrated.
Prognosis
Spinal cord infarction carries poor prognosis with
permanent and disabling sequelae. The mortality
rate of SCI based on published case series ranges
from 9 % to 23 % [1, 10, 147, 166]. SCI with
cardiac arrest, aortic rupture or dissection, and
high cervical involvement carries high risk of
mortality [147]. Among the survivors of the
acute episode, respiratory complications are the
leading cause of death [269]. The leading reported
secondary complications are typically pressure
sores, chills, and urinary sepsis, atelectasis, pneu-
monia, and deep vein thrombosis [84]. Prolonged
immobilization results in joint contractures and
heterotopic calcification [269]. When comparing
tetraplegics and paraplegics, an increased inci-
dence of urinary tract infections and pressure
sores is seen in the former [270]. The social and
economic impact of SCI extends to involve not
just the patient and the immediate family but the
community and society at large [84].
Short-Term Outcome
Initial neurological assessment following Ameri-
can Spinal Injury Association (ASIA) classifica-
tion proves to be the best predictor of prognosis
[1, 147]. Approximately 50 % of patients have
favorable short-term outcome, defined by ability
to walk without or with one assistive device and
S.R. Boddu et al.
no need for urinary catheterization at the time of
hospital discharge [134]. In a series of 36 patients,
the outcome at the time of discharge showed 57 %
confined to wheelchair, 25 % partly mobile with
ambulatory aids, and 18 % fully mobile.
Advanced patient age is a negative prognostic
factor for the functional outcome. Absent muscle
contraction at the time of admission has increased
risk of poor outcome and wheelchair bound
[147]. In a series of 57 patients with ASCIS,
logistic regression analysis showed that poor
grades at presentation (ASIA: A and B) and
female gender are independent predictors of poor
functional outcome [1].
Long-Term Outcome
The long-term functional outcome of ASCIS
depends on the degree of the initial neurological
deficits, especially the motor deficits [1]. Based on
a case series with a mean follow-up of 3 years,
improvement in outcome is possible long after
being discharged from the hospital. About 40 %
of patients will show some kind of improvement
[9]. As an example although about 40 % of
patients are wheelchair bound upon discharge, at
least 40 % of these will be able to walk indepen-
dently in the future [10]. In patients with venous
congestion related to SVMs, motor function is
more likely to improve compared to sensory and
bladder function (60 % vs. 30 %) [271]. Impotence
rarely improves. The presence of a sensory level is
a bad prognostic marker, suggesting irreversible
cord damage [272].
Future Direction
One should not forget the small footsteps in the
evolution of thrombolysis for cerebral stroke. Cer-
tainly the fascinating journey across intravenous
thrombolysis, intra-arterial thrombolysis,
mechanical thrombectomy, revolutionary contri-
bution from imaging, and incredible teamwork
was not instantaneous. With high index of suspi-
cion, better coordination, modern imaging, and
interventional techniques, spinal thrombolysis
47 Spinal Cord Infarction and Differential Diagnosis
could become a treatment of choice in appropriate
patients. Early diagnosis is the key for spinal
thrombolysis. Early initiation of resuscitation
measures to maintain the local perfusion and min-
imize secondary cord damage may widen the cur-
rent 3-h therapeutic window for intra-arterial
thrombolysis. Further research studies in this
area should be encouraged as a contribution
toward evidence-based medicine.
Summary
Thorough understanding of the arterial anatomy
and the rich intrinsic vascular anastomosis includ-
ing the anterior spinal artery is vital in the under-
standing of the spinal cord infarction. High index
of suspicion, awareness of the spinal cord infarc-
tion clinical syndromes, and early magnetic reso-
nance imaging with diffusion-weighted
acquisition in conjunction with the clinical assess-
ment and cerebrospinal fluid analysis help to dif-
ferentiate the spinal cord infarction from its
mimics. Catheter angiography helps to differenti-
ate between the embolic versus hypotensive
infarctions. Early diagnosis, maintaining the
local perfusion thus preventing the free radical-
mediated secondary damage, and timely throm-
bolysis in appropriate patients are the fundamen-
tals of spinal cord infarction treatment.
References
1. Nedeltchev K, Loher TJ, Stepper F et al (2004) Long-
term outcome of acute spinal cord ischemia syn-
drome. Stroke 35(2):560–565. doi:10.1161/01.
STR.0000111598.78198.EC
2. de Seze J, Stojkovic T, Breteau G et al (2001) Acute
myelopathies clinical, laboratory and outcome pro-
files in 79 cases. Brain 124(8):1509–1521.
doi:10.1093/brain/124.8.1509
3. Martinelli V, Comi G, Rovaris M et al (1995) Acute
myelopathy of unknown aetiology: a clinical, neuro-
physiological and MRI study of short- and long-term
prognostic factors. J Neurol 242(8):497–503
4. Woollam DH, Millen JW (1955) The arterial supply
of the spinal cord and its significance. J Neurol
Neurosurg Psychiatry 18(2):97–102
5. Singer D (1902) The pathology of so-called acute
myelitis. Brain 25(2):332–340
1173
6. Slager UT, Webb AT (1973) Pathologic findings in the
spinal cord. Arch Pathol 96(6):388–394
7. Sandson TA, Friedman JH (1989) Spinal cord infarc-
tion. Report of 8 cases and review of the literature.
Medicine (Baltimore) 68(5):282–292
8. Hirtz D, Thurman DJ, Gwinn-Hardy K, Mohamed M,
Chaudhuri AR, Zalutsky R (2007) How common are
the “common” neurologic disorders? Neurology
68(5):326–337. doi:10.1212/01.wnl.0000252807.
38124.a3
9. Cheshire WP, Santos CC, Massey EW, Howard JF Jr
(1996) Spinal cord infarction: etiology and
outcome. Neurology 47(2):321–330. doi:10.1212/
WNL.47.2.321
10. Robertson CE, Brown RD Jr, Wijdicks EFM,
Rabinstein AA (2012) Recovery after spinal cord
infarcts: long-term outcome in 115 patients.
Neurology 78(2):114–121. doi:10.1212/WNL.
0b013e31823efc93
11. Abercrombie J (1831) Pathological and practical
researches on diseases of the brain and the spinal
cord. Carey & Lea, Philadelphia
12. Tweedie A, Gerhard W (1842) A system of practical
medicine comprised in a series of original disserta-
tions. Lea & Blanchard, Philadelphia
13. Santillan A, Nacarino V, Greenberg E, Riina HA,
Gobin YP, Patsalides A (2012) Vascular anatomy of
the spinal cord. J Neurointerv Surg 4(1):67–74.
doi:10.1136/neurintsurg-2011-010018
14. Manelfe C, Lazorthes G, Roulleau J (1972) Arteries
of the human spinal dura mater. Acta Radiol Diagn
(Stockh) 13:829–841
15. Thron A (1988) Vascular anatomy of the spinal cord.
Neuroradiological investigations and clinical
syndromes. Springer, New York
16. Hong MK-Y, Hong MK-H, Pan W-R, Wallace D,
Ashton MW, Taylor GI (2008) The angiosome terri-
tories of the spinal cord: exploring the issue of preop-
erative spinal angiography. Laboratory investigation.
J Neurosurg Spine 8(4):352–364. doi:10.3171/SPI/
2008/8/4/352
17. Lasjaunias P, Berenstein A, TerBrugge K (2001) Clin-
ical vascular anatomy and variations. In: Surgical
neuroangiography, vol 1. Springer, Berlin
18. Piscol K (1972) Blood supply of the spinal cord and
its clinical importance. Schriftenr Neurol 8:1–91
19. Brockstein B, Johns L, Gewertz BL (1994)
Blood supply to the spinal cord: anatomic and
physiologic correlations. Ann Vasc Surg 8(4):
394–399
20. Lazorthes G, Poulhes J, Bastide G, Roulleau J,
Chancholle AR (1958) Arterial vascularization of
the spine; anatomic research and applications in
pathology of the spinal cord and aorta.
Neurochirurgie 4(1):3–19
21. Koshino T, Murakami G, Morishita K, Mawatari T,
Abe T (1999) Does the Adamkiewicz artery originate
from the larger segmental arteries? J Thorac
Cardiovasc Surg 117(5):898–905
1174
22. Hyodoh H, Shirase R, Akiba H et al (2007) Double-
subtraction maximum intensity projection MR angi-
ography for detecting the artery of Adamkiewicz and
differentiating it from the drainage vein. J Magn
Reson Imaging 26(2):359–365. doi:10.1002/
jmri.21024
23. Charles YP, Barbe B, Beaujeux R, Boujan F, Steib J-P
(2011) Relevance of the anatomical location of the
Adamkiewicz artery in spine surgery. Surg Radiol
Anat 33(1):3–9. doi:10.1007/s00276-010-0654-0
24. Garland H, Greenberg J, Harriman DG (1966) Infarc-
tion of the spinal cord. Brain J Neurol 89(4):645–662
25. Sliwa JA, Maclean IC (1992) Ischemic myelopathy: a
review of spinal vasculature and related clinical syn-
dromes. Arch Phys Med Rehabil 73(4):365–372
26. Gilles FH, Nag D (1971) Vulnerability of human
spinal cord in transient cardiac arrest. Neurology
21(8):833–839
27. Azzarelli B, Roessmann U (1977) Diffuse “anoxic”
myelopathy. Neurology 27(11):1049–1052
28. Rothman SM, Nelson JS (1980) Spinal cord infarc-
tion in a patient with sickle cell anemia. Neurology
30(10):1072–1076
29. Houdart R, Djindjian R (1966) Angiomas of the spi-
nal cord. Proc R Soc Med 59(8):787–790
30. Satran R (1988) Spinal cord infarction. Stroke
19(4):529–532
31. Whiteley AM, Hauw JJ, Escourolle R (1979) A path-
ological survey of 41 cases of acute intrinsic spinal
cord disease. J Neurol Sci 42(2):229–242
32. Shakir RA, Sulaiman K, Kahn RA, Rudwan M (1990)
Neurological presentation of neuro-Behçet’s syn-
drome: clinical categories. Eur Neurol 30(5):249–253
33. Gibb WR, Urry PA, Lees AJ (1985) Giant cell arteritis
with spinal cord infarction and basilar artery throm-
bosis. J Neurol Neurosurg Psychiatry 48(9):945–948
34. Sawaya GR, Kaminski MJ (1990) Spinal cord infarc-
tion after cocaine use. South Med J 83(5):601–602
35. Walden JE, Castillo M (2012) Sildenafil-induced cer-
vical spinal cord infarction. AJNR Am J Neuroradiol
33(3):E32–E33. doi:10.3174/ajnr.A2628
36. Adams HD, van Geertruyden HH (1956) Neurologic
complications of aortic surgery. Ann Surg 144(4):
574–609. Available at http://www.ncbi.nlm.nih.
gov/pmc/articles/PMC1465537/. Accessed 20 Mar
2014
37. Xenos ES, Abedi NN, Davenport DL et al (2008)
Meta-analysis of endovascular vs open repair for
traumatic descending thoracic aortic rupture.
J Vasc Surg 48(5):1343–1351. doi:10.1016/j.jvs.
2008.04.060
38. Amabile P, Grisoli D, Giorgi R, Bartoli J-M, Piquet P
(2008) Incidence and determinants of spinal cord
ischaemia in stent-graft repair of the thoracic aorta.
Eur J Vasc Endovasc Surg 35(4):455–461.
doi:10.1016/j.ejvs.2007.11.005
39. Maldonado TS, Rockman CB, Riles E et al (2004)
Ischemic complications after endovascular abdominal
S.R. Boddu et al.
aortic aneurysm repair. J Vasc Surg 40(4):703–709.
doi:10.1016/j.jvs.2004.07.032; discussion 709–710
40. Messé SR, Bavaria JE, Mullen M et al (2008) Neuro-
logic outcomes from high risk descending thoracic
and thoracoabdominal aortic operations in the era of
endovascular repair. Neurocrit Care 9(3):344–351.
doi:10.1007/s12028-008-9104-9
41. Bavaria JE, Appoo JJ, Makaroun MS et al (2007)
Endovascular stent grafting versus open surgical
repair of descending thoracic aortic aneurysms in
low-risk patients: a multicenter comparative trial.
J Thorac Cardiovasc Surg 133(2):369–377.
doi:10.1016/j.jtcvs.2006.07.040
42. Setacci F, Sirignano P, De Donato G et al (2010)
Endovascular thoracic aortic repair and risk of spinal
cord ischemia: the role of previous or concomitant
treatment for aortic aneurysm. J Cardiovasc Surg
(Torino) 51(2):169–176
43. Crawford ES, Crawford JL, Safi HJ et al (1986)
Thoracoabdominal aortic aneurysms: preoperative
and intraoperative factors determining immediate
and long-term results of operations in 605 patients.
J Vasc Surg 3(3):389–404
44. Gloviczki P, Cross SA, Stanson AW et al (1991)
Ischemic injury to the spinal cord or lumbosacral
plexus after aorto-iliac reconstruction. Am J Surg
162(2):131–136
45. Ross RT (1985) Spinal cord infarction in disease and
surgery of the aorta. Can J Neurol Sci 12(4):289–295
46. Mawad ME, Rivera V, Crawford S, Ramirez A,
Breitbach W (1990) Spinal cord ischemia after resec-
tion of thoracoabdominal aortic aneurysms: MR find-
ings in 24 patients. AJNR Am J Neuroradiol
11(5):987–991
47. Brusoni B, Colombo A, Merlo L, Marchetti G, Longo
T (1978) Hemodynamic and metabolic changes
induced by temporary clamping of the thoracic
aorta. Eur Surg Res 10(3):206–216
48. Hurst RW, Haskal ZJ, Zager E, Bagley LJ, Flamm ES
(1998) Endovascular stent treatment of cervical inter-
nal carotid artery aneurysms with parent vessel pres-
ervation. Surg Neurol 50(4):313–317; discussion 317.
Available at http://www.ncbi.nlm.nih.gov/pubmed/
9817452. Accessed 20 Mar 2012
49. Hughes JT, Brownell B (1966) Spinal cord ischemia
due to arteriosclerosis. Arch Neurol 15(2):189–202
50. Tubbs RS, Blouir MC, Romeo AK, Mortazavi MM,
Cohen-Gadol AA (2011) Spinal cord ischemia and
atherosclerosis: a review of the literature. Br J
Neurosurg 25(6):666–670. doi:10.3109/02688697.
2011.578774
51. Duggal N, Lach B (2002) Selective vulnerability of
the lumbosacral spinal cord after cardiac arrest and
hypotension. Stroke 33(1):116–121. doi:10.1161/
hs0102.101923
52. Slavin RE, Gonzalez-Vitale JC, Marin OS (1975)
Atheromatous emboli to the lumbosacral spinal
cord. Stroke 6(4):411–415
47 Spinal Cord Infarction and Differential Diagnosis
53. Coleman J (1898) Dissecting aneurysm. Trans R
Acad Med Irel 16(1):116–124
54. Tracci C, Cherry KJ Jr (2012) The aorta. In:
Townsend CM (ed) Sabiston textbook of surgery.
Saunders/Elsevier, Philadelphia, pp 1697–1724
55. Gaul C, Dietrich W, Friedrich I, Sirch J, Erbguth FJ
(2007) Neurological symptoms in type A aortic dis-
sections. Stroke 38(2):292–297. doi:10.1161/01.
STR.0000254594.33408.b1
56. Suzuki T, Mehta RH, Ince H et al (2003) Clinical
profiles and outcomes of acute type B aortic dissec-
tion in the current era: lessons from the International
Registry of Aortic Dissection (IRAD). Circulation
108(Suppl 1):II312–II317. doi:10.1161/01.cir.
0000087386.07204.09
57. Sladky JT, Rorke LB (1986) Perinatal hypoxic/ische-
mic spinal cord injury. Pediatr Pathol 6(1):87–101
58. Thompson TP, Pearce J, Chang G, Madamba J (2004)
Surfer’s myelopathy. Spine 29(16):E353–E356
59. Chang CWJ, Donovan DJ, Liem LK et al (2012)
Surfers’ myelopathy: a case series of 19 novice
surfers with nontraumatic myelopathy. Neurology
79(22):2171–2176. doi:10.1212/WNL.0b013e31
827595cd
60. Shamji MF, Maziak DE, Shamji FM, Ginsberg RJ,
Pon R (2003) Circulation of the spinal cord: an impor-
tant consideration for thoracic surgeons. Ann Thorac
Surg 76(1):315–321
61. Nakamoto BK, Siu AM, Hashiba KA et al (2013)
Surfer’s myelopathy: a radiologic study of 23 cases.
AJNR Am J Neuroradiol 34(12):2393–2398.
doi:10.3174/ajnr.A3599
62. Hirose G, Singer P, Bass NH (1971) Successful treat-
ment of posthypoxic action myoclonus with carba-
mazepine. JAMA 218(9):1432–1433
63. Mori S, Sadoshima S, Tagawa K, Iino K, Fujishima M
(1993) Massive spinal cord infarction with multiple
paradoxical embolism: a case report. Angiology
44(3):251–256
64. Naiman JL, Donohue WL, Prichard JS (1961) Fatal
nucleus pulposus embolism of spinal cord after
trauma. Neurology 11:83–87
65. Tosi L, Rigoli G, Beltramello A (1996) Fibrocarti-
laginous embolism of the spinal cord: a clinical and
pathogenetic reconsideration. J Neurol Neurosurg
Psychiatry 60(1):55–60
66. Srigley JR, Lambert CD, Bilbao JM, Pritzker KP
(1981) Spinal cord infarction secondary to
intervertebral disc embolism. Ann Neurol
9(3):296–301. doi:10.1002/ana.410090315
67. Mathew P, Todd NV, Hadley DM, Adams JH (1993)
Spinal cord infarction following meningitis. Br J
Neurosurg 7(6):701–704
68. Von Pohle WR (1996) Disseminated mucormycosis
presenting with lower extremity weakness. Eur Respir
J 9(8):1751–1753
69. Brown AC, Ray CE (2012) Anterior spinal cord
infarction following bronchial artery embolization.
1175
Semin Intervent Radiol 29(3):241–244. doi:10.1055/
s-0032-1326936
70. Gang DL, Dole KB, Adelman LS (1977) Spinal cord
infarction following therapeutic renal artery emboli-
zation. JAMA 237(26):2841–2842
71. Vujic I, Pyle R, Parker E, Mithoefer J (1980) Control
of massive hemoptysis by embolization of intercostal
arteries. Radiology 137(3):617–620. doi:10.1148/
radiology.137.3.7444046
72. Ozoilo K, Stein M (2013) Paraplegia complicating
embolization for bleeding intercostal artery in pene-
trating trauma. Inj Extra 44(8):70–73. doi:10.1016/j.
injury.2013.05.015
73. Heller SL, Meyer JR, Russell EJ (1996) Spinal cord
venous infarction following endoscopic sclerotherapy
for esophageal varices. Neurology 47(4):1081–1085
74. Tofuku K, Koga H, Yamamoto T, Yone K, Komiya S
(2008) Spinal cord infarction following endoscopic
variceal ligation. Spinal Cord 46(3):241–242.
doi:10.1038/sj.sc.3102092
75. Neal JM, Bernards CM, Hadzic A et al (2008) ASRA
practice advisory on neurologic complications in
regional anesthesia and pain mMedicine. Reg Anesth
Pain Med 33(5):404–415. doi:10.1016/j.rapm.2008.
07.527
76. Forbes G, Nichols DA, Jack CR Jr et al (1988) Com-
plications of spinal cord arteriography: prospective
assessment of risk for diagnostic procedures.
Radiology 169(2):479–484. doi:10.1148/radiology.
169.2.3174997
77. Gonzalez LF, Zabramski JM, Tabrizi P, Wallace RC,
Massand MG, Spetzler RF (2005) Spontaneous spinal
subarachnoid hemorrhage secondary to spinal aneu-
rysms: diagnosis and treatment paradigm. Neurosur-
gery 57(6):1127–1131; discussion 1127–1131
78. Krishna V, Lazaridis C, Ellegala D et al (2012) Spinal
cord infarction associated with subarachnoid hemor-
rhage. Clin Neurol Neurosurg 114(7):1030–1032.
doi:10.1016/j.clineuro.2012.01.037
79. Anthes DL, Theriault E, Tator CH (1996) Ultrastruc-
tural evidence for arteriolar vasospasm after spinal
cord trauma. Neurosurgery 39(4):804–814
80. Krings T, Lasjaunias PL, Hans FJ et al (2007)
Imaging in spinal vascular disease. Neuroimaging
Clin N Am 17(1):57–72. doi:10.1016/j.nic.2007.
01.001
81. Blatteau J-E, Gempp E, Simon O et al (2011) Prog-
nostic factors of spinal cord decompression sickness
in recreational diving: retrospective and multicentric
analysis of 279 cases. Neurocrit Care 15(1):120–127.
doi:10.1007/s12028-010-9370-1
82. Aharon-Peretz J, Adir Y, Gordon CR, Kol S, Gal N,
Melamed Y (1993) Spinal cord decompression sick-
ness in sport diving. Arch Neurol 50(7):753–756
83. Kim RC, Smith HR, Henbest ML, Choi BH (1984)
Nonhemorrhagic venous infarction of the spinal cord.
Ann Neurol 15(4):379–385. doi:10.1002/ana.
410150413
1176
84. Sekhon LH, Fehlings MG (2001) Epidemiology,
demographics, and pathophysiology of acute spinal
cord injury. Spine 26(24 Suppl):S2–S12
85. Fehlings MG, Sekhon LH (2000) Cellular, ionic and
biomolecular mechanisms of the injury process. In:
Tator CH, Benzel EC (eds) Contemporary manage-
ment of spinal cord injury: from impact to rehabilita-
tion. American Association of Neurological
Surgeons, New York, pp 33–50
86. Allen A (1911) Surgery of experimental lesion of
spinal cord equivalent to crush injury of fracture dis-
location of spinal column. A preliminary report.
JAMA 57:878–880
87. Allen A (1914) Remarks on the histopathological
changes in the spinal cord due to impact an experi-
mental study. J Nerv Ment Dis 31:141–147
88. Demopoulos H, Flamm ES, Seligman M (1979)
Membrane perturbations in central nervous system
injury: theoretical basis for free radical damage and
a review of the experimental data. In: Popp AJ
(ed) Neural trauma. Raven Press, New York,
pp 63–78
89. Anderson DK, Hall ED (1993) Pathophysiology of
spinal cord trauma. Ann Emerg Med 22(6):987–992
90. Collins WF (1983) A review and update of experi-
ment and clinical studies of spinal cord injury. Para-
plegia 21(4):204–219. doi:10.1038/sc.1983.34
91. Dusart I, Schwab ME (1994) Secondary cell death
and the inflammatory reaction after dorsal
hemisection of the rat spinal cord. Eur J Neurosci
6(5):712–724
92. Tator CH (1991) Review of experimental spinal cord
injury with emphasis on the local and systemic circu-
latory effects. Neurochirurgie 37(5):291–302
93. Young W, Koreh I (1986) Potassium and calcium
changes in injured spinal cords. Brain Res
365(1):42–53
94. Hall ED, Wolf DL (1987) Post-traumatic spinal cord
ischemia: relationship to injury severity and physio-
logical parameters. Cent Nerv Syst Trauma
4(1):15–25
95. Hall ED, Braughler JM (1982) Effects of intravenous
methylprednisolone on spinal cord lipid peroxidation
and Na+ + K+)-ATPase activity. Dose-response anal-
ysis during 1st hour after contusion injury in the cat.
J Neurosurg 57(2):247–253. doi:10.3171/jns.1982.
57.2.0247
96. Fehlings MG, Tator CH, Linden RD (1989) The effect
of nimodipine and dextran on axonal function and
blood flow following experimental spinal cord injury.
J Neurosurg 71(3):403–416. doi:10.3171/jns.1989.
71.3.0403
97. Tator CH, Fehlings MG (1991) Review of the sec-
ondary injury theory of acute spinal cord trauma with
emphasis on vascular mechanisms. J Neurosurg
75(1):15–26. doi:10.3171/jns.1991.75.1.0015
98. Senter HJ, Venes JL (1979) Loss of autoregulation
and posttraumatic ischemia following experimental
S.R. Boddu et al.
spinal cord trauma. J Neurosurg 50(2):198–206.
doi:10.3171/jns.1979.50.2.0198
99. Wolman L (1965) The disturbance of circulation in
traumatic paraplegia in acute and late stages: a path-
ological study. Paraplegia 2:213–226. doi:10.1038/
sc.1964.39
100. Sandler AN, Tator CH (1976) Regional spinal cord
blood flow in primates. J Neurosurg 45(6):647–659.
doi:10.3171/jns.1976.45.6.0647
101. Blisard KS, Follis F, Wong R, Miller KB, Wernly JA,
Scremin OU (1995) Degeneration of axons in the
corticospinal tract secondary to spinal cord ischemia in
rats. Paraplegia 33(3):136–140. doi:10.1038/sc.1995.30
102. Suh T, Alexander L (1939) Vascular system of
the human spinal cord. Arch Neurol Psychiatry
1939(41):660–676
103. Grassner L, Klausner F, Wagner M et al (2014) Acute
and chronic evolution of MRI findings in a case
of posterior spinal cord ischemia. Spinal Cord
52:S23–S24. doi:10.1038/sc.2013.165
104. Fried LC, Goodkin R (1971) Microangiographic
observations of the experimentally traumatized spinal
cord. J Neurosurg 35(6):709–714. doi:10.3171/
jns.1971.35.6.0709
105. Dohrmann GJ, Wick KM (1973) Intramedullary
blood flow patterns in transitory traumatic paraplegia.
Surg Neurol 1(4):209–215
106. Assenmacher DR, Ducker TB (1971) Experimental
traumatic paraplegia. The vascular and pathological
changes seen in reversible and irreversible spinal-cord
lesions. J Bone Joint Surg Am 53(4):671–680
107. Griffiths IR, Miller R (1974) Vascular permeability to
protein and vasogenic oedema in experimental con-
cussive injuries to the canine spinal cord. J Neurol Sci
22(3):291–304
108. Hsu CY, Hogan EL, Gadsden RH Sr, Spicer KM, Shi
MP, Cox RD (1985) Vascular permeability in exper-
imental spinal cord injury. J Neurol Sci 70(3):
275–282
109. Stewart WB, Wagner FC (1979) Vascular permeabil-
ity changes in the contused feline spinal cord. Brain
Res 169(1):163–167
110. Young W, Flamm ES (1982) Effect of high-dose
corticosteroid therapy on blood flow, evoked poten-
tials, and extracellular calcium in experimental spinal
injury. J Neurosurg 57(5):667–673. doi:10.3171/
jns.1982.57.5.0667
111. Young W, DeCrescito V, Tomasula JJ, Ho V (1980)
The role of the sympathetic nervous system in pressor
responses induced by spinal injury. J Neurosurg
52(4):473–481. doi:10.3171/jns.1980.52.4.0473
112. Atkinson PP, Atkinson JL (1996) Spinal shock. Mayo
Clin Proc 71(4):384–389. doi:10.1016/S0025-6196
(11)64067-6
113. Henderson CE (1996) Programmed cell death in the
developing nervous system. Neuron 17(4):579–585
114. Namura S, Zhu J, Fink K et al (1998) Activation and
cleavage of caspase-3 in apoptosis induced by
47 Spinal Cord Infarction and Differential Diagnosis
experimental cerebral ischemia. J Neurosci
18(10):3659–3668
115. Hockenbery D (1995) Defining apoptosis. Am J
Pathol 146(1):16–19
116. Crowe MJ, Bresnahan JC, Shuman SL, Masters JN,
Beattie MS (1997) Apoptosis and delayed degenera-
tion after spinal cord injury in rats and monkeys. Nat
Med 3(1):73–76
117. Emery E, Aldana P, Bunge MB et al (1998) Apoptosis
after traumatic human spinal cord injury. J Neurosurg
89(6):911–920. doi:10.3171/jns.1998.89.6.0911
118. Li GL, Brodin G, Farooque M et al (1996) Apoptosis
and expression of Bcl-2 after compression trauma to rat
spinal cord. J Neuropathol Exp Neurol 55(3):280–289
119. Tator CH, Koyanagi I (1997) Vascular mechanisms in
the pathophysiology of human spinal cord injury.
J Neurosurg 86(3):483–492. doi:10.3171/jns.1997.
86.3.0483
120. Agrawal SK, Fehlings MG (1996) Mechanisms of
secondary injury to spinal cord axons in vitro: role
of Na+, Na(+)-K(+)-ATPase, the Na(+)-H+ exchanger,
and the Na(+)-Ca2+ exchanger. J Neurosci 16(2):
545–552
121. Osterholm JL, Mathews GJ (1972) Altered norepi-
nephrine metabolism, following experimental spinal
cord injury. 2. Protection against traumatic spinal cord
hemorrhagic necrosis by norepinephrine synthesis
blockade with alpha methyl tyrosine. J Neurosurg
36(4):395–401. doi:10.3171/jns.1972.36.4.0395
122. Agrawal SK, Fehlings MG (1997) Role of NMDA
and non-NMDA ionotropic glutamate receptors in
traumatic spinal cord axonal injury. J Neurosci
17(3):1055–1063
123. Faden AI, Simon RP (1988) A potential role for
excitotoxins in the pathophysiology of spinal cord
injury. Ann Neurol 23(6):623–626. doi:10.1002/
ana.410230618
124. Demopoulos HB, Flamm ES, Pietronigro DD,
Seligman ML (1980) The free radical pathology and
the microcirculation in the major central nervous sys-
tem disorders. Acta Physiol Scand Suppl 492:91–119
125. Hall ED, Yonkers PA, Horan KL, Braughler JM
(1989) Correlation between attenuation of
posttraumatic spinal cord ischemia and preservation
of tissue vitamin E by the 21-aminosteroid U74006F:
evidence for an in vivo antioxidant mechanism.
J Neurotrauma 6(3):169–176
126. Hung TK, Albin MS, Brown TD, Bunegin L, Albin R,
Jannetta PJ (1975) Biomechanical responses to open
experimental spinal cord injury. Surg Neurol 4(2):
271–276
127. Faden AI, Jacobs TP, Holaday JW (1982) Compari-
son of early and late naloxone treatment in experi-
mental spinal injury. Neurology 32(6):677–681
128. Faden AI, Jacobs TP, Smith MT (1984) Evaluation of
the calcium channel antagonist nimodipine in exper-
imental spinal cord ischemia. J Neurosurg 60(4):
796–799. doi:10.3171/jns.1984.60.4.0796
1177
129. Wagner FC Jr, Stewart WB (1981) Effect of trauma
dose on spinal cord edema. J Neurosurg
54(6):802–806. doi:10.3171/jns.1981.54.6.0802
130. Anderson DK, Means ED, Waters TR, Spears CJ
(1980) Spinal cord energy metabolism following
compression trauma to the feline spinal cord.
J Neurosurg 53(3):375–380. doi:10.3171/jns.1980.
53.3.0375
131. Casha S, Yu WR, Fehlings MG (2001) Oligodendrog-
lial apoptosis occurs along degenerating axons and is
associated with FAS and p75 expression following
spinal cord injury in the rat. Neuroscience 103
(1):203–218
132. De la Torre JC (1981) Spinal cord injury. Review of
basic and applied research. Spine 6(4):315–335
133. Lou J, Lenke LG, Ludwig FJ, O’Brien MF (1998)
Apoptosis as a mechanism of neuronal cell death
following acute experimental spinal cord injury.
Spinal Cord 36(10):683–690
134. Masson C, Pruvo JP, Meder JF et al (2004) Spinal
cord infarction: clinical and magnetic resonance
imaging findings and short term outcome. J Neurol
Neurosurg Psychiatry 75(10):1431–1435. doi:10.1136/
jnnp.2003.031724
135. Geldmacher D (2012) Vascular diseases of the ner-
vous system: spinal cord vascular disease. In: Daroff
RB, Bradley WG (eds) Bradley’s neurology in clini-
cal practice, 6th edn. Saunders/Elsevier, Philadelphia,
pp 1096–1099
136. Cheng M-Y, Lyu R-K, Chang Y-J et al (2008)
Spinal cord infarction in Chinese patients. Clinical
features, risk factors, imaging and prognosis.
Cerebrovasc Dis 26(5):502–508. doi:10.1159/
000155988
137. Maynard FM Jr, Bracken MB, Creasey G et al (1997)
International standards for neurological and func-
tional classification of spinal cord injury. American
Spinal Injury Association. Spinal Cord 35(5):
266–274
138. Misulis K (2012) Hemiplegia and monoplegia. In:
Daroff RB, Mazziotta JC (eds) Bradley’s neurology
in clinical practice, 6th edn. Saunders/Elsevier,
Philadelphia, p 277
139. Dobkin B, Havton L (2012) Paraplegia and spinal
cord syndromes. In: Bradley’s neurology in clinical
practice, 6th edn. Saunders/Elsevier, Philadelphia, pp
286–289
140. Vertinsky AT, Krasnokutsky MV, Augustin M,
Bammer R (2007) Cutting-edge imaging of the
spine. Neuroimaging Clin N Am 17(1):117–136.
doi:10.1016/j.nic.2007.01.003
141. Roemer PB, Edelstein WA, Hayes CE, Souza SP,
Mueller OM (1990) The NMR phased array. Magn
Reson Med 16(2):192–225
142. Griswold MA, Jakob PM, Heidemann RM
et al (2002) Generalized autocalibrating partially
parallel acquisitions (GRAPPA). Magn Reson Med
47(6):1202–1210. doi:10.1002/mrm.10171
1178
143. Pruessmann KP, Weiger M, Scheidegger MB,
Boesiger P (1999) SENSE: sensitivity encoding for
fast MRI. Magn Reson Med 42(5):952–962
144. Thurnher MM, Bammer R (2006) Diffusion-weighted
MR imaging (DWI) in spinal cord ischemia. Neuro-
radiology 48(11):795–801. doi:10.1007/s00234-006-
0130-z
145. Marcel C, Kremer S, Jeantroux J, Blanc F, Dietemann
J-L, De Sèze J (2010) Diffusion-weighted imaging in
noncompressive myelopathies: a 33-patient prospec-
tive study. J Neurol 257(9):1438–1445. doi:10.1007/
s00415-010-5538-z
146. Novy J, Carruzzo A, Maeder P, Bogousslavsky J
(2006) Spinal cord ischemia: clinical and imaging
patterns, pathogenesis, and outcomes in 27 patients.
Arch Neurol 63(8):1113–1120. doi:10.1001/
archneur.63.8.1113
147. Salvador de la Barrera S, Barca-Buyo A, Montoto-
Marqués A, Ferreiro-Velasco ME, Cidoncha-Dans M,
Rodriguez-Sotillo A (2001) Spinal cord infarction:
prognosis and recovery in a series of 36 patients.
Spinal Cord 39(10):520–525. doi:10.1038/sj.
sc.3101201
148. Hajnal JV, Doran M, Hall AS et al (1991) MR imag-
ing of anisotropically restricted diffusion of water in
the nervous system: technical, anatomic, and patho-
logic considerations. J Comput Assist Tomogr
15(1):1–18
149. Stepper F, Lövblad KO (2001) Anterior spinal artery
stroke demonstrated by echo-planar DWI. Eur Radiol
11(12):2607–2610. doi:10.1007/s003300100926
150. Schlaug G, Siewert B, Benfield A, Edelman RR,
Warach S (1997) Time course of the apparent diffu-
sion coefficient (ADC) abnormality in human stroke.
Neurology 49(1):113–119
151. Nogueira RG, Ferreira R, Grant PE et al (2012)
Restricted diffusion in spinal cord infarction demon-
strated by magnetic resonance line scan diffusion
imaging. Stroke 43(2):532–535. doi:10.1161/
STROKEAHA.111.624023
152. K€uker W, Weller M, Klose U, Krapf H, Dichgans J,
Nägele T (2004) Diffusion-weighted MRI of spinal
cord infarction–high resolution imaging and time
course of diffusion abnormality. J Neurol
251(7):818–824. doi:10.1007/s00415-004-0434-z
153. Bammer R, Fazekas F, Augustin M et al (2000)
Diffusion-weighted MR imaging of the spinal cord.
AJNR Am J Neuroradiol 21(3):587–591
154. Gass A, Back T, Behrens S, Maras A (2000) MRI of
spinal cord infarction. Neurology 54(11):2195
155. Zhang J, Huan Y, Qian Y, Sun L, Ge Y (2005)
Multishot diffusion-weighted imaging features in spi-
nal cord infarction. J Spinal Disord Tech 18(3):
277–282
156. Fortuna A, Ferrante L, Acqui M, Trillò G (1995)
Spinal cord ischemia diagnosed by MRI. Case report
and review of the literature. J Neuroradiol
22(2):115–122
S.R. Boddu et al.
157. Weidauer S, Nichtweiss M, Lanfermann H, Zanella
FE (2002) Spinal cord infarction: MR imaging and
clinical features in 16 cases. Neuroradiology
44(10):851–857. doi:10.1007/s00234-002-0828-5
158. Beslow LA, Ichord RN, Zimmerman RA, Smith SE,
Licht DJ (2008) Role of diffusion MRI in diagnosis of
spinal cord infarction in children. Neuropediatrics
39(3):188–191. doi:10.1055/s-0028-1093335
159. Mull M, Thron A (2006) Spinal infarcts. In: Magnetic
resonance imaging in ischemic stroke. Springer,
Berlin, pp 251–267
160. Faig J, Busse O, Salbeck R (1998) Vertebral body
infarction as a confirmatory sign of spinal cord ische-
mic stroke: report of three cases and review of the
literature. Stroke 29(1):239–243
161. Mikulis DJ, Ogilvy CS, McKee A, Davis KR, Ojeman
RG (1992) Spinal cord infarction and fibrocarti-
lagenous emboli. AJNR Am J Neuroradiol 13(1):
155–160
162. Yuh WT, Marsh EE, Wang AK et al (1992) MR
imaging of spinal cord and vertebral body infarction.
AJNR Am J Neuroradiol 13(1):145–154. Available
at http://www.ajnr.org/content/13/1/145. Accessed
18 Mar 2014
163. Haddad MC, Aabed al-Thagafi MY, Djurberg H
(1996) MRI of spinal cord and vertebral body infarc-
tion in the anterior spinal artery syndrome. Neurora-
diology 38(2):161–162
164. Cheng M-Y, Lyu R-K, Chang Y-J et al (2009) Con-
comitant spinal cord and vertebral body infarction is
highly associated with aortic pathology: a clinical and
magnetic resonance imaging study. J Neurol
256(9):1418–1426. doi:10.1007/s00415-009-5126-2
165. Hurst WR (2009) Vascular disorders of the spine and
spinal cord. In: Atlas SW (ed) Magnetic resonance
imaging of the brain and spine. Wolters Kluwer
Health/Lippincott Williams & Wilkins, Philadelphia
166. Masson C, Leys D, Meder JF, Dousset V, Pruvo JP
(2004) Spinal cord ischemia. J Neuroradiol 31(1):
35–46
167. Chang Y, Jung T-D, Yoo DS, Hyun JK (2010) Diffu-
sion tensor imaging and fiber tractography of patients
with cervical spinal cord injury. J Neurotrauma
27(11):2033–2040. doi:10.1089/neu.2009.1265
168. Rajasekaran S, Kanna RM, Shetty AP, Ilayaraja V
(2012) Efficacy of diffusion tensor anisotropy indices
and tractography in assessing the extent of severity of
spinal cord injury: an in vitro analytical study in calf
spinal cords. Spine J 12(12):1147–1153. doi:10.1016/
j.spinee.2012.10.032
169. Luo C-B, Chang F-C, Teng MM-H, Chen S-S, Lirng
J-F, Chang C-Y (2003) Magnetic resonance
imaging as a guide in the diagnosis and follow-up
of spinal cord infarction. J Chin Med Assoc 66(2):
89–95
170. Ikuta F, Zimmerman HM (1976) Distribution of
plaques in seventy autopsy cases of multiple sclerosis
in the United States. Neurology 26(6 PT 2):26–28
47 Spinal Cord Infarction and Differential Diagnosis
171. Lycklama G, Thompson A, Filippi M et al (2003)
Spinal-cord MRI in multiple sclerosis. Lancet Neurol
2(9):555–562
172. Thurnher MM, Cartes-Zumelzu F, Mueller-Mang C
(2007) Demyelinating and infectious diseases of the
spinal cord. Neuroimaging Clin N Am 17(1):37–55.
doi:10.1016/j.nic.2006.12.002
173. Bachmann S, Kesselring J (1998) Multiple sclerosis
and infectious childhood diseases. Neuroepi-
demiology 17(3):154–160
174. Bot JCJ, Blezer ELA, Kamphorst W et al (2004) The
spinal cord in multiple sclerosis: relationship of high-
spatial-resolution quantitative MR imaging findings
to histopathologic results. Radiology 233(2):
531–540. doi:10.1148/radiol.2332031572
175. Evangelou N, DeLuca GC, Owens T, Esiri MM
(2005) Pathological study of spinal cord atrophy in
multiple sclerosis suggests limited role of local
lesions. Brain 128(Pt 1):29–34. doi:10.1093/brain/
awh323
176. Tench CR, Morgan PS, Jaspan T, Auer DP,
Constantinescu CS (2005) Spinal cord imaging in mul-
tiple sclerosis. J Neuroimaging 15(4 Suppl):94S–102S.
doi:10.1177/1051228405283292
177. Bot JCJ, Barkhof F, Lycklama à Nijeholt G
et al (2002) Differentiation of multiple sclerosis
from other inflammatory disorders and cerebrovascu-
lar disease: value of spinal MR imaging. Radiology
223(1):46–56. doi:10.1148/radiol.2231010707
178. Paty DW, Oger JJ, Kastrukoff LF et al (1988) MRI in
the diagnosis of MS: a prospective study with com-
parison of clinical evaluation, evoked potentials,
oligoclonal banding, and CT. Neurology 38(2):
180–185
179. Fazekas F, Offenbacher H, Fuchs S et al (1988)
Criteria for an increased specificity of MRI interpre-
tation in elderly subjects with suspected multiple
sclerosis. Neurology 38(12):1822–1825
180. Barkhof F, Filippi M, Miller DH et al (1997) Com-
parison of MRI criteria at first presentation to predict
conversion to clinically definite multiple sclerosis.
Brain 120(Pt 11):2059–2069
181. Kerr DA, Ayetey H (2002) Immunopathogenesis of
acute transverse myelitis. Curr Opin Neurol 15(3):
339–347
182. Choi KH, Lee KS, Chung SO et al (1996) Idiopathic
transverse myelitis: MR characteristics. AJNR Am J
Neuroradiol 17(6):1151–1160
183. Larner AJ, Farmer SF (2000) Myelopathy following
influenza vaccination in inflammatory CNS disorder
treated with chronic immunosuppression. Eur J
Neurol 7(6):731–733
184. Bakshi R, Mazziotta JC (1996) Acute transverse
myelitis after influenza vaccination: magnetic reso-
nance imaging findings. J Neuroimaging 6(4):
248–250
185. De Seze J, Lanctin C, Lebrun C et al (2005) Idiopathic
acute transverse myelitis: application of the recent
1179
diagnostic criteria. Neurology 65(12):1950–1953.
doi:10.1212/01.wnl.0000188896.48308.26
186. Transverse Myelitis Consortium Working Group
(2002) Proposed diagnostic criteria and nosology of
acute transverse myelitis. Neurology 59(4):499–505
187. Scotti G, Gerevini S (2001) Diagnosis and differential
diagnosis of acute transverse myelopathy. The role of
neuroradiological investigations and review of the
literature. Neurol Sci 22(Suppl 2):S69–S73
188. Nance JR, Golomb MR (2007) Ischemic spinal cord
infarction in children without vertebral fracture.
Pediatr Neurol 36(4):209–216. doi:10.1016/j.
pediatrneurol.2007.01.006
189. Alper G (2012) Acute disseminated encephalomyeli-
tis. J Child Neurol 27(11):1408–1425. doi:10.1177/
0883073812455104
190. Singh S, Prabhakar S, Korah IP, Warade SS, Alexan-
der M (2000) Acute disseminated encephalomyelitis
and multiple sclerosis: magnetic resonance imaging
differentiation. Australas Radiol 44(4):404–411
191. Khong P-L, Ho H-K, Cheng P-W, Wong VCN,
Goh W, Chan F-L (2002) Childhood acute dissemi-
nated encephalomyelitis: the role of brain and spinal
cord MRI. Pediatr Radiol 32(1):59–66. doi:10.1007/
s00247-001-0582-6
192. Callen DJA, Shroff MM, Branson HM et al (2009)
Role of MRI in the differentiation of ADEM from MS
in children. Neurology 72(11):968–973. doi:10.1212/
01.wnl.0000338630.20412.45
193. Baum PA, Barkovich AJ, Koch TK, Berg BO (1994)
Deep gray matter involvement in children with acute
disseminated encephalomyelitis. AJNR Am J
Neuroradiol 15(7):1275–1283
194. Pittock SJ, Lucchinetti CF (2006) Inflammatory trans-
verse myelitis: evolving concepts. Curr Opin Neurol
19(4):362–368. doi:10.1097/01.wco.0000236615.
59215.d3
195. Apak RA, Köse G, Anlar B, Turanli G, Topaloğlu H,
Ozdirim E (1999) Acute disseminated encephalomy-
elitis in childhood: report of 10 cases. J Child Neurol
14(3):198–201
196. Thron A, Caplan L (2003) Vascular malformations
and interventional neuroradiology of the spinal cord.
In: Neurological disorderscourse and treatment.
Academic, Amsterdam, pp 517–528
197. Koenig E, Thron A, Schrader V, Dichgans J (1989)
Spinal arteriovenous malformations and fistulae: clin-
ical, neuroradiological and neurophysiological find-
ings. J Neurol 236(5):260–266
198. Hurst RW, Kenyon LC, Lavi E, Raps EC, Marcotte P
(1995) Spinal dural arteriovenous fistula: the pathol-
ogy of venous hypertensive myelopathy. Neurology
45(7):1309–1313
199. Criscuolo GR, Oldfield EH, Doppman JL (1989)
Reversible acute and subacute myelopathy in patients
with dural arteriovenous fistulas. Foix-Alajouanine
syndrome reconsidered. J Neurosurg 70(3):354–359.
doi:10.3171/jns.1989.70.3.0354
1180
200. De Marco JK, Dillon WP, Halback VV, Tsuruda JS
(1990) Dural arteriovenous fistulas: evaluation with
MR imaging. Radiology 175(1):193–199.
doi:10.1148/radiology.175.1.2315480
201. Hurst RW, Grossman RI (2000) Peripheral spinal cord
hypointensity on T2-weighted MR images: a reliable
imaging sign of venous hypertensive myelopathy.
AJNR Am J Neuroradiol 21(4):781–786
202. Gomori JM, Grossman RI, Yu-Ip C, Asakura T (1987)
NMR relaxation times of blood: dependence on field
strength, oxidation state, and cell integrity. J Comput
Assist Tomogr 11(4):684–690
203. Krings T, Geibprasert S (2009) Spinal dural arterio-
venous fistulas. AJNR Am J Neuroradiol 30
(4):639–648. doi:10.3174/ajnr.A1485
204. Weinzierl MR, Krings T, Korinth MC, Reinges MHT,
Gilsbach JM (2004) MRI and intraoperative findings in
cavernous haemangiomas of the spinal cord. Neurora-
diology 46(1):65–71. doi:10.1007/s00234-003-1072-3
205. Balériaux DL (1999) Spinal cord tumors. Eur Radiol
9(7):1252–1258
206. Lowe GM (2000) Magnetic resonance imaging of
intramedullary spinal cord tumors. J Neurooncol
47(3):195–210
207. Chu BC, Terae S, Hida K, Furukawa M, Abe S,
Miyasaka K (2001) MR findings in spinal hemangio-
blastoma: correlation with symptoms and with angio-
graphic and surgical findings. AJNR Am J
Neuroradiol 22(1):206–217
208. Wippold FJ 2nd, Smirniotopoulos JG, Moran CJ,
Suojanen JN, Vollmer DG (1995) MR imaging of
myxopapillary ependymoma: findings and value to
determine extent of tumor and its relation to
intraspinal structures. AJR Am J Roentgenol
165(5):1263–1267. doi:10.2214/ajr.165.5.7572515
209. Miyazawa N, Hida K, Iwasaki Y, Koyanagi I, Abe H
(2000) MRI at 1.5 T of intramedullary ependymoma
and classification of pattern of contrast enhancement.
Neuroradiology 42(11):828–832
210. Richard S, Campello C, Taillandier L, Parker F,
Resche F (1998) Haemangioblastoma of the central
nervous system in von Hippel-Lindau disease. French
VHL Study Group. J Intern Med 243(6):547–553
211. Bou-Haidar P, Peduto AJ, Karunaratne N (2009) Dif-
ferential diagnosis of T2 hyperintense spinal cord
lesions: part B. J Med Imaging Radiat Oncol 53
(2):152–159. doi:10.1111/j.1754-9485.2009.02067.x
212. Chason JL, Walker FB, Landers JW (1963) Metastatic
carcinoma in the central nervous system and dorsal
root ganglia. A prospective autopsy study. Cancer
16:781–787
213. Costigan DA, Winkelman MD (1985) Intramedullary
spinal cord metastasis. A clinicopathological study of
13 cases. J Neurosurg 62(2):227–233. doi:10.3171/
jns.1985.62.2.0227
214. Kalayci M, Cağavi F, G€ ul S, Yenid€unya S, Açikgöz B
(2004) Intramedullary spinal cord metastases: diag-
nosis and treatment – an illustrated review. Acta
S.R. Boddu et al.
Neurochir (Wien) 146(12):1347–1354. doi:10.1007/
s00701-004-0386-1; discussion 1354
215. Watanabe M, Nomura T, Toh E, Sato M, Mochida J
(2006) Intramedullary spinal cord metastasis: a clini-
cal and imaging study of seven patients. J Spinal
Disord Tech 19(1):43–47. doi:10.1097/01.bsd.
0000188661.08342.2a
216. Mut M, Schiff D, Shaffrey ME (2005) Metastasis to
nervous system: spinal epidural and intramedullary
metastases. J Neurooncol 75(1):43–56. doi:10.1007/
s11060-004-8097-2
217. Haq A, Wasay M (2006) Magnetic resonance imaging
in poliomyelitis. Arch Neurol 63(5):778.
doi:10.1001/archneur.63.5.778
218. Ferraz-Filho JRL, dos Santos TU, de Oliveira EP,
Souza AS (2010) MRI findings in an infant with
vaccine-associated paralytic poliomyelitis. Pediatr
Radiol 40(Suppl 1):S138–S140. doi:10.1007/
s00247-010-1650-6
219. Hui AC, Wong KS, Fu M, Kay R (2000) Ischaemic
myelopathy presenting as Guillain-Barré syndrome.
Int J Clin Pract 54(5):340–341
220. Hemmer B, Glocker FX, Schumacher M, Deuschl G,
L€ucking CH (1998) Subacute combined degenera-
tion: clinical, electrophysiological, and magnetic res-
onance imaging findings. J Neurol Neurosurg
Psychiatry 65(6):822–827
221. Thurnher MM, Post MJ, Jinkins JR (2000) MRI of
infections and neoplasms of the spine and spinal cord
in 55 patients with AIDS. Neuroradiology
42(8):551–563
222. McArthur JC, Brew BJ, Nath A (2005)
Neurological complications of HIV infection. Lancet
Neurol 4(9):543–555. doi:10.1016/S1474-4422(05)
70165-4
223. Fitzgerald RH Jr, Marks RD Jr, Wallace KM (1982)
Chronic radiation myelitis. Radiology 144(3):
609–612. doi:10.1148/radiology.144.3.6808557
224. SreeHarsha CK, Rajasekaran S, Dhanasekararaja P
(2006) Spontaneous complete recovery of paraplegia
caused by epidural hematoma complicating epidural
anesthesia: a case report and review of literature.
Spinal Cord 44(8):514–517. doi:10.1038/sj.sc.
3101869
225. Gobin YP (1997) Classification and endovascular
treatment of spinal cord arteriovenous malformations
and fistulas. J Stroke Cerebrovasc Dis 6(4):282–286
226. Yamada N, Okita Y, Minatoya K et al (2000) Preop-
erative demonstration of the Adamkiewicz artery by
magnetic resonance angiography in patients with
descending or thoracoabdominal aortic aneurysms.
Eur J Cardiothorac Surg 18(1):104–111.
doi:10.1016/S1010-7940(00)00412-7
227. Shi HB, Suh DC, Lee HK et al (1999) Preoperative
transarterial embolization of spinal tumor: emboliza-
tion techniques and results. AJNR Am J Neuroradiol
20(10):2009–2015. Available at http://www.ajnr.org/
content/20/10/2009. Accessed 10 Mar 2014
47 Spinal Cord Infarction and Differential Diagnosis
228. Backes WH, Nijenhuis RJ (2008) Advances in spinal
cord MR angiography. AJNR Am J Neuroradiol
29(4):619–631. doi:10.3174/ajnr.A0910
229. Fereshetian A, Kadir S, Kaufman SL et al (1989)
Digital subtraction spinal cord angiography in
patients undergoing thoracic aneurysm surgery.
Cardiovasc Intervent Radiol 12(1):7–9
230. Kieffer E, Richard T, Chiras J, Godet G, Cormier E
(1989) Preoperative spinal cord arteriography in
aneurysmal disease of the descending thoracic and
thoracoabdominal aorta: preliminary results in
45 patients. Ann Vasc Surg 3(1):34–46
231. Kieffer E, Fukui S, Chiras J, Koskas F, Bahnini A,
Cormier E (2002) Spinal cord arteriography: a safe
adjunct before descending thoracic or
thoracoabdominal aortic aneurysmectomy. J Vasc
Surg 35(2):262–268
232. Williams GM, Roseborough GS, Webb TH, Perler
BA, Krosnick T (2004) Preoperative selective inter-
costal angiography in patients undergoing
thoracoabdominal aneurysm repair. J Vasc Surg
39(2):314–321. doi:10.1016/j.jvs.2003.09.039
233. Prestigiacomo CJ, Niimi Y, Setton A, Berenstein A
(2003) Three-dimensional rotational spinal angiogra-
phy in the evaluation and treatment of vascular
malformations. AJNR Am J Neuroradiol 24(7):
1429–1435. Available at http://www.ajnr.org/con
tent/24/7/1429. Accessed 8 Mar 2014
234. Ropper AE, Lin N, Gross BA et al (2012) Rotational
angiography for diagnosis and surgical planning in
the management of spinal vascular lesions. Neurosurg
Focus 32(5):E6. doi:10.3171/2012.1.FOCUS11254
235. Matsumoto M, Kodama N, Endo Y et al (2007)
Dynamic 3D-CT angiography. AJNR Am J
Neuroradiol 28(2):299–304
236. Takase K, Sawamura Y, Igarashi K et al (2002) Dem-
onstration of the artery of Adamkiewicz at multi-
detector row helical CT. Radiology 223(1):39–45.
doi:10.1148/radiol.2231010513
237. Lai P-H, Pan H-B, Yang C-F et al (2005) Multi-
detector row computed tomography angiography in
diagnosing spinal dural arteriovenous fistula: initial
experience. Stroke 36(7):1562–1564. doi:10.1161/
01.STR.0000170048.94718.07
238. Lai PH, Weng MJ, Lee KW, Pan HB (2006)
Multidetector CT angiography in diagnosing type I
and type IVA spinal vascular malformations. AJNR
Am J Neuroradiol 27(4):813–817
239. Si-jia G, Meng-wei Z, Xi-ping L et al (2009) The
clinical application studies of CT spinal angiography
with 64-detector row spiral CT in diagnosing spinal
vascular malformations. Eur J Radiol 71(1):22–28.
doi:10.1016/j.ejrad.2008.04.005
240. Farb RI, Kim JK, Willinsky RA et al (2002) Spinal
dural arteriovenous fistula localization with a tech-
nique of first-pass gadolinium-enhanced MR angiog-
raphy: initial experience. Radiology 222(3):843–850.
doi:10.1148/radiol.2223010826
1181
241. Pattany PM, Saraf-Lavi E, Bowen BC (2003) MR
angiography of the spine and spinal cord. Top Magn
Reson Imaging 14(6):444–460
242. Bowen BC, DePrima S, Pattany PM, Marcillo A,
Madsen P, Quencer RM (1996) MR angiography of
normal intradural vessels of the thoracolumbar spine.
AJNR Am J Neuroradiol 17(3):483–494
243. Nijenhuis RJ, Jacobs MJ, Jaspers K et al (2007) Com-
parison of magnetic resonance with computed tomog-
raphy angiography for preoperative localization of the
Adamkiewicz artery in thoracoabdominal aortic aneu-
rysm patients. J Vasc Surg 45(4):677–685.
doi:10.1016/j.jvs.2006.11.046
244. Jaspers K, Nijenhuis RJ, Backes WH (2007) Differ-
entiation of spinal cord arteries and veins by time-
resolved MR angiography. J Magn Reson Imaging
26(1):31–40. doi:10.1002/jmri.20940
245. Ali S, Cashen TA, Carroll TJ et al (2007) Time-
resolved spinal MR angiography: initial clinical expe-
rience in the evaluation of spinal arteriovenous
shunts. AJNR Am J Neuroradiol 28(9):1806–1810.
doi:10.3174/ajnr.A0639
246. Saindane AM, Boddu SR, Tong FC, Dehkharghani S,
Dion JE (2014) Contrast-enhanced time-resolved
MRA for pre-angiographic evaluation of suspected
spinal dural arterial venous fistulas. J Neurointerv
Surg 7:135–140. doi:10.1136/neurintsurg-2013-
010981
247. Kaufmann TJ, Kallmes DF (2008) Diagnostic cere-
bral angiography: archaic and complication-prone
or here to stay for another 80 years? AJR Am J
Roentgenol 190(6):1435–1437. doi:10.2214/AJR.
07.3522
248. Mull M, Nijenhuis RJ, Backes WH, Krings T,
Wilmink JT, Thron A (2007) Value and limitations
of contrast-enhanced MR angiography in spinal arte-
riovenous malformations and dural arteriovenous fis-
tulas. AJNR Am J Neuroradiol 28(7):1249–1258.
doi:10.3174/ajnr.A0612
249. Luetmer PH, Lane JI, Gilbertson JR, Bernstein MA,
Huston J 3rd, Atkinson JLD (2005) Preangiographic
evaluation of spinal dural arteriovenous fistulas with
elliptic centric contrast-enhanced MR angiography and
effect on radiation dose and volume of iodinated con-
trast material. AJNR Am J Neuroradiol 26(4):711–718
250. Eddleman CS, Jeong H, Cashen TA et al (2009)
Advanced noninvasive imaging of spinal vascular
malformations. Neurosurg Focus 26(1):E9.
doi:10.3171/FOC.2009.26.1.E9
251. Di Chiro G, Wener L (1973) Angiography of the
spinal cord. A review of contemporary techniques
and applications. J Neurosurg 39(1):1–29.
doi:10.3171/jns.1973.39.1.0001
252. Kaufmann TJ, Huston J 3rd, Mandrekar JN, Schleck
CD, Thielen KR, Kallmes DF (2007) Complications
of diagnostic cerebral angiography: evaluation of
19,826 consecutive patients. Radiology 243(3):
812–819. doi:10.1148/radiol.2433060536
1182
253. Chen J, Gailloud P (2011) Safety of spinal angiogra-
phy: complication rate analysis in 302 diagnostic
angiograms. Neurology 77(13):1235–1240.
doi:10.1212/WNL.0b013e3182302068
254. Restrepo L, Wityk RJ, Grega MA et al (2002) Diffu-
sion- and perfusion-weighted magnetic resonance
imaging of the brain before and after coronary artery
bypass grafting surgery. Stroke 33(12):2909–2915.
Available at http://www.ncbi.nlm.nih.gov/pubmed/
12468790. Accessed 26 Apr 2012
255. Baba H, Tomita K, Kawagishi T, Imura S (1993)
Anterior spinal artery syndrome. Int Orthop
17(6):353–356
256. Jauch EC, Saver JL, Adams HP Jr et al (2013) Guide-
lines for the early management of patients with acute
ischemic stroke: a guideline for healthcare profes-
sionals from the American Heart Association/Ameri-
can Stroke Association. Stroke 44(3):870–947.
doi:10.1161/STR.0b013e318284056a
257. McGarvey ML, Cheung AT, Szeto W, Messe SR
(2007) Management of neurologic complications of
thoracic aortic surgery. J Clin Neurophysiol 24
(4):336–343. doi:10.1097/WNP.0b013e31811ec0b0
258. Cheung AT, Weiss SJ, McGarvey ML et al (2002)
Interventions for reversing delayed-onset postopera-
tive paraplegia after thoracic aortic reconstruction.
Ann Thorac Surg 74(2):413–419; discussion 420-421
259. Wallace MC, Tator CH, Frazee P (1986) Relationship
between posttraumatic ischemia and hemorrhage in
the injured rat spinal cord as shown by colloidal
carbon angiography. Neurosurgery 18(4):433–439
260. Kirsch M, Berg-Dammer E, Musahl C, Bäzner H,
K€uhne D, Henkes H (2013) Endovascular manage-
ment of spinal dural arteriovenous fistulas in
78 patients. Neuroradiology 55(3):337–343.
doi:10.1007/s00234-013-1134-0
261. Anderson DK, Means ED, Waters TR, Green ES
(1982) Microvascular perfusion and metabolism in
injured spinal cord after methylprednisolone treat-
ment. J Neurosurg 56(1):106–113. doi:10.3171/
jns.1982.56.1.0106
262. Anderson DK, Saunders RD, Demediuk P et al (1985)
Lipid hydrolysis and peroxidation in injured spinal
cord: partial protection with methylprednisolone or
vitamin E and selenium. Cent Nerv Syst Trauma
2(4):257–267
263. Hall ED, Braughler JM (1982) Glucocorticoid mech-
anisms in acute spinal cord injury: a review and ther-
apeutic rationale. Surg Neurol 18(5):320–327
264. Fehlings MG, Tator CH (1999) An evidence-based
review of decompressive surgery in acute spinal cord
injury: rationale, indications, and timing based on
experimental and clinical studies. J Neurosurg
91(1 Suppl):1–11
265. Fehlings MG, Perrin RG (2005) The role and timing
of early decompression for cervical spinal cord injury:
update with a review of recent clinical evidence.
S.R. Boddu et al.
Injury 36(Suppl 2):B13–B26. doi:10.1016/j.
injury.2005.06.011
266. La Rosa G, Conti A, Cardali S, Cacciola F, Tomasello
F (2004) Does early decompression improve neuro-
logical outcome of spinal cord injured patients?
Appraisal of the literature using a meta-analytical
approach. Spinal Cord 42(9):503–512. doi:10.1038/
sj.sc.3101627
267. Fehlings MG, Vaccaro A, Wilson JR et al (2012)
Early versus delayed decompression for traumatic
cervical spinal cord injury: results of the Surgical
Timing in Acute Spinal Cord Injury Study
(STASCIS). PLoS One 7(2):e32037. doi:10.1371/
journal.pone.0032037
268. Van Middendorp JJ, Hosman AJF, Doi SAR (2013)
The effects of the timing of spinal surgery after trau-
matic spinal cord injury: a systematic review and
meta-analysis. J Neurotrauma 30(21):1781–1794.
doi:10.1089/neu.2013.2932
269. DeVivo MJ, Kartus PL, Stover SL, Rutt RD, Fine PR
(1989) Cause of death for patients with spinal cord
injuries. Arch Intern Med 149(8):1761–1766
270. Young JS (1979) Spinal cord injury: associated gen-
eral trauma and medical complications. Adv Neurol
22:255–260
271. Patsalides A, Knopman J, Santillan A, Tsiouris AJ,
Riina H, Gobin YP (2011) Endovascular treatment of
spinal arteriovenous lesions: beyond the dural fistula.
AJNR Am J Neuroradiol 32(5):798–808.
doi:10.3174/ajnr.A2190
272. Da Costa L, Dehdashti AR, terBrugge KG (2009)
Spinal cord vascular shunts: spinal cord vascular
malformations and dural arteriovenous fistulas.
Neurosurg Focus 26(1):E6. doi:10.3171/
FOC.2009.26.1.E6
Suggested Reading
Krings T, Geibprasert S (2009) Spinal dural arteriovenous
fistulas. AJNR Am J Neuroradiol 30(4):639–648
Krings T, Lasjaunias PL, Hans FJ et al (2007) Imaging in
spinal vascular disease. Neuroimaging Clin N Am
17(1):57–72
Krishna V, Lazaridis C, Ellegala D et al (2012) Spinal cord
infarction associated with subarachnoid hemorrhage.
Clin Neurol Neurosurg 114(7):1030–1032
Lasjaunias P, Berenstein A, TerBrugge K (2001) Clinical
vascular anatomy and variations. In: Surgical
neuroangiography, vol 1. Springer, Berlin
Nair S, Gobin YP, Leng LZ et al (2013) Preoperative
embolization of hypervascular thoracic, lumbar, and
sacral spinal column tumors: technique and outcomes
from a single center. Interv Neuroradiol 19(3):377–385
Patsalides A, Knopman J, Santillan A, Tsiouris AJ,
Riina H, Gobin YP (2011) Endovascular treatment of
47 Spinal Cord Infarction and Differential Diagnosis
spinal arteriovenous lesions: beyond the dural fistula.
AJNR Am J Neuroradiol 32(5):798–808
Saindane AM, Boddu SR, Tong FC, Dehkharghani S, Dion
JE (2014) Contrast-enhanced time-resolved MRA for
pre-angiographic evaluation of suspected spinal dural
arterial venous fistulas. J Neurointerv Surg 7:135–140
Santillan A, Nacarino V, Greenberg E, Riina HA, Gobin
YP, Patsalides A (2012) Vascular anatomy of the spinal
cord. J Neurointerv Surg 4(1):67–74
1183
Setacci F, Sirignano P, De Donato G et al (2010)
Endovascular thoracic aortic repair and risk of spinal
cord ischemia: the role of previous or concomitant
treatment for aortic aneurysm. J Cardiovasc Surg
(Torino) 51(2):169–176
Thurnher MM, Cartes-Zumelzu F, Mueller-Mang C (2007)
Demyelinating and infectious diseases of the spinal
cord. Neuroimaging Clin N Am 17(1):37–55
 Index

A Arteriotomy, 197–198
Acetazolamide, 1035 Arteriovenous malformations, 79, 262, 264, 1028, 1081,
Acetylsalicylic acid, 1027 1086
ACTA2, 1027 Arteriovenous malformations (AVMs), 79, 262–264, 397
Acute disseminated encephalomyelitis (ADEM), 1152–1153 Artery of Adamkiewicz (AKA), 1129–1130
Acute hemiparesis, 1011 Artery-to-artery embolism, 1015
Acute polyneuropathy, 1158 Artifacts, 586, 588
AIDS associated myelopathy, 1159 Ascending pharyngeal artery, 91
Alagille, 1027 Asymptomatic Carotid Atherosclerosis Study (ACAS), 192
Amyloid angiopathy, 396 Asymptomatic ICAS, 206
Anastomosis, 294, 296–298, 300 Atherosclerosis, 96–97, 171–172, 174
Ancillary test, 867 AVM, 991, 1029
Anesthetic management, 540 Axial fat-suppressed, 1025
Aneurysms, 199–200, 393–394, 397, 522
definition of, 552
iron age, 554–563 B
modern era, 571–577 Balloon remodeling, 554–563, 579
renaissance, 563–568 Basilar artery fenestration, 165–166
stone age, 553 Basilar artery strokes, 1016
Angiogenesis, 304, 306–307, 316, 318 Bell palsy, 1019
Angiography, 294–295, 297, 299, 1089 Bernoulli’s principle, 52–53
CT, 623–625 Black-blood, 100–101, 130
digital subtraction, 627–629 Bleeding disorders, 1028
super-selective angiography, 631 Blister aneurysm, 522–531
Angio-MR, 1079 Blister-like aneurysm, 552, 574
Anoxia, 332, 356–359 Blood blister aneurysm, 522
Anterior cerebral artery variations, 25 Blood vessels, 304–307
Anterior spinal artery (ASA), 1128, 1129, 1134 Bone infarction, 443, 455–456
Anticoagulant therapy, 1035 Bow Hunter’s syndrome, 184–188
Anticoagulation, 1027 Brain arteriovenous malformation (bAVM)
Antiplatelet therapy, 1027 angioarchitecture, 629
Antithrombotic therapy, 1011, 1013 arterial aneurysms, 630–631
Aortic arch, 88 computed tomography, 623–625
Apparent diffusion coefficient (ADC), 942, 945, 1014, definition, 605–606
1030, 1146, 1150 digital subtraction angiography, 627–629
Arachnoid granulations, 950 embolization, 633–634
Arrhythmia, 1025 embryology, 612–614
Arterial dissection, 1024–1025 epidemiology, 611–612
Arterial ischemic stroke, 1010 hemorrhage risk, 608–611
Arterial spin labeling (ASL), 460, 848–849 magnetic resonance imaging, 625–627
Arterial tortuosity syndrome, 147 microsurgical resection, 633
Arterial vasospasm, 510 natural history, 606
Arterials ischemic stroke, 1030 nidal size, 631
Arteriopathies, 1014, 1016, 1024 pathophysiology, 614–622

# Springer Science+Business Media New York 2016 1185

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6
1186 Index

Brain arteriovenous malformation (bAVM) (cont.) Cerebral sinovenous thrombosis (CSVT), 1010, 1030
radiotherapy, 634–635 Cerebral vascular malformations, 720
venous drainage, 631 Cerebral vasculitis, 698, 700, 703
Brain Cerebral venous and sinus thrombosis (CVST), 936
amphibians, 9 Cerebral venous system, 956–957
birds, 10 Cerebrovascular disease, 223, 227
death, 58, 866–892 Cerebrovascular reactivity, 1022
fishes, 6, 8 Cerebrovascular reserve (CVR), 1022
hemorrhage, 552 Childhood arterial ischemic stroke (CAIS), 1010
ischemia, 470–471 Choroidal stage, 15–16
mammals, 10–12 Chronic cerebrospinal venous insufficiency (CCSVI),
reptiles, 9 954–955, 962–965, 968, 971–974
retraction, 541 Churg-Strauss syndrome, 769
tumor, 1029 Circle of Willis, 48
Brainstem infarction, 1016 Clip inspection, 543
Bypass surgery, 1023 Clipping, 506, 513, 553, 556, 563, 574
Clip-wrapping, 526–527
Clot retrievers, 427–428
C Coagulopathy, 241
Capillary malformations, 988 Cognitive decline, 1024
Capillary telangiectasias, 742 Cognitive dysfunction, 1036
Cardiac catheterization, 1025 Cognitive outcome, 1036
Cardiac surgery, 1025 Coil(ing), 504, 554, 571
Cardiomyopathy, 1025 Coil embolization, 584–586, 591
Carotid aneurysms, 154–155 COL4A1, 1027
Carotid artery, 97–98 Collateral circulation, 208, 215, 294, 300
stenting, 108 Collateralization, 1023
Carotid artery dissection (CAD), 116–135 Color Doppler ultrasound, 1033
Carotid artery surgery Common carotid artery, 88–89, 93
aneurysms, 199–200 Complications, vasospasm and post-treatment, 589–599
carotid body tumours, 198–199 Computed tomography (CT), 332, 353, 920, 1011
pre-operative study, 192–195 Computed tomography angiography (CTA),
technique of, 196–198 391, 585–586, 589–590, 599, 623, 948,
Carotid body tumours, 198–199 Congenital heart disease, 1019, 1025
Carotid cavernous fistula (CCF), 658–680 Congenital neurovascular syndromes
Carotid endarterectomy, 98, 106, 108 AVM/AVF syndromes, 1047–1051
Carotid stenosis, 143 cavernous malformations, 1055, 1057
Carotid-vertebrobasilar variant, 165–167 MELAS, 1057, 1059
Carotidynia, 156–157 moyamoya disease, 1042–1043
Cavernoma, 720 PHACE syndrome, 1043–1047
Cavernous hemangiomas, 720 Sturge Weber syndrome, 1051–1053
Cavernous malformations (CMs), 266–267, 994–996 venous malformations, 1054–1055
Central nervous system vasculitis, 285, 693 Connective tissue disorders, 147, 150, 1025
Cerebral amyloid angiopathy, 244–248 Continuous EEG, 1027
Cerebral aneurysm, 583–586, 590 Continuous flushing technique, 69
Cerebral arteriopathies, 1015 Contrast-enhanced CT, 947
Cerebral autosomal dominant arteriopathy with subcortical Contrast-enhanced venous MRA, 944
infarcts and leukoencephalopathy (CADASIL), Conventional angiography (CA), 1015, 1023
762–763 Conversion disorders, 1019
Cerebral autosomal recessive arteriopathy with subcortical Cord sign, 945, 1035
infarcts and leukoencephalopathy (CARASIL), Corkscrew vessels, 939
763–764 Corpus callosum, 1036
Cerebral fat embolism, 449 Cortical venous drainage, 647
Cerebral hemodynamics, 223 Cortical venous reflux, 665
Cerebral infarction, 694, 698 Cortical venous thrombosis (CVT), 937
Cerebral, ischemia, 710 Cortical-subcortical hypoattenuating, 1012
Cerebral microbleeds, 251 Cranial ultrasound (CUS), 1012, 1030
Cerebral microhemorrhages, 247 Craniocervical artery dissection (CCAD), 44–45
Index 1187

Crescent-shaped rim, 1025 E

CREST trial, 192 Eagle syndrome, 157–158
Crossed cerebellar diaschisis, 800 ECASS, 1017
Cryptogenic stroke, 108–109 Echo-planar imaging (EPI), 1146
CT angiography. See Computed tomography angiography EFNS-guidelines, 795
(CTA) Ehler Danlos syndrome, 147
CT venography (CTV), 1032, 1035 Embolectomy, 432–433
CTA. See Computed tomography angiography (CTA) Emboli, 331–332, 357
Cytotoxic edema, 945 Embolization, 633–634, 991
iatrogenic, 1138
spinal angiography and, 1138
D transarterial, 650
D-dimer, 938 transvenous, 648, 650
Decompressive surgery, 1018 Embryonic Carotid-Basilar system anastomoses, 90
Deep cerebral veins, 938 Empty delta sign, 947, 1035
Deep cerebral venous thrombosis (DVT), 937 End diastolic velocity (EDV), 193
Dehydration, 1032 Endarterectomy, 196–198
Delivery, 857 Endocarditis, 1025
Dense vein sign, 945 Endovascular embolization, 671, 673, 1106, 1113
Descending corticospinal tract, 1036 Endovascular procedures, 1028
Detachable balloon, 553–554, 579 Endovascular techniques, 584–586
Developmental venous anomaly (DVA), 733–736 Endovascular therapy, 530–531
Diagnostic delay, 1011 Endovascular treatment, 426–430, 432
Diffusion, 1079 Epidural hematomas, 405–407
Diffusion tensor imaging (DTI), 457, 459, 1036, European Carotid Surgery Trial (ECST), 192–193
1084–1085, 1149 Evidence based medicine, 416–418
Diffusion weighted imaging (DWI), 942, 944, 1013–1014, External carotid artery (ECA), 89–92, 294–295,
1030, 1127 297, 299
spinal cord, 1146–1150 External-internal carotid arteries anastomoses, 92
Diffusion-weighted MRI, 870 Extracorporeal membrane oxygenation (ECMO), 1025
Diffusion-weighted sequences, 945 Extracranial internal carotid artery, 44
Digital subtraction angiography (DSA), 122, 585, Extra-cranial venous abnormalities, 954–955, 960, 962,
588, 939–940, 1127, 1158, 1161 971, 973–974
Digital subtraction cervico-cerebral angiography (DSA)
catheter and wire, 69–70
patient preparation, 68 F
preprocedural assessment, 68 Fabry’s disease, 752–753
procedure, 68–69 Facial artery, 91
projections, 70–83 Fast spin-echo (FSE), 1145
vascular access, 69 Fat supressed T1 sequence, 124
Diplopia, 1032 Fetal distress, 1031
Direct bypass, 852 Fetal posterior cerebral artery, 18
Direct signs, 939 Fiber tractography (FT), 1149
Dissection, 1015–1016 Fibromuscular dysplasia (FMD), 42–43
Dominant anterior choroidal artery, 19 angiography, 141–142
Doppler effect, 49, 50, 53 natural history and clinical features, 140–141
Doppler ultrasound, 1033 types, 141
Double lumen, 1025 Flame shape stenosis, 1025
Down’s syndrome, 1021 Flow, blood, 306–307, 309, 311
Duplications, 24 Flow disruptors, 579
Dural arteriovenous fistulas (DAVFs), 264, 397 Flow-diverters (FDs), 529, 571–579
Dural sinus thrombosis, 937 Fluid-attenuated inverted recovery (FLAIR), 1014
Dynamic Contrast Enhanced Computed Tomography Focal cerebral arteriopathy (FCA), 1020
(DCE-CT), 311 Fractional flow, 219, 227
Dynamic relaxivity contrast enhanced MRI (DRCE-MRI), Frontotemporal craniotomy, 544
310–311 Functional MRI (fMRI), 871
Dynamic susceptibility contrast enhanced MRI (DSCE- Fusiform, 570, 574
MRI), 308–310 Fusion patterns variations, basilar artery, 20–24
1188 Index

G spinal cord, 1127

Gamma knife surgery (GKS), 634–635 spinal venous, 1138–1139, 1142
GDCs, 555, 571 thromboembolic, 1136
Giant cell arteritis (GCA), 41–42, 144 venous, 1133
Global brain hypoxia, 471, 486, 492 Infection, 117
Glossopharyngeal neuralgia, 786–787 Inflammatory, 1015
Glue, 650 Integrin, 319–321
Gradient-echo (GRE), 1145, 1165 Internal carotid artery (ICA), 89, 294–295, 299
Guidelines, for tPA administration, 417 occlusion, 58–60
Internal maxillary artery, 92
Interventional treatment, 940
H Intimal fibrodysplasia, 1024
Headache, 1032 Intimal flap, 1025
Hemicraniectomy, 1018 Intra-arterial thrombolysis, 940
Hemifacial spasm (HFS), 784–786 Intracardiac shunt, 63
Hemiparesis, 1030 Intracranial aneurysms, 60, 74, 258, 522–523,
Hemiplegic cerebral palsy, 1030–1031 525, 527–528, 531, 536
Hemoglobin, 385, 388 bone resection, 541
Hemolytic anemia, 441 brain retraction, 541
Hemorrhagic neoplasm, 276–281 complication, 512–513, 546
Hemorrhagic stroke, 445, 447, 1011, 1028, 1030 conservative management, 508
clinical presentation, 385 clip inspection, 543
contrast enhanced MRA, 393–394 craniotomy selection, 545
magnetic resonance imaging, 387, 390 endovascular treatment, 504, 506
non-contrast computed tomography, 385–386 epidemiology, 498
vascular malformations, 397 follow-up procedure, 547
Hemorrhagic transformation, 1012 imaging, 501–503
of infarct, 270–274 permanent clipping, 542
Herniation, 1029 postoperative care, 545–546
High resolution magnetic resonance imaging (HR-MRI), prognostic factors, 513
129 surgical clipping, 536
HII. See Hypoxic ischemic injury (HII) surgical treatment, 506–508
Homocystinuria, 1025, 1027 subarachnoid dissection, 541
Hydrocephalus, 508–510, 1029 surgical clipping, 537
Hyperdense artery sign, 1012 symptoms, 499–500
Hyperdense dural sinus sign, 1035 temporary clipping, 542–543
Hypertension, 395 vascular control, 542
Hypertensive hemorrhage, 234–238 Intracranial arterial stenosis, 219, 224
Hypoperfusion, 1016 Intracranial atherosclerosis (ICAS), 206–207, 209
Hypoxia, 304, 306–307, 316, 332, 356–359 Intra-cranial dissection, 1025
Hypoxic ischemic injury (HII), 470–472, 476, 478–486, Intracranial dural fistula
488–492 classification, 643–645
Hypoxic-ischemic encephalopathy (HIE), 1030 imaging, 647–648
natural history, 646–647
physiopathology, 642–643
I symptoms, 646
Idiopathic transverse myelitis (ITM), 1151–1152 treatment, 648–655
Incidence, 1010–1011 Intracranial hemorrhage, 608, 611, 665–666, 828, 855,
Indirect bypass, 852–853 1012, 1035
Indirect signs, 939 Intracranial hypertension, 1035
Indocyanine green (ICG) fluorescent videoangiography, Intracranial pressure, 1032
543 Intracranial revascularization, 433
Infarct(s), 370–373, 376, 379 Intramural hematoma, 1025
Infarction Intraoperative digital subtraction angiography (DSA), 543
hemorrhagic spinal cord, 1138 Intravenous fibrinolysis (IVF), 425, 430–431
hemorrhagic, 1133 Ischemia, 331, 345, 1075–1077, 1079
ischemic spinal cord, 1132–1136, 1143 Ischemic stroke, 96, 98, 108, 206–207, 425–429, 445
ischemic, 1133 Ivy sign, 1022
Index 1189

J MRA. See Magnetic resonance angiography (MRA)

Japan Adult Moyamoya (JAM) Trial, 857 MRI with venography (MRV), 1032
Jugular vein, 900 Multi-contrast MRI, 102
Jugular vein reflux, 954 Multi-detector computed tomography-angiography
(MDCTA), 195
Multiple sclerosis (MS), 955, 962–966, 969, 971,
J 974, 1150–1151
Kawasaki disease, 769 Mural hematoma, 116, 118
Myelitis
acute, 1127
L transverse, 1146, 1151
Labyrinthine hemorrhage (LH), 456 Myelopathy, 1112, 1119
Labyrinthitis ossificans (LO), 456 Myxoma, 1025
Lactic acidosis, 1027
Lacunar infarct, 330, 355–356
Lindegaard ratio, 56 N
Lingual artery, 91 Neck veins, 957, 966, 970
Loss of gray matter–white matter differentiation, 1012 Neoplasms, 1028
Lumbo-peritoneal shunt, 1035 Neural tube/spinal cord, 1095
Lymphatic malformations, 988–991 Neuro-Behcet’s disease, 775, 1158
Neurofibromatosis, 150–151
Neurofibromatosis type1, 1021, 1027
M Neuroimaging, 208
Magnetic resonance angiography (MRA), 391–392, Neuromyelitis optica, 1158
394, 585–586, 588–589, 595, 625, 831, 834, Neuroprotection, 1027, 1035
844, 881, 940 Neuroprotective therapy, 1011, 1013
Magnetic resonance imaging (MRI), 96, 103, 337, 355, Neurosarcoidosis, 1151, 1158
831, 836–837, 923 Neurovasculitides, 684–687, 695, 698, 700, 705, 707, 714
Malignant infarct, 1012 Non-accidental injury, 1028
Marfan syndrome (MS), 150 Non-accidental trauma, 1025
Mastoiditis, 1032, 1035 Non-atherosclerotic vascular disease
Mechanical index (MI), 64 craniocervical artery dissection, 44–45
Mechanical recanalization, 940 extracranial internal carotid artery, 44
Meningitis, 1035 fibromuscular dysplasia, 42–43
Methemoglobin, 389, 393 giant cell arteritis, 41–42
Micro-Doppler ultrasonography, 543 moyamoya, 43–44
Microsurgery, 633 Takayasu arteritis, 40–41
Microsurgical clipping, 528, 584 Non-contrast T1-weighted images, 1025
Middle cerebral artery, 1015 Non-convulsive seizures, 1027
Migraine, 1019 Non-progressive childhood primary angiitis of the central
characterization, 792 nervous system, 1020
diagnostic criteria without and with aura, 792 Non-traumatic intracranial hemorrhage
MRI scans, 811 cerebral amyloid angiopathy, 244, 248
neuroimaging indications, 794–795 coagulopathy, 241
pathophysiology, 793–794 CT and MRI, 239
and stroke, 796, 807 GRE T2* images, 248
T2 hyperintensities, 807–811 hemorrhagic neoplasm, 276–281
Migrainous infarction, 796–797, 803 hemorrhagic transformation of infarct, 270–274
Mitochondrial encephalopathy, 1027 hypertensive hemorrhage, 234–238
Mitochondrial myopathy, Encephalopathy, Lactic nontraumatic subarachnoid hemorrhage, 251–262
Acidosis, and Stroke-like symptoms (MELAS), radiation-induced disease, 283
1057–1059 susceptibility weighted imaging, 248–251
Moyamoya, 1011, 1015, 1024, 1027 vascular malformations, 262–270
Moyamoya disease (MMD), 758, 819, 858, 1042–1043 vasculitis, 285
Moyamoya syndrome, 43–44, 449–451 venous thrombosis, 274–276
MR angiography. See Magnetic resonance angiography Non-traumatic subarachnoid hemorrhage, 251–262
(MRA), North American Symptomatic Carotid Endarterectomy
MR spectroscopy (MRS), 460, 871 Trial (NASCET), 192–193
1190 Index

O Propagation, 1035
Occipital artery, 91 Prothrombotic states, 1032
Onyx, 650 Pseudoaneurysm, 1025
Ophthalmic artery, 49, 60 Pseudonormalization, 1030
Optic nerve sheath fenestration, 1035 Pterional craniotomy, 545
Optic radiation, 1036 Puff of smoke, 1021
Osteomyelitis, 456
Otitis media, 1032
Q
Quantitative assessment, 947
P Quantitative MRI (qMRI), 460–461
Pacchionian granulations, 950
Papilledema, 1032, 1035
Parainfectious, 1015 R
Paraplegia, 1108 Radiation, 1013, 1015
Parenchymal changes, 939 microhemorrhage, 283
Parenchymal hemorrhage, 939 myelitis, 1159
Parenchymal infarcts, 1033 vasculopathy, 283
Partially-thrombosed aneurysm, 570, 574 Radiation-induced microangiopathy, 287
Patent foramen ovale, 63 Radicular artery, 1128, 1130–1131, 1143, 1149
Peak systolic velocity (PSV), 193 Recurrence, 121, 207, 216
Pediatric ASPECT score, 1018, 1036 Rehabilitation, 1027
Pediatric ICH score, 1030 Resistive index (RI), 53
Pediatric NIHSS, 1018 Revascularization, 1015
Pediatric stroke, 1057, 1059 Reversible cerebral vasoconstriction syndrome (RCVS), 756
Perfusion, 845, 847 Risk factors, 1011, 1019
imaging, 223, 459–460 Risk stratification, 207–208, 229
scintigraphy, 889 RNF213, 1027
Perinatal hemorrhagic stroke, 1031 Rupture, 499, 504
Perinatal period, 1030
Perinatal strokes, 1030
Periventricular leukomalacia, 1031 S
Periventricular venous infarction, 1030 Sclerotherapy, 987
Permanent clipping, 543 Segmental artery, 1128
Permeability, endothelial membrane, 306, 308–310 anastomose, 1096
Persistent trigeminal artery, 19 Seizures, 1011, 1019
PFO, 1026 Serial lumbar punctures, 1035
PHACE syndrome, 1027, 1043, 1047 Shamblin’s classification, 198
Phase-contrast venous MRA, 944 Sickle cell disease (SCD), 57–58, 440, 463, 1011, 1021
Plaque imaging, 205, 229 Signal-to-noise ratio (SNR), 1145, 1147
Plaque rupture, 98, 107 Silent cerebral infarct (SCI), 447, 449
Polyarteritis nodosa (PAN), 768 Single photon emission computed tomography (SPECT),
Polycythemia, 1035 461, 847–848, 852
Positron emission tomography (PET), 308, 316, 318–319, Sinus pericranii (SP), 744, 993–994
461, 847 Sinus thrombosis, 937
Posterior auricular artery, 92 Sloan ratio, 56
Posterior circulation, 169, 180–181 Spastic cerebral palsy, 1030
Posterior circulation stroke, 1015 Spinal angiogram, 1105
Posterior Reversible Encephalopathy Syndrome (PRES), Spinal cord, 1075, 1079, 1081, 1086
454, 753 vascular neurophylogeny, 3–4
Posterior spinal artery (PSA), 1129, 1141, 1144 Spine, 1068
Post-stroke dystonia, 1036 arteries, 1096, 1101
Post-varicella angiopathy, 1020 veins, 1101–1102
Prechoroidal stage, 13, 15 Spontaneous intracranial hemorrhage, 244. See also Non-
Pregnancy, 857 traumatic intracranial hemorrhage
Presumed perinatal ischemic stroke, 1030 Stenosis, 1014
Primary angiitis of CNS (PACNS), 764 Stent(s), 135, 524–526, 529–531, 563, 591, 593
Proatlantal artery, 20 retrievers, 427, 429–430
Index 1191

String of beads, 1024 Time of flight, 392–393

String of pearls, 1022 Time-resolved MRA, 944
Stroke, 172–176, 299, 364, 376–377, 443–444, 449, 606, TOF-MRA, 1014
612, 685, 693 Torcula, 1033
mimics, 1011, 1013 Tortuosity, 1014
Stroke-like episodes, 1027 tPA, 1011, 1013, 1028
Sturge Weber syndrome (SWS), 1051, 1053 Tractography, 1036
Sub acute combined degeneration, 1158–1159 Transcranial Doppler (TCD), 448, 461
Subarachnoid dissection, 541 screening, 1024
Subarachnoid hemorrhage, 498, 523–526, 528–531, 939 Transforaminal approach, 49
complications, 508, 512 Transient cerebral arteriopathy, 1015, 1020
DSA-negative, 503 (see also Intracranial aneurysm) Transient ischemic attack, 206
Subarachnoid hemorrhage (SAH), 385, 404–405 Transtemporal approach, 48
Subclavian steal syndrome, 185 Traumatic head injury, 1028
Subdural hematomas, 405, 407 Trigeminal autonomic headaches, 795
Superficial temporal artery, 92 Trigeminal neuralgia, 781, 783
Superficial temporal artery-to-middle cerebral artery Tumors, 1068, 1074
(STA-MCA) anastomosis, 852, 857 Turbulent flow, 1014
Superior oblique myokymia, 787
Superior sagittal sinus, 937
Superior thyroid artery, 91 U
Surface coils, 129 Ultrasound, 1033
Susceptibility-weighted angiography (SWAN), 944 Unenhanced CT, 946
Susceptibility weighted imaging (SWI), 837, 845,
977, 1014, 1029, 1033
Susceptibility-weighted magnetic resonance imaging V
(SWI) Valvular heart disease, 1025
capillary telangiectasias, 742 Varicella zoster virus (VZV), 773
cerebral cavernous malformation, 731 Vascular anomalies, 1029
DVA, 739 Vascular disorder, 606. See also Brain arteriovenous
Symptomatic ICAS, 206 malformation
Systemic lupus erythematosus (SLE), 771 Vascular neuroembryology, 12
accessory MCA, 18–19
adult pattern, 17
T choroidal stage, 15–16
T2 shine-through, 1014 dominant anterior choroidal artery, 19
T2-weighted, 1014 duplications, 24
T2*WI gradient-echo MRI sequences, 942 fetal posterior cerebral artery, 18
Takayasu arteritis (TA), 40–41, 142, 144 fusion patterns variations, basilar artery, 20, 24
criteria for diagnosis, 40 persistent hypoglossal artery, 19
etiology, 40 persistent trigeminal artery, 19
treatment, 41 prechoroidal stage, 13
types, 40 proatlantal artery, 20
Tandem occlusions, 430, 437 6 weeks, 16
Technic MR, 1080 7 weeks, 16–17
Temporal arteritis. See Giant cell arteritis (GCA) 9 weeks, 17
Temporary clipping, 542 Vascular neurophylogeny, spinal cord, 3–4
Territorial distributions, 1015 Vascular territories, 364, 367, 378
Thalami, 939 Vascular venous imaging, 919
Thalamic hemorrhage, 1033 Vasculitis, 142–146
Thermal index, 64 Vasculopathy, 451–454, 1015
Thrombectomy, 426–430, 1028 Vasogenic edema, 945
Thrombocytopenia, 1031 Vaso-occlusion, 441
Thromboembolic, 1025 Vasospasm, 48, 54, 57, 1029
Thrombolysis, 1170, 1172 Vein of Galen aneurysmal malformation, 999, 1002
Thrombophilia, 1026 Venous edema, 939
Thrombophlebitis, 1032 Venous infarction, 939
Thrombus, 1032 Venous malformations, 985, 988
1192
Venous thrombosis, 274–276, 399
Venous time-of-flight (TOF-) MRA, 944
Ventricular assist device, 1025
Vertebral artery, 88–92, 172
dissection, 177
fenestration, 166
stenosis, 172, 177
Visual field deficits, 1032, 1035
Index
W
Wallerian degeneration, 1031, 1036
Watershed infarct, 330, 353
Watershed territory, 1132
Watershed/border zones, 1016
White matter lesions (WMLs), 807, 811
Williams, 1027
Wrapping, 526–527