The Hematology Journal (2002) 3, 56 ± 60

ã 2002 The European Haematology Association All rights reserved 1466 ± 4680/02 $25.00
www.nature.com/thj

Patterns of mortality in sickle cell disease in adults in France and England

VeÂronique Perronne1, Marilyn Roberts-Harewood2, Dora Bachir1, FrancËoise Roudot-Thoraval3,

Jean-Marie Delord4, Isabelle Thuret5, Annette Schae€er1, Sally C Davies2, FreÂdeÂric GalacteÂros1 and
Bertrand Godeau*,1
1
Centre de la dreÂpanocytose, HoÃpital Henri Mondor, CreÂteil, France; 2Haematology deÂpartment, Central Middlesex Hospital,
London, UK; 3DeÂpartement de biostatistique, HoÃpital Henri Mondor, CreÂteil, France; 4Service de MeÂdecine Interne, Centre
hospitalier du Lamentin, Lamentin, France; 5Service de peÂdiatrie, HoÃpital de la Timone, Marseille, France

Introduction: An understanding of the causes of death among patients with sickle cell disease
may be informative for both epidemiology and pathogenesis. This information should aid
anticipation of dangerous clinical conditions, counselling patients and design of preventive
therapies.
Patients and methods: All deaths known to four European sickle cell disease centres over a
10-year period were retrospectively analysed. The circumstances of death were classi®ed as
follows: (1) acute sickle related vaso-occlusion; (2) chronic organ failure related to sickle cell
disease; (3) infection; and (4) miscellaneous causes.
Results: Sixty-one adult patients (mean age: 32+11 years) died during the study period.
Twelve patients suddenly died at home; most of them exhibited symptoms of vaso-occlusion
but in eight patients, the cause of death was unknown. The primary cause of death in the 53
evaluable patients was sickle related vaso-occlusion (27 out of 53; 51%) which manifested
mainly by acute multiorgan failure (n=13) and acute chest syndrome (n=9). Ten of the 27
patients (37%) who died in these circumstances had an apparent mild disease before their
deaths. Ten patients (19%) died of documented infection. Ten of the evaluable patients (19%)
died of a chronic terminal visceral involvement related to sickle cell disease which was mainly
liver cirrhosis. Four patients died by suicide or because of refusal of care and two patients
died of iatrogenic complication.
Conclusion: The primary cause of death in adults appears to be vaso-occlusive, even in
patients with no overt organ-system failure. Our results emphasise that the circumstances of
death in sickle cell disease are di€erent between adults and children. The deaths among adults
appear not to be easily assigned to a few preventable causes as they are in children.
The Hematology Journal (2002) 3, 56 ± 60. DOI: 10.1038/sj/thj/6200147

Keywords: death; sickle cell disease; acute chest syndrome; vaso-occlusion; pregnancy

Introduction

An understanding of the causes of death among
patients with sickle cell disease (SCD) in di€erent
geographical areas may be informative for both the
epidemiology and the pathogenesis of the disease. This
information should aid anticipation of the dangerous
clinical conditions, counselling patients and design of
preventive therapies. Causes of death are well known in
children with SCD in whom Streptococcus pneumoniae
sepsis, acute anaemia related to splenic sequestration,
severe haemolysis and stroke, are the main causes.1 ± 4
Penicillin prophylaxis and counselling parents about
the importance of detection of an enlarging spleen, the
*Correspondence: B Godeau, Service de MeÂdecine Interne 1, HoÃpital
Henri Mondor, 51 avenue du MareÂchal de Lattre de Tassigny, 94000
CreÂteil, France;
Tel: + 33 1 49 81 20 76; Fax: +33 1 49 81 27 72;
E-mail: bertrand.godeau@hmn.ap-hop-pairs.fr
Received 6 June 2001; accepted 21 September 2001
dangers of fever and increasing pallor have dramati-
cally improved the prognosis for children.5 ± 7 In
contrast, there is less information available on the
causes and circumstances of death in adults with SCD.
The two large published studies focusing on adults
have been conducted in the USA and Jamaica2,8 and no
large study has been performed in Europe. The aim of
our study was to determine the circumstances of death
in a population of adults followed in European tertiary
SCD centres over the last decade.
Patients and methods
The study was conducted in three SCD centres in
France (HoÃpital Henri Mondor, CreÂteil; HoÃpital de La
Timone, Marseille; HoÃpital du Lamentin, Lamentin)
and one centre in England (Central Middlesex

Hospital, London). More than 1000 adult patients over
16 years were regularly followed by these centres.
About 62% of patients have SS disease (mean age
29.5+10 years), 29% have SC disease (mean age
35+11 years) and 9% have Sb thalassemia disease
(mean age 32+13 years) (these data are those of Henri
Mondor Hospital which are representative of all the
centres).
The medical records of all patients who died over a
10-year period were retrospectively analysed. For each
patient, the steady state haematological parameters
(foetal haemoglobin, leucocytes count and haemoglo-
bin value) were obtained. SCD genotypes were
established by combining family study, haemoglobin
analysis by isoelectric focusing, citrate agar electro-
phoresis and direct demonstration of b globin gene
mutation. The presence of a b Thalassemia allele
combined to bS gene was ascertained by family study,
but also by HPLC methods (at distance of transfu-
sion). These results were matched to the direct
demonstration of heterozygosity at the bS locus by
DNA study and in most cases by characterisation of
the b Thalassemia mutations. We assessed the patients
for recognised and chronic organ damage related to the
SCD (mainly eye, lung, heart, liver, kidney and bone),
the number of hospitalisations during the year prior to
death and any treatments directed at the prevention of
complications including hydroxyurea and transfusion
requirement, as well as the number of packed red
blood cell concentrates.
The circumstances of death were classi®ed as follow:
(1) Acute sickle related vaso-occlusion including acute
multi-organ failure syndrome, predominating acute
chest syndrome and stroke.
Acute multi-organ failure syndrome was de®ned as
the association of acute dysfunction of the lungs, liver
and kidney attributed to a sequestration of sickle cells
causing microvascular occlusion as previously pro-
posed.9 The organ failure may be associated with fever,
a rapid fall in haemoglobin (Hb) level and platelet
count, encephalopathy and rhabdomyolysis. Diagnosis
of acute chest syndrome was based on the presence of
fever or chest pain, associated with new pulmonary
in®ltrates on chest X-ray.10 Stroke included cerebral
infarction caused by occlusion of one of the major
vessels and/or intracranial haemorrhage;
(2) Chronic organ failure related to SCD including
debilitating stroke, congestive heart failure, cirrho-
sis, pulmonary hypertension and chronic renal
failure;
(3) Infection with microbiologically documented in-
fectious organisms; and
(4) Miscellaneous causes.
Statistical analysis
Results are expressed as mean+1 standard deviation
or as percentage + 95% con®dence interval. Compar-
isons are made by means of chi square test (or Fisher's
Death and sickle cell disease
V Perronne et al
57
exact test when necessary) for categorical data and non
parametric Mann Whitney test for quantitative data. A
P value 50.05 was considered signi®cant.
Results
Sixty-one patients (34 males, 27 females) with SCD
[47 with HbSS (77%), nine with HbSC (15%), three
with HbSb+ thalassemia (5%) and two with HbSb0
thalassemia (3%)] died during the study period. The
mean age at death was 32+11 years (range 16 ± 60
years) and did not signi®cantly di€er between SS
patients and others (31+11 years vs 36+12 years;
P=0.16). However, in eight homozygous patients who
had HbF level 410% (mean HbF value: 13.5+6%),
the mean age at death was in the fourth decade as
compared those who had lower HbF levels 510%
(mean HbF value: 3.9%+2.1) in whom the mean age
at death was in the third decade (42+7 years vs
31+10 years, P=0.05). Mean number of chronic
damages related to SCD at the time of death was a
1.8+3 while 13 patients (21%) had no chronic visceral
involvement. Half of the patients (30 out of 61; 49%)
were treated with repeated transfusions, and three
with hydroxyurea in addition. Two of the eight
patients with HbF level 410% received repeated
transfusions associated with hydroxyurea in one
patient. The mean number of chronic damages related
to SCD was higher in this group treated with repeated
transfusion or hydroxurea demonstrating a more
severe disease (2.7+0.9 vs 1+1; P50.001). Only one
patient was treated with hydroxyurea alone. Most
patients (43 out of 61; 70%) were regularly followed
in our centres. The others were seen only when the
patient su€ered acute complications. The majority of
the patients died in hospital (49 out of 61; 80%) and
most in intensive care units (30 out of 61; 49%).
Twelve patients died suddenly at home; most of them
exhibited symptoms of vaso-occlusion, but in eight
patients, the cause of death was classi®ed as unknown
because of insucient data.
Patient characteristics according to their circum-
stances of death are summarised in Table 1. The
primary cause of death in the 53 evaluable patients
was acute sickle-related vaso-occlusion (27 out of 53;
51%) which was manifested by acute multi-organ
failure (n=13), acute chest syndrome (n=9) and
stroke (n=5) (Table 1). Four of the 27 patients who
died of sickle-related vaso-occlusion were pregnant
and had an apparently mild disease before their
deaths. Two of them had SC, and two had
respectively SS and Sb. They died of acute chest
syndrome (n=3) and multi-organ failure (n=1) in
the immediate post-partum period. Post-mortem
examination was performed in two acute chest
syndrome (ACS) cases and revealed fat embolism
and pulmonary artery thrombosis, respectively. One
male patient also with apparently mild disease died
of ACS during the postoperative period after surgical
repair of an abdominal hernia. Ten of the 27
The Hematology Journal
 Death and sickle cell disease
V Perronne et al
58
Table 1 Circumstances of death in 61 adults with sickle cell disease
Mean value and
standard deviation Mean number of Number of patients
Number of patients of HbF value in Chronic visceral receiving repeated
Number of patients (%) with SS patients with SS involvement related transfusion and/or
Causes of death (%) Mean age (years) disease disease to SCD hydroxyurea (%)
Acute sickle related
vasco-occlusion 27 (44%) 31+12 20 (74%) 4.6+2.9 1.5+1.4 10 (37%)
Infection 10 (16%) 32+10 8 (80%) 6.8+1.8 1.1+0.9 3 (30%)
Terminal chronic
visceral involvement 10 (16%) 32+1 7 (70%) 5.2+4 2.8+1.1 7 (70%)
Suicide or refusal of care 4 (7%) 32+10 4 (100%) 3.8+4.3 2.8+1.5 3 (75%)
Iatrogenic complication 2 (3%) 21+1 2 (100%) 4 2+0 2 (100%)
Unknown 8 (13%) 34+10 6 (75%) 6.8+6.2 2.4+0.9 4 (50%)

Table 2 Comparison of patients who died of acute chest syndrome (n=9) with patients who died of another vaso-occlusive event (n=18)
Death related to acute chest Death related to acute multiorgan
syndrome (n=9) failure (n=13) or stroke (n=5) P value
Mean age (years) 24+3 34+14 0.2
Number of patients with SS disease
(%) 6/9 (66%) 14/18 (77%) 40.5
Mean value and standard deviation
of HbF in patients with SS disease 4.8+4 4.1+2.6 40.5
Number of patients (%) with
percentage of F Hb 4/=10% among
the patients with SS disease 1/9 (11%) 1/18 (6%) 1
Percentage of patients receiving
repeated transfusion and/or
hydroxurea 22% 50% 0.2
Percentage of patients who have
received more than 20 erythrocyte
concentrates 22% 69% 0.04
Mean number of visceral
involvement related to SCD 0.8+1.3 1.8+1.3 0.06

patients (37%) who died of acute sickle-related vaso-
occlusion did not have chronic visceral involvement
related to SCD and had not received repeated
transfusions or hydroxyurea; they mainly died of
ACS (n=6) (Table 2).
Ten patients (19%) died of documented infection.
Four patients died of Streptococcus pneumoniae
meningitis, three of whom were infected by HIV.
Other infectious causes of death were infection with
Gram negatives rods [Escherishia coli (n=2), Enter-
obacter cloacae (n=1)], Staphylococcus aureus (n=1),
Plasmodium falciparum (n=1), and varicellae pneumo-
nitis (n=1).
Ten other patients (19%) died of chronic terminal
visceral involvement related to SCD (cirrhosis, n=6;
chronic renal failure, n=3; pulmonary hypertension,
n=1). In ®ve of the six patients, cirrhosis was partly
related to transfusion haemosiderosis associated in
three patients with Hepatitis C virus infection. Three
of these ®ve patients died of hepatic failure despite the
use of deferoxamine.
Two patients committed suicide (4%, 95%) ic [0.5 ±
13%] and two patients (4%) who su€ered from severe
SCD with mutivisceral involvement refused transfu-
sions and died, respectively, of acute heart failure and
terminal chronic renal failure. Two other patients died
The Hematology Journal
of iatrogenic complications: haemorrhagic shock after
liver biopsy and digitalis intoxication.
Comments
Our study is the ®rst large study focused on the
circumstances of death in a population of adults
followed in European SCD centres. Although it is a
retrospective study which su€ers from possible bias
and under reporting, sickle-related vaso-occlusion was
the primary cause of death in about half of the
patients. Moreover, this incidence is likely to be an
underestimate because most patients who died sud-
denly at home were classi®ed in the group of `death of
unknown origin' while some of them had complained
of pain, a sign of vaso-occlusion. Similar results8
suggest that pain episode is the main circumstance of
death in American adults with SCD. It has also been
demonstrated2 that one predominant cause of death in
SCD patients older than 20 years is vaso-occlusion,
particularly ACS, in Jamaica. Fatal vaso-occlusion is
manifested mainly by acute multi-organ failure and
ACS. In contrast, death related to stroke in our adults
in the European study was rare, while it is a more
common cause of death in children.

This study highlights the risk that adults who appear
relatively ®t are susceptible to acute multi-organ system
failure and may die suddenly during a painful episode,
and particularly of ACS. Over one third of our patients
who died of vaso-occlusion did not have chronic
visceral involvement related to SCD and had not
received modifying therapy such as repeated transfu-
sions or hydroxyurea. This emphasises the need for
prevention and education for all SCD patients, even
those adults with apparently mild disease. In addition,
patients who exhibit signs of severe vaso-occlusion
should be encouraged to present early, allowing more
time for e€ective treatment and admission into the
intensive care unit for those who are most severely
a€ected. The fact that four pregnant women with
apparently mild disease suddenly died of vaso-occlu-
sion highlights the fact that pregnancy remains a risk
period for acute severe events in SCD, even in patients
with SC or Sb disease.2,4
Ten of our patients (19%) died from infection: four
from Streptococcus pneumonia meningitis which is one
of the primary causes of death in children.6 Three of
whom were HIV-infected. SCD patients with HIV
infection are at high risk for fatal Streptococcus
pneumonia infection11 and the risk of death related to
Streptococcus infection appears low, in contrast, in
SCD adults who are not infected with HIV.
The incidence of death from chronic organ failure
related to SCD was low in our study (520%), the
primary cause being cirrhosis, while only three patients
died of chronic renal failure. This result contrasts with
the pattern of chronic organ involvement reported
from North America and Jamaica, where kidney failure
is the predominant cause.2,8 In SCD, vascular hepatic
lesions can be responsible for acute and/or chronic
ischaemia that may be severe.12 However, cirrhosis in
our patients was mainly due to haemosiderosis and
hepatitis C as a consequence of recurrent transfusions
References
1 Rogers DW, Clarke JM, Cupidore L, Ramlal AM,
Sparke BR, Serjeant GR. Early deaths in Jamaica
children with sickle cell disease. Br Med J 1978; i: 1515.
2 Thomas AN, Pattison C, Serjeant GR. Causes of death in
sickle-cell disease in Jamaica. Br Med J 1982; 285: 633.
3 Leikin SL, Gallagher D, Kinney TR, Sloane D, Klug P,
Rida W. Mortality in children and adolescents with sickle
cell disease. Pediatrics 1989; 84: 500.
4 Gray A, Anionwu EN, Davies SC, Brozovic M. Patterns
of mortality in sickle cell disease in the united kingdom. J
Clin Pathol 44: 1991; 459.
5 Serjeant GR, Serjeant BE. Management of sickle cell
disease; lessons from the Jamaican cohort study. Bl
Review 1993; 7: 137.
6 Bunn HF. Pathogenesis and treatment of sickle cell
disease. N Engl J Med 1997; 337: 762.
7 Davis H, Schoendorf KC, Gergen PJ, Moore Jr RM.
National trends in the mortality of children with sickle
cell disease, 1968 through 1992. Am J Pub Heal 1997; 87:
1317.
Death and sickle cell disease
V Perronne et al
59
and was `iatrogenic' in some ways. It highlights areas
for improvement possibly through better methods of
viral inactivation of blood products and better iron
chelation treatment. Blood transfusion can be life-
saving and can improve the morbidity associated with
recurrent neurological complications or ACS, but is
not a panacea in itself. Strategies should be instituted
to recognise patients who will develop haemosiderosis.
The incidence of death related to suicide or refusal
of care was relatively high in our study and higher than
those observed in the general population in Europe,13
although similar to those observed in other chronic
severe diseases such as multiple sclerosis.14 Moreover, it
is possible that some patients who were not under
regular follow-up in our centres and who suddenly died
of acute complications had psychiatric problems
explaining their delay in seeking hospital care. Even
though this result must be interpreted with caution due
to the small number of patients, it suggests that poor
compliance, cultural and social issues with psychiatric
morbidity may be underestimated in SCD adult
patients and should be considered in patient manage-
ment.
Our results demonstrate that the circumstances of
death in adults with SCD are multifactorial, thus less
easily addressed that in children where simple measures
have had a signi®cant impact on the morbidity and
mortality of the disease. Whilst the primary cause of
death in adults appears to be vaso-occlusive, with or
without overt organ failure, patients with mild disease
also died suddenly. Only through addressing post-
partum care, comorbid conditions including psychiatric
conditions, compliance with therapy, patient awareness
of disease complications, prevention of haemosiderosis,
chronic visceral involvement with SCD and iatrogenic
complication, can we attempt to prolong the life
expectancy of these patients.
8 Platt O, Branbilla DJ, Rosse WF, Milner PF, Castro O,
Steinberg MH, Klug PP. Mortality in sickle cell disease.
Life expectancy and risk factors for early death. N Engl J
Med 1994; 330: 1639.
9 Hassel KL, Eckman JR, Lane PA. Acute multiorgan
failure syndrome: a potentially catastrophic complication
of severe sickle cell pain episodes. Am J Med 1994; 96:
155.
10 Maitre B, Habibi A, Roudot-Thoraval F, Bachir D,
Desvaux Belghiti D, Galacteros F, Godeau B. Acute
chest syndrome in adults with sickle cell disease:
therapeutic approach, outcome and results of bronch-
oalveolar lavage in a monocentric series of 107 episodes.
Chest 2000; 117: 1386.
11 Godeau B, Bachir D, Schae€er A, Brun-Buisson C, Billy
I, Portos JL, Galacteros F. Severe pneumococcal sepsis
and meningitis in human immunode®ciency virus-
infected adults with sickle cell disease. Clin Infect Dis
1992; 15: 327.
The Hematology Journal
 Death and sickle cell disease
V Perronne et al
60
12 Charlotte F, Bachir D, NeÂnert M, Mavier P, GalacteÂros
F, Dhumeaux D, Zafrani ES. Vascular lesions of the liver
in sickle cell disease. Arch Path lab Med 1995; 119: 46.
13 Anonymous. World health report 2000 health systems.
WHO-OMS. Avalaible on line: http://www.oms.ch/
14 Koch-Henriksen N, Bronnum-Hansen H, Stenager E.
Underlying cause of death in Danish patients with
multiple sclerosis. Results from the Danish multiple
sclerosis registry. J Neurol Neurosurg Psychiatry 1998;
65: 56.

The Hematology Journal