Interdisciplinary Science Reviews

ISSN: 0308-0188 (Print) 1743-2790 (Online) Journal homepage: http://www.tandfonline.com/loi/yisr20

Platinum Complexes for the Treatment of Cancer

Barnett Rosenberg

To cite this article: Barnett Rosenberg (1978) Platinum Complexes for the Treatment of
Cancer, Interdisciplinary Science Reviews, 3:2, 134-147, DOI: 10.1179/030801878791926119

To link to this article: http://dx.doi.org/10.1179/030801878791926119

Published online: 20 Nov 2013.

Submit your article to this journal

Article views: 40

View related articles

Citing articles: 110 View citing articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=yisr20

Download by: [Monash University Library] Date: 02 July 2016, At: 07:58
 Platinum Complexes for
the Treatment of Cancer
PROFESSOR BARNETI ROSE BERG
Michigan Slale University, East Lansing, Michigan, USA

Some platinum coordInation complexes are active anticancer drugs in animal and man. Tbi new cia s of
chemotherapeutics was dIscovered during the course of investigation of the electric field eOect on bacterial
growth. The platinum electrodes electrolyzed during the experiment, relea ing a platinum complex which caused
complete ce alion of cell division in the bacterial rods. With this filamentation assay system, we were able to
identify the specific chemical as cis-dichlorodiammineplatinum([J), a comple known ince 1848. The bacterial
tudies with many uch complexe suggested to us the generalizations that charged platinum complexe were
bacteriocidal, while the neutral platinum complexes induced filamentation and, in lysogenic bacteria, Iy is. The
Downloaded by [Monash University Library] at 07:58 02 July 2016

neutral complexes have ignificant activity again t transplantable, virally induced, and chemicaUy induced
cancers in animals. They are synergistic with almost every other anlicancer drug in current u e. In man kidney
toxicity i the do e limiting ide effect but this is now completely ameliorated by simply hydrating the patient. The
drug in combination therapy, ha proven to be 'potentially curable' for all form of te tkular cancer. Other
cancers where the drug activity has begun to approach this are head and neck cancer, and ovarian cancer. Activity
again t the other major cancers i now being studied. The mechanism of action at a molecular level appear to
depend upon a primary Ie ion formed on the cellular D A by the platinum complex. ThIs serendipitous
discovery has led to a new class of anticancer agents, metal coordinalion complexe , which may prove to be of
significant value.

Cancer research ha e calated enormously in this past
decad . D pite thi , it i till a rare occurrence for a
new cla of anticanc r drug to urface. In the ran-
dom screening of chemicals, only about J in 10000
shows significant activity to warrant further te t . It i
all the m re interesting when a new clas of drug
emerg fr m an unexpected area of chemistry.
Therefore it is not urpri ing that the report of
anticancer activity of platinum coordination com-
plexes from this laboratory in 1969 evoked trong but
mixed reactions. There was joy among t inorganic
chemi t , who, for too long, have been largely ex-
cluded from medical re arch; and th r wa kepti-
cism among medical scientist, who are d eply con-
ditioned to consider heavy metal compounds as
poi on . Now that the fir t platinum drug, cis-
dichlor diammineplatinum(Il), ha proved to be of
value in treating many human cancer, it may not be
inappropriate to de cribe brieRy the hi tory of its
discovery and the subsequent tortuous development
leading to it clinical use.
EARLY HISTORY
The tory b gin in 1961, when I left the Phy ic
Department at New York University to
the Biophysics Department at Michigan
versity. With this change of department
134 INTERDISCIPLINARY SCIENCE REVIEWS, VOL. 3, NO.2,
an obligation to orient my r arch more toward
biology. In my earlier reading I had been fa cinated
by the microphotographs of the mitotic figur in c II
in process of division. They called to a phy ici t'
mind nothing 0 much a the shape of an electric or
magn tic dipole field, th kind one ees with iron
filings over a bar magnet.
If such a dipole may be involved in cell divi ion - a
some had earLier speculated - then by tickling the
dipole with electromagnetic radiation of a re onant
frequency or a ubharmonic, to avoid the radiofre-
quency heating f cell, it may ab rb ome energy
which mayor may not b detrim ntal t the c II.
Admittedly this is an overburden of 'may ,but I wa
intrigued by the idea of an experimental test. Having
no comp t nc in bioi gy - f w phy icist have - L.
VanCamp joined the laboratory to do the te t. We et
up a continuous culture apparatus for the cells, but
included in the growth chamber a set of platinum
electrodes. Platinum of cour e, i known to b quit
inert in a biologic environment.
The electrode were powered by an audi amplifier
who e input frequ ncy wa t by an audio oscillat r.
The impedance of the chamber, 6 n, wa j.) rfectly
matched to the output impedance of the amplifier. To
test the proper functioning of [he apparatu b for
putting in mammalian cells we used the common
help found bacterium Escherichia coli. These, and prokaryotic
State Uni- cells generally, do not show mitotic figure in divi ion.
there came After the bacterial population reach d a t ady tat ,
CCC-0308-0 18 /7 /0003-0134$07.00/1
1978 © He den & on LId, 197
 the electric field was turned on. The density of bac-
teria tarted to decline, and we were in danger of
having an a epticchamber. When the field wa turned
off, the den ity returned to normal after a few hours.
A rather triking effect, but how striking we did not
realize until we examined the bacterial cell in the
effluent of the chamber. The bacterial rods normally
look like the picture in Fig. la, rods about 2-5 ~m
long, with a 1 ~m diameter. After an exposure to the
electric field they appeared a in Fig. l(b); long
filament , up to 300 times the usual length. ow thi
required an explanation. The effect wa not due to a
direct action of the electric field n the bacterial cell
but rather to electroly i product from the platinum
electrodes.
We now brought our chemi t, T. Kriga , in to
i alate and identify these product. He cI arly iden-
tified it as a platinum containing compound probably
ammonium chloroplatinate (NH4h[PtCI6]. We were
(a) (b)
omewhat nonplus ed, however, when addition of
Figure 1. (al Scanning electron microphoto-
Downloaded by [Monash University Library] at 07:58 02 July 2016

thi compound at the detected concentration to bac-

terial cell in test tube culture led, not to filamenta-
tion, but to bacteri cidal activity. Many experiments
later we found that a olution of thi compound, after
standing on our laboratory shelf for a few week, wa
able to produce a mall amount of hart filament.
Some quick studies showed that light wa the
nece ary agent for the change,' and we were now
deep into the photochemi try of platinum. In retro-
spect, this wa not urpri ing. Platinum compound
antedate iIver in photography. Ultraviolet light
cau ed a eries of chemical reactions in the olution,
leading from the charged into a final neutral
pecie [pev ( H3hCI4]. Bacterial test of the epa-
rated intermediate and final neutral product showed
graph of normal E. coli (gram-negative rods).
(bl Scanning electron microphotograph of
E. coli grown in medium containing a few
parts per million of cis-dichloro-
diammineplatinum(II). Same magnification in
both pictures. The platinum drug has inhibited
cell division, but not growth, leading to long
filaments. These pictures were taken by D.
Beck of Bowling Green University.
that the latter wa the chemical causing filamentation
and was chemically identical to th electr Iytically
formed agent. A. Thomson, in our laboratory yn-
thesized the neutral pecie by known ch mical tech-
niques and te ted it. It had no activity!

PROFESSOR BARNETT ROSENBERG has been at the Michigan State Uni-

versity since 1961, and is now Distinguished Professor of Biophysics. Born
and bred a New Yorker, he completed his Bachelor's degree in Physics at
Brooklyn College and began his graduate career at the Eidgen6ssiche
Technische Hochschule in Zurich. In 1949, after this wanderjahr, he reo
turned to New York University to obtain the Masters and PhD degrees in
experimental physics, followed by a postdoctoral year at the now Courant
Institute of Mathematical Sciences. He worked in industry at Westinghouse
Electric Corporation as a Senior Research Physicist, but returned again to
New York University as a Project Director on a biophysical project in the
Department of Physics, With the late Leroy G. Augenstein, he organized the
Graduate Department of Biophysics at Michigan State University.
In earlier years, his research interests in Biophysics were the study of
solid state properties of biological molecules and systems such as
bimolecular lipid memb ranes. The work described in this review paper has
been a continuing preoccupation up to the present, with a side excursion
into the interesting area of thermodynamics of aging and death.
Address: The Department of Biophysics, Michigan State University, East
Lansing, Michigan 48824, USA.

Professor Barnett Rosenberg with his long-time associate, Mrs Loretta VanCamp in their Laboratory at Michigan
State University, discussing the latest research results obtained from animal experiments with mice. In their work,
as the author states in this review' the entire panoply of disciplines in chemistry and biology were needed.' Before
any newly designed drugs are assayed on men, they are, of course, first tried on mice and other animals.

© Heyden & Son Ltd, 1978 INTERDISCIPLINARY SCIENCE REVIEWS, VOL. 3, ND. 2, 1978 135
 (a) (b) I H
probably Salvar an developed by P. Ehrlich ab ut
the turn of the century as a specific for syphilis. It wa
~' also almo t the last of the line. For, in the first half of
thi century, rapid progre in organic chemistry and
H) I
biochemistry produced a proliferation of antibac-
terial drugs culminating in the enorm u Iy ucce ful
sulphonamide and finally, the antibiotics. Thi uc-
Cl cess fixed the attitude of th next generation of
(oj NH~C1 scienti t and metal complex were largely ignored
thereafter.
The antibacterial activity of some platinum group
HJf4-ICl metal complexe wa fir t studied by F. P. Dwy rand
CI CI his co-worker in 1953. They found the relatively
inert chelated complexe of ruthenium with
Figure 2. Molecular structures of anticancer
phenanthroline to be quite good bacterio talic and
active (cis configurations) and nonactive
(trans configurations) platinum complexes. bacteriocidal agent against Gram-p itive micro-
(a) cis-dichlorodiammineplatinum(II); (b) organisms. Unfortunately, these charg d c mpl e
trans-dichlorodiammineplatinum(ll); (c) cis- al 0 produced a evere neuromu cular toxicity,
tetrachlorodiammineplatinum(IV); (d) trans- 'curare like' behavior, which limit d their u e to
tetrach lorodiam mi neplatinum( IV). topical (skin) administration. Limited clinical trial
Downloaded by [Monash University Library] at 07:58 02 July 2016

did e tablish a usefulne for these comple 'e in the

W had only one remaining possibility. The neutral treatment of some skin infections such a dermatosis,
compound exi t in two isomeric m difications; the dermatomycosi and others but little furth r work
trans-[pev ( H3)2C14] and the cis-[pev ( H3)2CL4]. was done to bring these complexe into general u e.
The former wa the more thermodynamically stabl Our laboratory first called attenti n to the bacterial
and was the one we first prepared. We now synth- effects of the impler complex in 1 65. Over th
e ized the cis configuration and finally achi ved next few year, in cooperativ studies with mic-
complete ucce . robiologi ts a number f paper were published
Platinum ha two dominant valence state, + 2 and de cribing a multiplicity of effect on micr organi m
+4. The lower state form quare planar complexe , cau ed by various complexe of platinum group
and the latter forms octahedral compLexes. We now metals: platinum palladium, ruthenium, rhodium
synthe ized the 2 + complex, and it al 0 was active in osmium and iridium.
forming filament. Thus the two active chemical are C n ider fir t the filamentation effect. Trial of
cis-[PtII(NH3hCh] and cis-[Pt'V (NH3)2C14]. These many complexes e tab Ii h d that mainly tho e com-
structure are shown in Fig. 2. The trans tructures plexes which were neutral and had n lectrically
have the two similar chemical groups (ligands) on charged ions in olution, markedly inhibited cell divi-
oppo ite sides of the molecule and both trans specie sion in bact ria. The cis configuration was activ ,the
3fe inactive at low concentration (parts per million in trans wa not. They did not inhibit gr wth unl the
olution), but begin to uppress growth at higher cone ntrations were greatly increa d. Th Y were
concentration. a ociated in the cell primarily with nucleic acid
Now we had done a trange thing, for by the (RNA and DNA) and with orne so]uble proteins.
circuitou route de cribed, we had di covered a com- Gram-negative rods were the mo t en itive to thi
pound fir t ynthe ized in 1 48 and known a effect, Gram-po itive rod much Ie sand sph rical
Peyrone's Chloride. The molecular structural differ- bac.illi (cocci) not at a]1. F rming a filament wa not a
ences between the cis and trans complexe had been terminal event for the bacterium. If the platinum
olved by Werner in 1 90, who, in 0 doing, e tab- complex was removed from the o]ution, r the fila-
ti hed the ba i of modern coordination chemistry. ments tran ferred to a normal m dium uitabl for
What little value we added by thi whole exerci e was growth, the filament begin to divide into n rmal
the use of a biologic te t for identification of the bacteria, looking much like a string of sau ag in the
complex, thu establishjng a clear and interesting proce , growing into quite normal colonie. The
biologic activity of orne coordination complexe of division occur all along the length of th filament and
platinum. not just at the end. This wa quite a different pattern
from the filamentation cau ed by chemical uch a
the nitrogen mustard, where filamentati n i a termi-
nalevent.
THE EFFECTS OF PLATINUM COMPLEXES The difference may reside in th {act that the
o BACTERIAL CELL nitrogen mu tard block DNA synth i and each
such filament contain only a mall number of copie
ClinicaJ u e of metal complexes, particularly of ar e- of the genetic information (genome) wherea th
nic, antimony and mercury in the treatment of bacter- platinum complex does not top new DNA synthe i
ial infection ha a long history. The noble t cion was in bacteria at the concentration causing filaments to

136 INTEflDISCIPLINARYSCIENCEREVIEWS,VOl. 3, NO.2, 1978 © Heyden & Son LId, 1978

 appear, and th D A exists in multiple genome that had been previously infected with a bact rial
copies a continuou trand or large clump through- viru (A -bacteriophage). In these lysogenic bacteria,
out the filament. Thi by the way, i quite different the genetic information f the virus has been incorpo-
from the effects of these complexe on mammalian rated into the cell, but it i repre ed that it i n t
cell a di cussed below. It i al 0 on of the major normally detectable. It replicat during c II divi ion
differences in the biological effects of the nitrogen along with the bacterial D A and 0 i not 10 t or
mu tards, bifunctional alkylating agents and potent diluted out after many divi ion .
anticancer agent and the platinum complexe . This is the bacterial equivalent to low or lat ot
The complexe which form ions in olution such as virus infections in mammals and man. Th e ba trial
(NfLhfPtCI4] which ionize to PtCI~- and strain are called lysogenic, ince a number of phy i-
2( H4t are quite poisonous to the bacteria, cau ing cal agents uch a X-ray or U.v. light and ome
a large cell kill at low concentrations, and few or no chemical uch a the nitr g n mu tard and car-
filament . The e ion react with proteins in the cyto- cinogens, can derepre the viral g n me cau ing an
pia m of the cell almo t exclu ively, compared to the active viral infection leading to th di lution - th
trong nucleic acid association of the neutral Iysi - of the cell. These effects are easily m a ured
complexe. when the bact ria are grown in te t tub culture. Th
Mea urement of the e vari u reaction required platinum complexe , for example, are added as a few
a en itive technique for detecting th minute parts per miJlion concentration in the growth
amounts of platinum incorporated by tb cell. This medium. The bacteria grow, forming filament for
necessitated the use of a radioactive isotop of about three hour then rather quickly, th milky
platinum as a tracer. E. Renshaw and A. Thorn on opacity of the culture dimini he , and in a few hour
Downloaded by [Monash University Library] at 07:58 02 July 2016

produced the i otope t91Pt y irradiating an iridium the culture i water clear; the cell hav allly ed. Th
foil in the proton beam of the Michigan State Uni- cis-[Ptll(NH3)2Ch] complex i extremely efficient in
versity Cylotron, chemically eparating the platinum inducing uch ly i -Ie than 0.1 ppm in th cultur
i otope from the other metals pres"ent and synthe iz- produces a detectable effect.
ing the charged and the neutral complexe for the It hould be pointed out h r that all the platinum
bacterial te t . In recent year ,the radioactive isotope complexes wh.icb are active anticancer agent ar al 0
195mpt ha been generously prepared for us at Oak efficients inducers. Those complexes which are not
Ridge Laboratories by K. Poggenberg. This also is a active, do not cau e Iy i . S far, there i a c mplete
y-ray emitting isotope with a three day half-life. Thi isomorphi m betwe n th et f active anticancer
mean a bout of hectic activity in our laboratory with complexes and the et of effici nt inducer. arlier,
each delivery to accomplish all our experiment we had believed that a good c rrelati n xi ted
before the level of radioactivity diminishes below our between anticancer active compl and th
detection thre hold. filament-f rming complexe . After a whil , how v r,
R. Gillard and hi co-workers at Kent University exception in both cia es occurred which decreas d
have extended the e tudie and hown that organic our faith in this correlati n. The correlation of lytic
complexes of rhodium produce imilar effects. G. induction and anticancer activity has held up well. In
Gale and his associates at the Medical Univer ity of fact, when R. Adam on at the ational nc r In ti-
South Carolina developed a parallel story of the tute reported the anticancer activity of gallium alt
photochemical transformations and filamentation of we te ted the e and, indeed they did induce Iy i in
bacteria by the cis i orner of the neutral complex with ly ogenic bacteria.
iridium in tead of platinum. Thus the experience now We then te ted alt of oth r gr up JIIa elem nt ,
accumulated ugge t a generality of the bacterial aluminium and indium, and th t pr ved t be
phenomena with the complexe of the other platinum inducer. It wa only after we had predict d but n t
group metal . publi hed, th activity of aluminum and indium in
Certainly the development of new bacteriocidal these te t that Adam on reported them active a
agents particularly tho e which are active against anticancer agent . The verificati n of the prediction
Gram-negative bacteria, i a very de irable goal. tended to reinforce our belief in th c rr lati nand
However, the report of anticancer activity of these more important, in the po ibility of a imilar
complexes hifted tbe weight of re earch to thi more mechani m of action in the tw appar ntly di imilar
urgent problem. And ju t as the electric field experi- effect.
ment wa bypa ed - temporarily I hope - so too the In 1953, A. Lwoff, had reviewed evidence hawing
bacteriocidal utility wa put in limbo by more exciting that water soluble mutagen, carcinogen and anti-
developments. Before moving on to the develop- cancer drug were potent induc r of ly ogenic
ment a third bacterial effect need di cu ion ince it bacteria, a strong hint that underlying he four
provide orne in ight into the po ible mechani m of different effects there was a common mechanism and
how cancer are affected by these complexes. that it involved an interaction of the cau ative agent
S. Vasilukova, nee Reslova a young with cellular D A. The importance of the agent-
Czech lovakian microbiologi t, and an ex-student of caused Ie ion in the D A in these processe wa
J. Drobnik who contributed much to our microbial further enhanced when Va ilukova returned t h r
experiments worked with trains of E. coli bacteria native land and performed the experiment called

© Heyden & on Ltd, 197 INTERDISCIPLINARY SCIENCE REVIEWS, VOl. 3, NO.2, 1978 137
 'indirect induction. In this case a strain of nonlyso-
genic bacteria with the sexual transduction factor, P
was treated with the platinum complex. These cells
were allowed exually to conjugate with a lysogenic
strain, F-,. which had not been treated with the pla-
tinum complex. In this process only a portion of the
D A of the cell is transferred. Yet the recipient cells
were induced to lyse. Later on I will peculate on the
sequence of events arising from this correlation in
order to account for the anticancer action of the
platinum drugs. But first I must carry the story for-
ward to the di covery of their utility a cancer drug.
THE ANTICANCER ACfIVITY OF PLATINUM
COMPLEXES
By 1968 we had achieved a certain degree of under-
tanding of the bacteriologic effect of the platinum
complexe and we had ynthe ized and repeatedly
Downloaded by [Monash University Library] at 07:58 02 July 2016
tested the cis -dichlorodiammineplatinum(II) which
we now took as a model for the active neutral com-
plexe . We were primed to try the cbemical again t a
cancer. The logic wa omewhat nai've: the complex
topped cell divi ion in bacteria at concentration
without marked toxicity, perhaps then it would stop
cell division in tumors which grow rapidly, without
unacceptable toxicity to the host animal. J. Toth-
Allen fir t determined the afe do e level which
could be injected into the peritoneal cavity of mice.
The dose which killed 50% of the animals (LDso) wa
about 13 mg of the drug per kilogram of animal body
weight. A dose of 8 mg kg-1 was nonlethal.
VanCamp then implanted in these mice a tandard
tran plantable animal tum r the olid Sarcoma-180.
Thi wa admini tered a a 10 mg piece of tumor
tissue inoculated beneath the skin under one armpit.
The tumor fragment increased its mass about 100
time over the next ten day. Th tumor could be cut
out, since it remained localized, - non-meta tatic-
and weighed. The tandard protocols of the National
Cancer Institute called for implantation of the tumor
on day 0; injection of the drug on day 1; and sacrifice
of the animal on day 8. The average tumor ize of the
treated group i divided by the average tumor size of
the untreated group, the negative control. For a drug
to be considered effective against the tumor, the
treated to control (TIC) tumor size ratio should be
less than 0.5. Our first test values were well below
this. We repeated this te t more than half a dozen
times to be sure it was not a peculiarity due to our
inexperience. It wa not; each new test reconfirmed
the activity.
We also te ted a number of other neutral platinum
complexe and in this new biological effect, we
again saw the stereospecificity that had occurred in
the bacterial te t ; the cis configuration were active,
the COfre ponding trans configurations were not. The
implication of this are ignificant: it show that the
platinum complexe retain their geometry in the
138 INTERDISCIPLINARYSCIENCEREVIEWS.VOl. 3. NO.2. 1978
biologic environment, they were not degraded to
heavy metal ions which could be non pecific poi on .
The specific chemical reaction leading t th biologi-
cal effect wa sensitive to mol cular geom try and
wa rna t likely to involve a macromolecule such as a
protein or a nucleic acid. We also were presented with
a simple test to determine the significant chemical
reaction. Both cis and tran complex und rgo
roughly imilar, multiple reaction in the cell, but
obviou Iy only tho e reaction which the ci config-
uration can undergo, but the (ran configuration can-
not are likely to be significant.
After confirming these re ult I contacted G. Zub-
rod, head of the chemotherapy branch of th ationaJ
Cancer Institute and apprized him of the result. I wa
invited to discuss this with hi as ociates at Bethesda,
Maryland. After my short lecture, which was received
with perceptible, but understandable coolnes 1 left
sample of the four complexe to b te ted in th ir
tumor creen, th L1210 I ukemia in mice. few
month later I wa informed that th complexe were
also active in their system, and it wa suggested to me
that a grant propo al to pur u thi re earch w uld
not be unfavorably received. A propo al was duly
submitted, and approved, and N. .1. support ha
continued ever ince.
But mea'nwhile, we had tried a variant of the pro-
tocols and this gave us the first hint of the true potency
of these complexes. Instead of injecting the com-
plexes on day 1, we waited until the tum r wa ab ut
a gram in weight (in a 20 g mall !) and th n inj cted
the drugs (on day 8). All the tumor r gr d and all
the animal were cured. A time equence photograph
of two mice is hown in ig. 3. This wa an unu ual
result since we were not aware of any other anticancer
drug capable of regres ing large SarcQma-l 0
tumor. A w uld be exp cted, the urviving animal
showed strong immulogic rejection of r implant of
the same tumor up to the longest time te ted, ] 1
months. The animals lived out their full life expec-
tancy; about 30 month and died of normal age
related, cau es.
With confirmed activity against two diff r nt ani-
mal tumor we were ready to publi h the preliminary
results, which we did in a hart paper in Nature in
1
1969. An American journal f alma t equal distinc-
tion had turned tb manu cript d wn b au a
referee had commented that it wa not n t worthy
since a many new drug with activity w r bing
found. Indeed, there wa a tr ng po ibiIity that th
Nature paper would be lost in a crowd of imilar
reports of new anticancer chemicals that were flood-
ing the literature at the time. It wa re cued from
potential oblivion by the intere t and g ad grac of
Professor Sir Alexander Haddow, then head of the
Chester Beatty Institute in London.
Curiou Iy, he had an intuitive feeling that platinum
complexe might be effective anticancer agents, and
had already tested orne earlier, with ut ucce . On
hearing of our re ult he had th n w c mplexe
synthesized and te ted again t a differ nt tumor y-
© Heyden & Son Lid. 197
 Control sarcoma 180

Died day 21

day 8 day 12 day 16 day 20

Treated - single injection cis - [Pt II (NH3hCI2J -8 mg kg -1 on day 8
Downloaded by [Monash University Library] at 07:58 02 July 2016

day 8 day 12 day 16 day 20 day 36

Figure 3. Time sequence photographs of two mice with solid Sarcoma-laO tumors. The mouse at
the top was an untreated negative control. She died on day 21 when the tumor weighed about 3 g.
The bottom mouse was in the group treated on day 8 with an intraperitoneal injection of
cis-dichlorodiammineplatinum(II). Her tumor was completely regressed six days after treatment,
and she died of age related causes almost 3 years later.

tern, a myeloma tumor (ADJ/PC6) in mice, and again field, are concerned with it. A tabulati n of th best
confirmed the activity. He wrote to me of the re ult animal te t re uIt is hown in Tabl 1.
obtained at the Chester Beatty In titute and extended The best result are indicat d here olety to convey
an invitation to visit with him and orne of his col- a qualitative impression of the activity. Mo t te t
leagues, which I accepted with alacrity. Thi began a were performed with small number of animal , mak-
trong cooperative group in Britain which included ing stati tical analy e meaning]e . Be ide, th p r-
T. A. Connor and J. J. Rob rt of the Che ter Beatty ver ity of animal re p n es which indicate a lack f
Institute R. J. P. William of Oxford University, and appreciation or knowledge of all th imp rtant vari-
M. Tobe at University Colleg London al] of whom able nece sary to control, mak ither the average
have contributed much to the advance of this new or b t re ults of dubious num rica] valu . N verth -
re earch field. Ie , we have compiled a Ii t of 16 tum r type,
[Per onal comment: We lat r learned that R. including tran plantable tumors, chemically derived
Mason, who had helped in ur arli r bacterial tumors (from carcinogen) and virally derived tumors
tudie had ent orne cis-dichloro- (from oncogenic viru e ). The drug i active again tall
diammineplatinum(lI) to a friend to t t for anti- type. The conclu ion that may b drawn from the e
cancer activity in 1966. His friend ov rdo ed the te t are that the drug:
animal , they all died and he reported back that the
drug was too toxic! There mu t urely be a les on 1. exhibit marked rather than marginal
omewhere in this story.] antitumor activity;
In recent year, cis-dichlorodiammin platinum(II) 2. ha a broad spectrum f a tivity again t drug
ha been te ted again t a wide variety of anima] re i tant as well as drug ensitive tumor ;
tumor .2 While thi drug is by no mean the mo t 3. i active again t low growing a well a rapidly
active of the platinum complexe , it wa the fir t growing tumor ;
cho en by the ational Cancer In titute to be lated 4. is active again t tumor normally in ensitive to
for clinical trials. his fact made tudies with thi drug S' phase (D replicative tage) inhibit r ;
more imperative, and rna t further research making 5. regre es tran plantabl a well a hemically
up the bulk of the, by now, ver 6 0 paper in the and virally induced tumors;

© Heyden Son LId. 197 INTEROISCIPUNARY SCIENCE REVIEWS, VOL 3, NO.2, 1978 139
 Table 1. Best Resulls of the Anti-tumor Activity of cis-dichlorodiammineplatinum(U) in AnimaJ y tem

Tumor Host Best Results

Sarcoma-laO solid Swiss white mice T/C"'2-10%"

Sarcoma-180 solid (advanced) Swiss white mice 100% cures
Sarcoma-180 ascites Swiss white mice 100%cures
b
Leukemia L1210 BDF, mice % ILS"'379%; 4/10cures
Primary Lewis lung carcinoma BDF, mice 100% inhibition
Ehrlich ascites BALB/c mice % ILS;=300%
c
Walker 256 carcino-sarcoma (advanced) Fisher 344 rats 100% cures; T.I.>50
Dunning leukemia (advanced) Fisher 344 rats 100% cures
P388 lymphocytic leukemia BDF, mice % ILS", 533%;b 6/10 cures
Reticulum cell sarcoma C+ mice % ILS= 141%b
8-16 melano-carcinoma BDF1 mice % ILS '" 279%.b 8/10 cures
ADJ/PC6 BALB/c mice 100% cures; T.I.=8c
AK leukemia (lymphoma) AKR/LW mice % ILS = 225%;b 3/10 cures
Ependymoblastoma C57BL/6 mice % ILS=141%;b 1/6 cures
Rous sarcoma (advanced) 15-1 chickens 65% cures
DMBA-induced mammary carcinoma Sprague Dawley rats 77% total regressions
3/9 free of all tumors
ICI 42, 464-induced myeloid and lymphatic leukemias Alderly Park rats % ILS=400%b

" T Ie ~ _T_u_m_o_r_m_a_s_s_in_tr_ea_t_ed_a_ni_m_a_'s
x 100.
Downloaded by [Monash University Library] at 07:58 02 July 2016

Tumor mass in control animals

b % ILS:= % increase in lifespan of treated over control animals.
c TI =Therapeutlc index (LD50/ED90J. ED90:= effective dose to inhibit tumors by 90%.

6. ha hown no animal specificity since it work in political infighting that characterize so much f d-
mice and rats, either inbred or random bred, and in ence eem to have b en muted by tbe urgency of the
chickens' problem at hand.]
7. is useful for di eminated (e.g. leukemias) as
well as solid (e.g. arcoma) tumor;
. is potent, in that it can rescue animal when MOLECULAR TRUCTURE DETERMIN
injected a few days before death from certain types THE ANTICANCER ACfIVlTY
of tumor.

Thu the credential of the drug, and by implication,
other in th cia of platinum group complexes, a an
active anticancer agent in animal i well e tabli h d.
We were then faced with a serie of que tion , the
an wers to which were urgently needed and which
required for the e ao wer expert competence in
coordination chemistry biochemistry, biophysics
molecular biology phy iology, pathology, phar-
mac logy electron microscopy, immunology and
finally, clinical medicine. In hart the entire panoply
of discipline in chemi try and biology were needed.
We alone could not do it, nor could any small group of
laboratorie . A worldwide network of cooperating
laboratories wa called for, and e tabli hed. They
were supported by public fund, cancer ocietie , and
to a generous degree, tbe platinum indu try.
[Per anal comment: I recall two strict admonitions
from my major profes or when I informed him of my
growing intere t in biophysics. The e were, not to
work with medical doctor untrained in re earch and
to avoid cancer re earch, ince many had tami hed
their reputation from a malignant neglect of scien-
tific objectivity in their de ire to do omething useful.
I have broken both injunction but I cannot ay that I
am sorry. With a very few exceptions all connected
with the network impressed me as dedicated eLfless
humane scientists. The expected ego c1ashe and
Of the myriad que tion ari ing from this di covery
of the anticancer activity in mice of orne platinum
complexe , one which we did feel c mp tent t
attack, particularly ince we had the advic of orne
very able inorganic chemi t , R. Ma n of u ex
Univer ity R. J. P. William of Oxford niver ity
and M. Tobe of University olleg London, wa the
so called structure-activity relati n. Without requir-
ing detailed knowledg of the molecular interactions
of the chemical in the biologic sy tern, we imply
induced tructural change in the molecule by known
synthetic techniqu and t ted them again t a tan-
dard mou e-tumor. Ii many variation are tried, th n
a catalogu of th e chemical, with a impl numeri-
cal mea ure of their activity houJd exhibit orne
regularities. This allows them to b gr up d in ub-
classes, and each subcla can b analyzed for com-
mon chemical propertie . Th more chemicals that
are tested, the harper will be the subcla ification.
However, a reasonable limit had to be et for the e
ynthe es.
As an example, on tart with one metal f the
platinum group platinum i elf. It ha the two major
valence tate +2 and +4. Take the latter. It can
a ociate in an octahedral complex with i ligand,
atomic group bound to the m tal. Th individual
ligand may be chosen from a larg group, but let us

140 SCIENCEREVIEWS,VOL. 3, NO.2, 1978

INTERDISCIPLINARY © Heyden & on Ltd. 197
 re trict it to just ten. Therefore, for this one metal
valence state we have ab ut one million potential
variations. Obviou Iy the required manpower to
ynthesize, purify and characterize such numbers of
chemical i b yond the world's capacity even if aJl
laboratorie were recruited for thi Ie purpo e; to
ay nothing of the 30 milli n mic required.
A dra tic c mpromi e was caJled for, and here the
intuition of the exper in c ordination chemi try wa
es ential. Actually, only about 1000 complexe hav
been tudied in variou laboratorie . Of the e, ab ut
10-20% are active. The molecular tructure of ome
of the most active comple e are hown in Fig. 4. The
number of 'active' is this large simply because mo t
have tarted with a known active complex and pro-
duced mall variations upon it. In a large moras of
chemical, we have found through luck or cleverness,
a few mall island of succe and we stray far from
these only at ome peril. It is not a very satisfying
situation when endangered grant renewal i the pen-
alty for boldne s.
Downloaded by [Monash University Library] at 07:58 02 July 2016
everthele s, in the area explored ome common
features have emerged which link tructure to activ-
ity. We embody these here in a set of 'rules of thumb'
since they can hardly lay claim to general validity.3
The e are:
1. the complexes exchang onJy ome of their
ligands quickly in reactions with biological
molecule;
2. the complexe hould be electrically n utral,
although the active form may be charged after
undergoing ligand exchanges in tbe animal'
3. the geometry of the complexe are either square
planar or octahedral;
Figure 4. Molecular structures of new
platinum group metal complexes with high
activity against animal cancers. (a) dich-
loroethylenediammineplatinum(lIl; (b) sub-
stituted (R) malonatodiammineplatinumlll);
leI cis -dichloro-bis-(cyclohexylamine)plati n-
um(ll); (d) sulfato-1,2 diaminocyclohexane-
platinumOl); (e) rhodium(ll) carboxylate.
© Heyden & Son Ltd, 1978
4. two cis monodentate or one bidentate leaving
group (exchangeable ligand) are required; the cor-
re ponding trans i omer of the monodentat I av-
ing groups are generally inactive'
5. the rate of exchange of the e group hould fall
into a restricted region since too high a reactivity
will mean that the chemical reacts immediately
with blood con tituents and never get to th tum r
cell while too Iowa reactivity would allow it t get
to the cell , but they would do n thing once there;
6. the leaving groups hould be approximately
3.4 apart on th mol cule (an intere ting
number inc the pacing b tween the step f th
Watson-Crick D A ladder is also 3.4 A)·
7. the groups across the mole ule fr m the leaving
groups sbould be strongly bonded, relatively inert
amine type systems.
We certainly do not intend these rule t re tri t
future research, but only to encompass a large
amount of past experience with platinum(lI) com-
plexe . Obviou Iy xcepti n will, and have already,
occurred. For example the high activity f bid nlat
leaving group uch a oxalate and mal nat (e
structures of Fig. 4) first synthesized by M. I are and
J. Hoe chele in thi laboratory are not encompa ed'
nor is the effect of cyclic amin , developed by obe,
which decrea e the Jubility of the comple es, but
markedly enhance the antitum r activity. H re
studie of the relative olubilitje in oil and wat r,lh
partition coefficient, may b ignificant in d termin-
ing activity.
We can now pre ent a broad outlin of the fate f
the drug, cis-dichlorodiammineplatinum II), after
injection into the p ritoneal cavity of the mou e.
Within minute the drug leaves the cavity thr ugh the
blood and lymph circulation. The high chI ride con-
centration of the e extracellular fluid prevent the
chlorides from leaving th molecul thus maintaining
the tructural integrity. The intact drug i rapidly
excreted in the urine, with a half-life in th b dy of
about one h ur. Th excreted drug is 95% th
unchanged molecule but about 5% i attached to
proteins. The drug j pas ively tran ported acr the
cellular membrane - no active tran port (carrier)
is nece sary. Once inside the cell, the lower chloride
content of the cytoplasm (1/30 of that outside the
cell) allows the chloride to exchange with water
according to the following cheme:
[Ptll(NH3hCI210+ H20¢ [Ptll(NHJh(H20)Cll +CI-
[Ptll(NHJh(H20)Clt+H20¢[Ptll H2MH20)212 + 1-
Depending on the hydrogen ion cone ntration the
HzO may be changed to (OH)-. Thi aquated specie
react primarily with the nitrogen of the 0 Aba es
leading to the primary Ie ion re pon ible for the
anticancer effect. While the formula for the diaquo
pecie implie a impl, ingle tructure, we have
recently di covered that it i slightly m re compli-
cated than that. In fact, isolation of cry tal p ci of
INTERDISCIPLINARY SCIENCE REVIEWS, VOL. 3, NO.2, 1978 141
 the diaquo complex under slightly different condi-
tions have yielded one monomer, one hydroxy 1·20
bridged dimer, two hydroxy bridged trimers a tet-
ramer and two other not yet re olved crystal forms.
Thi emerged from a cooperative tudy between B.
Lippert of this laboratory and C. J. L. Lock of
McMaster Univer ity. It is not yet clear what role if
3H -Leucine
any the e variou tructure have in th anticancer
activity or toxicity of the parent drug. 0·80
l:
I have been purpo ely nebulou so far on where the ..•.o 3 H - Uridine
DNA i , and what ites of the 0 A are involved. In 'o"
Q.
order to be more concrete we require a short descrip- o00,60
tion of the molecular biology studie . .!:
III

..•.>
'"
Gi
a:
WHAT DO THE PLATINUM COMPLEXES DO 0'40
3 H - Thymidine
TO MAMMALI CELLS?

Two options are available generally to study the 0·20

effect of drug on c II . fir t inject the drug in
Downloaded by [Monash University Library] at 07:58 02 July 2016

animal excise the desired cells and examine these

for change , the in vivo sy tern' second u purifi d
cells growing in ti ue culture, the in vitro sy 'tern. he o 2 4 6 10 24h
former i more relevant, but the latter i cientifically Incubation time in 5·0 ~M cis-I Pt "(NH3)2CI21 media
'cleaner'. Both should be done, and in the ca e of the
Figure 5. The effects of cis-dichloro-
platinum drug, were imultan ou Iy and indepen- diammineplatinum(lll on macromolecular
dently p rf rmed in two laborat rie. Both th syntheses in human amniotic cells in tissue
technique were u ed by Gale and hi a ociates at the culture at a concentration (5J.LM=1 ppm)
Medical Univer ity of South Carolina whiJe only the similar to that found in tumor cells in animals
treated with a therapeutic dose of the drug.
second was done in our laboratory. There i general
DNA synthesis is measured by the incorpora-
agreement on the result obtained by the two tion of radioactively labeled thymidine and is
4
methods. severely and persistently inhibited. The synth-
By the u of radioactively labelled precur or esis of RNA, measured by radioactive uridine.
chemical, the cell' ability to ynthe ize mac- and protein. measured by radioactive leucine
is not significantly different from control
romolecules such a D A, R A and proteins, after
(nontreated) cells represented by the horizon-
treatment with the drug, can b measured. A typical tal bar at 1.00.
re ult i hown in Fig. 5, for exp ur of the cell to
the equival nt level of drug found in tumor ti ue of a
treated animal. The ynthesis of new total DNA i the hypothe i that the anticancer a tlVlty f the
electively and per i tently inhibited. Total R A and platinum drugs ari es from a primary attack n D
protein synthese are not markedly affected until The battle to di cover the molecular m hani m of
much higher drug dose levels, which are frankly toxic action wa therefore, joined n the fi Id of metal
to the cell , are u ed. Th level of inhibition of DNA complex interactions with DNA, and numerou other
ynthe i i dose depend nl. It on t i low, taking laboratories entered the fray. The b oty ha b en
about 4-6 h after drug expo ore to reach a nadir. It rich embarra ingly o.
was surprising that there was not a large cell kill at the We now kn w that ci -dichlorodiammine-
therapeutic do e level. The cell fir t grew into giant platinum(II) can cro link the two trand of the
c II which, after a few day, hawed th app aranc doubl helix of D A, an exciting discovery since thi
of many nuclei and eventually divided into a number type of linkage had already b n invoked to acc unt
of single cell . I will return to thi important re ult for the anticancer activity of the bifunctional alkylat-
later. ing agents such as the nitrogen mustard . It wa made
H. Harder, then in our laboratory but now at almost simultaneously in three laborat rie ,but mo t
George Wa hington Univer ity, wa re pon ible for elegantly by J. J. Robert and hi c -work r at the
these tudie. He also checked that the synthesis of Che ter Beatty In titute. It can at 0, apparently
precursor molecule for the D A, and the tran port era link two neighboring ba e tacked on a single
of the e acro membrane wa not respon ible for trand, which ignificantly, the complex in the trans
the inhibition. He ha more r cently hown that th configuration hould not do. The platinum drug doe
ability f the D A to act a a templat for new not react with the sugar-phosphate backb n of the
synthe is is strongly inhibited by the platinum drug. strands, but only with the bases. or doe it appear t
These results can be rna t rea onably explained by intercalate between the ba es. It react mo t trongly

142 INTERDISCIPLINARY SCIENCE REVIEWS, VOL. 3, NO.2, 1978 © Heyden on Lid. 197
 with the G-C rich regions of DNA, and can, through
the technique of gradient centrifugation, be used to
characterize the relative G-Cj A-T content of
DNA.
The platinum complex-D A reaction is very
lowly rever ible in vitro, but it may be removed more
rapidly witbin the cell by the actiolls of D A repair
enzyme. The two available exchangeable group of
the platinum can react at two ites on a given ba e
(primarily the purines, guanine and adenine), or with
ingle ites on two different base, or finally, a ingle
ligand of platinum exchange at only one ite of a
ba e. It will take a considerable period to sort out the
multiplicity of such reaction and to identify finally
one or more a the nece ary lesion for anticancer
activity. In the meantime, it is clear to many of u that
metal complex interactions with nucleic acids are too
poorly understood, and too important to remain so.
It must surely be nagging the reader by now, as it
ha u for ome year that in aJl of the ab ve work n
clear di tinction ha emerged between effects on
Downloaded by [Monash University Library] at 07:58 02 July 2016
tumor cell and on normal cells. The same effects
qualitatively and very likely, quantitatively appear in
many cell type ,and yet we have ugge ted that these
effects are the primary lesion leading to anticancer
activity. Chemical tudies of DNA are no doubt
important, but ince we cannot ay in what way DNA
differ in cancer cell from normal cells, we cannot
answer the question of why the cancers are killed and
not the animals. The question is by no mean trivial. It
cut to the heart of cancer chemotherapy.
Justifiably, some people are unhappy that we have
not yet di covered drug with higher curativ~ power
against cancer. However, many researchers, myself
included, are surprised that we have discovered so
many that are so good, because we do not know why.
All cancer drug are cellular poi on , but not all
cellular poi ons are cancer drug. If the drugs were
not eventually more poisonou to cancer cells than
normal cell , we would not be injecting them into
patient. With the po ible exception of
a paraginase, we have not yet been able to seize
upon a unique exploitable characteristic of cancer
cells to produce specific, or even selective tumor cell
kill. Yet in the ho t animal thi can, and does, occur
with presently useable drugs. This admittedly irn-
plistic, logic lead naturally to the invocation of the
ho t-tumor interaction rather than the drug-tumor
interaction a the ource of specificity in the anti-
cancer activity of drugs. Such specificity is usually
as ociated with the ho t' immune re pon e.
HOW DOE ELECTIVE C CER DESTRUC-
TID OCCUR?
Here I hould like to touch on a more p culative ide
of the research. We had to face up to the trong
evidence at the molecular level that tbe platinum drug
produced a lesion on the DNA of cells which did not
© Heyden & Son Ltd, 1978
necessarily lead to cell death, and in any ca e, wa not
restricted to cancer cells alone, and the final clinical
observation that the cancer disappeared in the ani-
mal without unacceptable side effect. There i a
wide gap b tween molecular biology and clinical
results. Could we bridge ome or all of it with testable
hypotheses?
The specificity of the cure i a good clue, ince a we
noted, specificity i u ually a dated with the h t'
immunologic respon e . I there any evid nce for
such re ponses of the host? There is, but thi evid nce
consists mainly of an accumulation of weak argu-
ment which cannot be summed to make a trong
argument. Briefly, these are: our earlie t creening
studie of coordination complexe brought to light
the peculiar result that some complexe incr a ed the
rate of tumor growth by about 200% compared to
untreated controls. This is consistent with the already
established ugge tion that the ho t animal exert
some con traint on the gr wing tum r thr ugh
immunologic reaction, and if these con traints are
inhibited (by immunosuppressive agents) without
these agents simultaneously exerting antitumor activ-
ity then increa ed tumor growth rat are xp cted.
The second involvement of th immune ystem
occurred when we were able to cure large olid
Sarcoma-lBO tumor in ICR mice. The cured
animal rejected any new attempt to reimplant this
tumor up to 11 month later. They have obviou Iy
developed a heightened immunologic reactivity
for this tumor. Interestingly enough the cure of
small tumors did not produce such an immunologic
rejection reaction.
This experiment also produced a third unexpected
result. It has been accepted since the classic work of
H. E. kipper and co-worker at th uth Tn
Research In titute, that at least for leuk mia, a given
dose of a drug kill a con tant fraction of the tumor
cells pr. ent, in fact a first order kinetic proce s. Yet
we are able to cure mall tumor and larg tum r
with a given optimal do of the platinum drug but not
intermediate ized tumor. Thi i not en ible if one
considers direct cell kill only. imilarly, if the optimal
do e cures the large tumors, then a much smaller dose
should cure the smaller tumors. This, too, is contrary
to our exp riments. Something other than direct
chemical cell kill mu t be operating to achiev cure .
Dead Sarcoma-1BO cell injected into mice do not
cause tumors, neither do they induce an immun
reaction to reimplanted live tumor cell. Here one
mu t be cautious since only mall number of live
implanted cell (- 40) can ev ntually lead to larg
tumors and death. But cell treated with the platinum
drug at low concentrations, 100 times Ie than the
concentration required to produce extensive cell kill,
implanted in the mice do not produce tumor, but
do induce an immunologic rejection of pristine
tumor cells implanted two weeks later. This experi-
ment is difficult to interpret without invoking the
immune system of the host in cau ing tum r cell
death.
INTERDISCIPLINARY SCIENCE REVIEWS. VOl. 3, NO.2. 1978 143
 If the immune system is involved, then one could
anticipate that modulating the ho nimmune comp -
tence should modulate the anticancer activity of the
drug. Preliminary experiments by P. Conran in this
laboratory, and now at the University of Connecticut
suggest. that thi i true. Decreasing the
immunocompetence of mice by hydrocortisone injec-
tion decrease the cure rate of the platinum drug
against Sarcoma-180 in ICR mice; while conver ely,
the non pecific immune stimulant, zymo an,
increa es the cure rate again t the Sarcoma-180 in
BALB/c mice. Unfortunately the e ystem are not
a 'immunologically clean' a on would like so the
experiment are now being repeated using acceptable
systems hoth in our, and Conran's, lahoratorie .
One of Tobe' complexe, cis-dichloro-bis-
(cyclohexylamine)-platinum(II) wa te ted by T. A.
Connor again t the ADJ/PC6 myeloma tumor in
mice. It cured the tumor completely at a dose 1/500
of the LDso. Such specificity, e pecially in the
ab ence of any evidence of selective tumor uptake of
Downloaded by [Monash University Library] at 07:58 02 July 2016
the drug is utterly inconsistent with the direct cell
toxicity hypothesis. A ho t re ponse must again be
invoked, one with high specificity. And again, the
immune system alone has that characteristic.
Finally, we can count the number of tumor cells a a
function of time in the animal after a platinum drug
treatment known to produce large percent cure. The
A cite Sarcoma-180 in ICR mice i ideal for this
purpo e. We inject four million cell into the
peritoneal cavity on day O. They multiply rapidly to
two thousand million cells 15 days later, remain-
ing localized in the cavity and killing 100% of the
animals. Now we inject the platinum drug on day 1,
and acrifice small numbers of the animal every day.
We wash, clean, and count the tumor cell . The
re ult are shown in Fig. 6. The cell divide about 2-4
times, increasing in number up to 40 million cell by
day 4-5 before a turnaround occurs and th cell
number drop to 0 OD days 9-] O. If direct tumor cell
kill by the drug wa operative we hould expect a fast
decline from about eight million cells to 0 on day 2
with no further cell d~visions and no continued
growth. This is contrary to th experimental results
and again sugge t a host mechani m for tumor cell
destruction.
We propose that the e argument are con istent
with, but do n t corroborate, the hypothesi that the
platinum drug enhances the antigenic character of the
tumor cells, tipping the balance in favor of the ho t'
immunologic intervention to de troy the cancer. The
que tion then becomes, how doe the platinum drug
accompli h this? . Now we return to the derepression
story based on Reslova's work. In brief, Fig. 7 out-
line one potential equence of molecular event in a
mammalian cell which could produce the de ired
re ult.
Without making a definit commitment, let u
as ume the hypothesis that expre ion of viraJ D Ai
the cau ative factor in the cell tran formation to a
cancer state. There i certainly a ignificant body of
144 INTERDISCIPLINARY SCIENCE REVIEWS. VOL. 3. NO.2, 1978
100
80
<D~
~
~
!!! 60
"ii
<>
0
E
-
~
0
~
40 " -,
q
\
E
z=' \
20 \
\
16
\
12
C\
8 ......
0
.•.•..
4 - _D_
O
Figure 6. The growth of Ascites Sarcoma-' 80
tumor cells in untreated leA mice (solid Iinel
and in treated mice (dashed line). The treat-
ment consisted of five injections of
1.5 mgkg-' given on day 1 after inoculation of
4 000 000 tumor cells on day O. The number of
cells continues to increase by repeated cell
division up to day 4 and then slowly decreases
to zero cells (cured animals).
experiments indicating that thi i true in many mam-
mal, but olid evidence in human till elude u . The
viral genome, incorporated in the cellular gnome, i
completely repre d for long period c mpar d to
rna t cells division times. A in the c f Iy g nic
bacteria, a wide vari ty of ch mical and phy ical
agent are able to cause a depre sian of a small part of
this latent viral genome, enough ay to code the
production of one or two protein . Th e protein
transform the cell. The exi tence of temperature en-
sitive mutants for cell tran formati n tell us that the
production of as little a one virally c ded protein i a
neces ary (but not ufficient) cau e f cell tran forma-
tion. This small numb r of proteins i also the cau e of
the antigenicity of the tum r cell. The chemical and
phy ical agents cau ing d repre ion are therefore,
carcinogen. If now we add the platinum drug to the
cell, it effectively derepresses a larger fraction or all of
the viral genomic information. Thi inevitably lead
to the production of a larger number and vari ty of
protein, and thi enhance th antigenicity of the
cell.
Thi cherne, incidentally, provide a simple expla-
nation of 'Haddow s paradox' that i ,certain cia se
of chemical and agents that cau canc r can al 0
cure cancer. Th difference between cau e and ure
indicated here i imply a quantitative one. It i the
amount of derepression of th viral genome.
© Heyden n Ltd. 1978
 Viral infection

C3
cCJ Inregration
~ • [J ph •••

~ Uv .. X-rays, carcinogens

Viral replication
@D ~ T~•• lo'm"

Virulent infection
[(J ,-----Pt
~A.h9'.'
complex

® Increased antigenicity
©
Downloaded by [Monash University Library] at 07:58 02 July 2016

© Antigens

Possible virulent infection

Figure 7. A schematic diagram of one possible hypothesis of the molecular action of cis-
dichlorodiammineplatinum(lI) leading to an enhanced antigenicity of tumor cells by the derepres-
sion of virally coded information latent in the cell.

orne experimental information con istent with

our hypothesis exists. V. Vonka and co-workers at
the Institute of Sera and Vaccines in Prague, were
able to induce up to 300% increa e in the number of
Ep tein Barr virus (an oncogenic Herpes type virus)
po itive cell in a culture of Burkitt lymphoma cells
(EB3) by treatment with cis-dichlorodiammine-
platinum(II). The induction of the new, virus a 0-
dated antigens was monitored both by an indirect
immunofluore cence test for the coat proteins of the
virus appearing at the cell surface, and by the vi u-
alization of virus like particles in the treated cells by
electron microscopy.
Thus the platinum drug, at lea tin thi ca e, cau e
the hypothesized derepre ion in a cancer cell line,
and ba enhanced the antigenicity of the cells. While
the enhanced antigenicity hypothesis i consistent
witha large body of information and does bridge the
gap between molecular events involving tbe platinum
drug interaction with cellular DNA and the host
immunologic intervention, it till leaves unexplained
the detailed mechanisms of derepres ion of latent
viral genomes, its role in cell transformation and the
nature of the immune re ponse. Fortunately, many
laboratorie are working on the e particular prob-
lem.
HOW EFFECITVE ARE THE PLATINUM
DRUGS AGAI ST HUMAN CANCERS?
The end of the story mu t be the an wer t thi
question, for if the drugs are no improvement ver
pre ent drug therapie , coordination complex a
anticancer agent will pa into hi tory and be forgot-
ten. Happily the emerging c1inicaJre ult do sugge t a
significant treatment improvement in, at lea t, a
number of types of cancer .5
The first platinum drugs entered human clinical
trials in 1972 after a sene of preclinical tudie of the
toxic ide effect of different do e level in dog and
monkey. The e tudie e tablished a afe low tart-
ingdo e, and predicted certain toxicitie which would
occur when the do e level wa e calated. The m t
severe side effect wa the damag to the kidney, with
less stringent toxicity to the blood forming element
of the bone marrow intense nausea and vomiting
(which we have good reason to believe is due to a
central nervous sy tern effect and not to ga t£Ointe ti-
nal damage) and orne destruction of the ound
detecting hair cells of the organ of Corti in th inner
ear leading to tran ient high frequency hearing los ,
and in a few ca es, total deafne . The kidney damage

© Heyden & Son Ltd, 1978 INTERDISCIPLINARY SCIENCE REVIEWS, VOL, 3. NO.2, 1978 145
 is a evere limiting side effect on the allowed do e
levels. Nevertheless, below these dose level anti-
cancer activity wa reported.
The patient entered into fir t clinical trials of a new
drug are u ually terminally ill with cancers that are no
longer re pan ive to any treatment, and who have
igned an informed consent agreement. It cannot be
expected that uch patient, who e physiologic vital-
ity ha been verely apped both by the illness and
the sub equent therapy, particularly with drug which
are immuno uppre sive, will show marked re pan e
to the new drug. But invaluable information on the
side effects in humans and appropriate dose levels
and schedules can be obtained. In the ca e of the
platinum drug, the predicted side effects occurred
plu the hear-ing loss effect, which could not be ea ily
predicted from animal studies, just a the nau ea and
vomiting effect which was fir t found in dogs and
monkey could not have been predicted from all our
mice and rat studies, since the e latter animal cannot
Downloaded by [Monash University Library] at 07:58 02 July 2016
vomit.
Still orne patient did benefit from the treatment,
with partial or complete remissions in about 20% of
the patients, for various time duration. It was noted
early on that the drug eemed to produce more
re pan e in patient with genito-urinary cancer,
-80% for testicular cancer, >90% for ovarian
carcinomas, head and neck cancer (-40%), and
orne lymphoma (- 40%), than in other types of
cancer. The e preliminary re ults tend to augment the
Impression that the platinum drug i active against a
fairly broad spectrum of cancers. One total failure
noted was colon carcinoma, a very drug re istant
cancer. Even here, however, a bright note occurred
when th platinum drug is used in combination
chemotherapy as described later.
At this earlier stage, a far e timate of the clinician'
attitude toward the platinum drug was encapsulated
in the remark, 'thi is too interesting a drug to drop,
but too dangerou for extensive use'. The danger
referred to wa the kidney toxicity, an unfamiliar ide
effect with the older cancer drug.
In the second phase of the clinical trials, the drug
was used on early patients in a crossover pattern with
other drug . Tho e patients not responding to the first
drug were ubsequently treated with the econd drug
and vice versa. I hould point out that although
variou phase and protoc I of te ting are neatly
blocked out by the Nationa] ancer Institute in the
United States, the e delineation tend to be some-
what blurred in actual practice sinc th medical
doctor' first priority is to benefit the patient rather
than exactly fulfill an experimental protocol. Previous
experience with other drug in Phase II tests generally
produced higher re ponse rates, and the platinum
drug wa not an exception.
hemotherapy of cancer with ingle agent ha not
yet proven to be of ignificant long-term value.
Within thi la t decade large improvement in
re pan e rate occurred when drugs were admini -
tered in combination. The desiderata for uch combi-
146 INTERDISCIPUNARY SCIENCE REVIEWS, VOL. 3, NO.2, 1978
nations are synergi m, or at lea t additivity, of anti-
cancer effects, and I s than additivity for any on
type of do e limiting ide-effect. A number of uch 3-,
4- and even 5- drug combination ha b en dev I p d
with encouraging results in patient, but true yner-
gism is relatively rare. The platinum drug wa
reported by J. Venditti and his co-worker at
the ational Cancer Institute to produce ynergistic
action in animal tumor system with a large number of
other anti-cancer drugs. This wa verified hortly
thereafter by R. Speer and H. Ridgeway at the
Wadley In titute of Molecular Medicine in Dalla,
Texa , and a number of other laboratories including
our own. Thi provided the necessary rationale for
combination therapy with the platinum drug.
The first success wa found by F. An field and R. A.
Ellerby at the Univer ity of Wi can in. They treated
terminally ill colon carcinoma patients with the cis-
dichlorodiammineplatinum(II) drug and 5-
fluorouracil. Three of nine such patient hawed
objective (greater than 50% reduction of tumor
masses) respon e for varying duration. This i
presently under further tudy at the May Clinic in
Rochester, Minne ota.
The tudies of J. Wallace J. Holland and co-
workers at the Roswell Park Memorial In titute in
Buffalo, New York, on testicular cancer probably
saved the platinum drug from premature burial. They
reported re pon in 13 of 15 patients, even of
which were complete and of long duration. Her
again another frustration in cancer re arch was
manife 1. Drug do exi t which can, in a large percen-
tage of ca e , cau e complete r gres ions - disap-
pearance of all detectable symptoms of cancer, but
only for limited times. The tumor then return with a
vengeance and is usually refractory to further treat-
ments. Why not cure?
A cure i defined by the fact that the pati nt
achieve a normal life expectancy and di of cau e
not related to the cancer. Op rati nally however,
thi cannot be an acceptabl definition. In tead,
experience dictate that if a given cancer doe n t
return within am p cified tim (usually ab ut 5
years) the patient is considered cur d. Therefor
improved therapies which pu h the survival rate
further out in time could cro the threshold to pro-
duce cure. Thi may, in the near future occur with
testicular cancers.
E. Cvitkovic, l. Krakoff and co-worker at the
Sloan Kettering Memorial Center for ancer and
Allied Diseases in New York mad th ne t pu h
toward this goal. vitkovic rea on d that heavy metal
kidney toxicity can be ameliorated by administering
an a motic diuretic agent such as D-mannit I, whil
hydrating the patient. Since platinum is a heavy m tal
cau ing kidney to icity, why not try this treatm nt?
While the logic i not totally defen ible, hi xperi-
mental re ult were effective - he wa abl t increa e
the afelyadmini tered do e of th platinum drug by
300%, with a concommitant increa e in efficacy. The
platinum drug wa th n added to their tandard mul-
©He den n Ltd. L978
 tiple drug therapy which on average, put about 50%
of the patients in complete remi sian for a median life
expectancy of 6 month. The new combination ha
now produced greater than 90% complete remis-
ion , who e mean duration cannot b e tabli hed
inee in the ucceeding 14 months there have b en
little or no relap e. imiIar good response in the
treatment of testicular cancer with th platinum drug
in combination therapy were arso reported by L.
Einhorn and co-worker at Indiana University
Medical Center in Indianapoli , Indiana, and by C.
Merrin at Ro well Park Memorial In titute.
Thi optimi tic progre s towards achieving com-
plete remi ion in large percentages of the patient
and for increasing duration i now happening in treat-
ment of ovarian carcinoma. The fir t indication of
activity of the platinum drug in thi cancer was
reported by E. Wilt haw of the Royal Marsden
Ho pital in London in 1973. Again, combination
therapy bas proved more efficaciou than single
agents. H. Bruckner J. Holland and a ociate at
Downloaded by [Monash University Library] at 07:58 02 July 2016
Mount Sinai School of Medicine in New York
developed a combination of the platinum drug with
adriamycin and more recently have added cycJo-
LITERATURE CITED
1. B. Rosenberg,l. VanCamp, J. E. Trosko and V. H. Mansour,
Platinum compounds: A new class of potent antitumor
agents. Nature (London) 222, 385--386 (1969).
2. B. Rosenberg, Platinum coordination complexes in cancer
chemotherapy. NatUfwissenschaften 50, 399-406 (1973).
3. A J. Thomson, R. J. P. Williams and S. Reslova, The
chemistry of complexes related to cis-Pt(NH"hCI2 - an
antitumor drug, Struer. Bonding (Berlin) 1" 1, (1972).
4. T. A. Connors and J. J. Roberts (Editors), Platinum Coordi-
nation Complexes in Cancer Chemotherapy pp. 79-97.
Springer-Verlag, New York (1974).
© Heyden & on Ltd, 1978
phosphamide, and are reporting response rate of
70% and better, with longer durations of re pon e.
Cancers of the head and neck are now treated quite
effectively by a combination chemotherapy utilizing
the high do e platinum drug, with bleomycin infu ion,
as developed at the Memorial Sloan-Kettering
Cancer Center by R. E. Wittes and his co-worker.
More recently, A. Yagoda and a ociates at the same
institution have shown the marked efficacy against
bladder carcinomas of the platinum drug, alone' and
in combination chemotherapy. And finally both
Merrin's group and Holland's group hay now added
pro trate carcinoma to thi growing Ii t f re pan ive
cancer.
It j now certain that combination therapi with
the platinum drug are a afe, effective treatment for a
number of different typ of cancer. Thi will undoub-
tedly be extended in the future to a number of other
cancers; to a variation of therapies to improv th
present resuits; and finally to include econd and
third generation platinum coordination complexe
which are now known, from animal te t , t b much
superior to cis-dichlorodiammineplatinum(II).
5. Recent clinical results are described in The Proceedings of
the III International Symposium on Platinum Coordination
Complexes in Cancer Chemotherapy. Published in J. Clin.
Hemato/. Oneo/. 7, Parts 1 and 2 (January 1977).
The manuscript was received 20 February 1977
© Heyden and Son Ltd, 1978
INTERDISCIPLINARY SCIENCE REVIEWS, VOl. 3, NO.2, 1978 147