`
`
`
`
`
`-
`
`Hospira v. Genentech
`Hospira v. Genentech
`IPR2017-00737
`IPR2017—00737
`Genentech Exhibit 2038
`
`Genentech Exhibit 2038
`
`

`

`Principles and Practice
`
`Cancer Chemotherapy and Biotherapy:
`
`

`

`
`
`Cancer Chemotherapy
`and Biotherapy:
`Principles and Practice
`
`SECOND EDITION
`
`
`
`EDITED BY
`
`Bruce A. Chabner, M.D.
`Chief, Hematology/Oncology
`Clinical Director
`
`Massachusetts General Hospital Cancer Center
`Boston, Massachusetts
`
`Dan L. Longo, M.D.
`Director, Biological Response Modifiers Program
`Division of Cancer Treatment
`National Cancer Institute
`
`Frederick, Maryland
`
`
`
`
`n Lippincott - Raven
`
`
`
`Philadelphia - New York
`
`_It______
`
`

`

`
`
`Copyright © 1996 by Lippincott—Raven Publishers. All rights reserved. This book is protected by copy-
`right. No part of it may be reproduced, stored in a retrieval system, or transmitted, in any form or by
`any means—electronic, mechanical, photocopy, recording, or otherwise—without the prior written
`permission ofthe publisher, except for brief quotations embodied in critical articles and reviews. Printed
`in the United States of America. For information write Lippincott—Raven Publishers, 227 East Wash—
`ington Square, Philadelphia, PA 19106.
`
`Library of Congress Cataloging-in-Publications Data
`
`Cancer chemotherapy and biotherapy : principles and practice / edited
`by Bruce A. Chabner, Dan L. Longo -—- 2nd ed.
`p.
`cm.
`Rev. ed. of: Cancer chemotherapy.
`Includes bibliographical references and index.
`ISBN 0—397-514184 (hard : alk. paper)
`1. Cancer—Chemotherapy.
`2. Cancer—Immunotherapy.
`(Dan Louis), 1949—
`.
`II. Cancer chemotherapy.
`[DNLM: 1. Neoplasms—drug therapy.
`2. Biological Products—therapeutic use.
`3. Antineoplastic Agents—therapeutic use. 4. Chabner, Bruce. QZ 267 C21515 1996]
`RC271.CSC32219
`1996
`616.99’4061—dc20
`DNLM/DLC
`for Library of Congress
`
`I. Longo, Dan L
`
`95-38920
`CIP
`
`The material contained in this volume was submitted as previously unpublished material, except in
`the instances in which credit has been given to the source fi'om which some of the illustrative material
`was derived.
`Great care has been taken to maintain the accuracy ofthe information contained in the volume. How—
`ever, neither Lippincott—Raven Publishers nor the editors can be held responsible for errors or for any
`consequences arising from the use of the information herein.
`The authors and publisher have exerted every effort to ensure that drug selection and dosage set forth
`in this text are in accord with current recommendations and practice at the time of publication. How-
`ever, in view of ongoing research, changes in government regulations, and the constant flow of infor-
`mation relating to drug therapy and drug reactions, the reader is urged to check the package insert for
`each drug for any change in indications and dosage and for added warnings and precautions. This is
`particularly important when the recommended agent is a new or infrequently employed drug.
`Materials appearing in this book prepared by individuals as part of their official duties as US.
`Government employees are not covered by the above-mentioned copyright.
`
`987654321
`
`

`

`L553..
`
`Cancer Chemotherapy and Mot/army, second edition,
`edited by Bruce A. Chabner and Dan L. Longo.
`Lippincott—Raven Publishers, Philadelphia ©1996
`
`
`
`_-
`
`_
`
`..
`
`Platinum Analogues
`Eddie Reed, Meenakshee Dabholkor, and Bruce A. ChObner
`
`..
`
`.
`
`
`
`
`
`The antitumor activity ofthe platinum complexes was dis-
`covered as the result ofa fortuitous observation by Rosen-
`berg et all,2 during a study of the effects of electric
`current on growing bacteria. When alternating current
`was delivered through platinum electrodes to a growing
`bacterial culture, the bacterial cells stopped dividing and
`grew into long filaments. The same result was seen when
`an attempt was made to grow the bacteria in fresh
`medium that previously had been subjected to the elec—
`tric current. Because filamentous growth was known to
`occur in bacteria subjected to alkylating agents or radia-
`tion, Rosenberg suspected that an active substance may
`have entered the medium, possibly through the release of
`soluble platinum compounds from the electrodes. It was
`found that platinum was released by electrolysis as hexa—
`chloroplatinate, which, in the presence of ammonium
`salts and light, generated the platinum complex cis-
`Pt(II)(NH3)2C12, which is now known as the anticancer
`drug cisplatin. This agent has become the primary build—
`ing block for regimens that cure patients with testicular
`carcinoma and produces high response rates in patients
`with small cell carcinoma ofthe lung, bladder cancer, and
`ovarian cancer. The key features of cisplatin and its active
`and less nephrotoxic analogue, carboplatin, are shown in
`Tables 14—1 and 14-2.
`
`CHEMISTRY
`
`have fixed bond angles and spacial configuration. The
`spacial configuration depends on whether the oxidation
`state of the platinum is +2 or +4; these oxidation states
`are usually designated as Pt(II) or Pt(IV), respectively. In
`Pt(II) complexes, the platinum atom has four bonds
`directed to the corners of a square at which the four
`ligand atoms are located (Fig. 14-1). In Pt(IV) complexes,
`there are six bonds and six ligands: four in a planar square
`configuration as in Pt(II), plus one directly above and
`one directly below the platinum atom, an octahedral con~
`figuration. Because the bonds are fixed and not readily
`exchangeable, the complexes can have distinct isomers,
`such as cis— and tram-Pt(II)(NH3)2C12 (Fig. 14-1). The
`cis isomer is the antitumor drug cisplatin. The importance
`of steric conformation is highlighted by the fact that the
`trans isomer has virtually no antitumor activity.
`Like covalent bonds in general, the bonds to the plat-
`inum atom tend to have energy barriers that limit the rate
`ofbond formation or dissociation. Some platinum bonds,
`such as those to nitrogen, are essentially irreversible under
`physiologic conditions, while others, such as the chloride-
`platinum(II) bond in cisplatin, are more labile.
`Displacement reactions can occur in Pt(II) complexes
`in which one or both ligands are displaced by a compet-
`ing nucleophile, in analogy to the reactions of alkylating
`agents. Thus the Pt(II) complexes have significant chem-
`ical and biologic similarities to alkylating agents.
`
`Covalent Bond Character
`
`Oxidation States
`
`Platinum is in the third row of transition metals in the
`periodic table and has eight electrons in the outer d shell.
`Palladium and nickel, which occupy analogous positions
`in the second and first transition series, have similar con-
`figurations of outer electrons but little or no antitumor
`activity. Because the platinum atom has a much larger
`total number of electrons, however, the orbitals of its
`- Outer electrons are more polarizable, and bonds that are
`formed from these orbitals have more covalent character.
`The covalent nature of these bonds confers two essen-
`tial properties, stereospecificity and energy barriers. Stereo-
`Schzficity means that the bonds from the platinum atom
`
`Although both Pt(II) and Pt(IV) complexes can have
`potent antitumor activities, it is unclear whether Pt(IV)
`complexes may undergo ligand displacement reactions
`and form covalent bonds under physiologic conditions.
`The activity of Pt(IV) complexes may be due to their
`reduction in vivo to the active Pt(II) state.3
`
`Displacement Reactions
`
`The biologic actions of Pt(II) complexes are due to
`displacement reactions, which cause the platinum to
`become stably bound to DNA, RNA, proteins, or other
`
`357
`
`
`
`@3315“s
`a?“
`
`\AaGiant?i
`"tauaver-.mm
`
`
`
`
`
`..-.;.a.znull-
`
`

`

`358
`
`CANCER CHEMOTHERAPY AND BIOTHERAPY
`
`
`
`Table 14-]. Key Features of Cisplatin
`
`
`
`Metabolism:
`
`Pharmacokinetics:
`
`Elimination:
`
`
`
`ligands such as sulfur. This type of mechanism could
`
`Dose: 20mg/m2/day times 5 days every 3—4 weeks
`Mechanism of action: Covalently binds to DNA bases and disrupts DNA funcu'on; In some settings, toxicity may be related to_drug
`binding to proteins.
`Inactivated intracellularly and in the blood stream by sulfhydryl groups. Drug covalently binds to glutathione,
`metallothioneins, and sulfhydryls on proteins.
`After IV bolus,
`‘1/2a =
`tl/ZB = approx 60 min
`t1fl'y = approx 24 h
`After “high dose,” two: = 30 min
`Approximately 25% of an administered dose is excreted from the body during the first 24 hours. Of that
`portion eliminated, excretion is: renal >90% and bile <10%.
`Systemic thiosulfates (IV thiosulfate, possibly WR—2721, possibly diethyldithiocarbamate) may theoretically
`inactivate drug systemically.
`Renal insufliciency with Mg2+ wasting
`Nausea and vomiting
`Peripheral neuropathy
`. Auditory impairment
`Myelosuppression (thrombocytopenia primarily; WBC suppression seen with carboplatin)
`. Visual impairment (rare)
`. Hypersensitivity (rare)
`. Seizures (rare)
`Use with caution in the presence of other nephrotoxic drugs (aminoglycosides, etc)
`. Monitor electrolytes, Mg“, and Ca2+
`Maintain high urine flow during therapy
`Aggressive premedicau'on with antiemetics is usually helpful (ondansetron or metoclopromide i decadron)
`Full dose can be administered if 24-hour creatinine clearance is >60 ml/h and patient can tolerate vigor-
`ous hydration. If 24-hour creatinine clearance is less than 60 ml/h, carboplatin (also given with vigorous
`hydration) should be considered.
`
`Drug interactions:
`
`Toxicity:
`
`Precautions:
`
`
`
`mesaer°°\I°~sm>s~.~re
`
`alkylating agents such as nitrogen mustards. Ligand
`displacement reactions in platinum complexes occur with
`retention of the spatial configuration of the bond angles,
`in contrast to alkylating agents, where configuration typ-
`ically becomes inverted.
`The stability of the binding of different ligands to
`Pt(II) varies greatly. Although the binding to sulfur or
`nitrogen sites is essentially irreversible under physiologic
`conditions, the stability of binding to halogens is much
`lower, in the order I‘ > Br‘ > Cl‘, and the stability of
`binding to H20 is weaker still.
`The selectivity among different sites of binding, how-
`ever, may be determined more by the relative reaction
`rates than by the equilibrium binding stabilities. The bind-
`ing to sulfhydryl groups generally is very rapid as well as
`very strong. The binding to amines, however, is relatively
`slow, even though it is strong. Binding may, ofcourse, be
`hindered by steric factors. Another important determinant
`of displacement reactions is the rate of dissociation.
`The dissociation of sulfur or nitrogen ligands is gener-
`ally negligible under physiologic conditions (except pos-
`sibly as noted below). The aquo (H20) ligand dissociates
`rapidly, whereas chloride dissociates more slowly. The
`dissociation rate also can be influenced by the identity of
`the other ligands in the complex, especially by the ligand
`located trans to the dissociating ligand. This “kinetic
`trans effect” is exerted most strongly by highly polarizable
`
`critical biomolecules. All the antitumor active complexes
`are bifunctional in that they can form, by successive dis—
`placement reactions, two stable bonds under physiologic
`conditions so as to produce a covalent cross-link between
`two nucleophilic atoms of a macromolecule. In this
`respect, the Pt(II) complexes are similar to bifunctional
`
`
`
`'—
`Table 14-2. Key Features of Corboplafin
`
`sees.
`
`i
`
`3‘82
`ism
`efi4n
`
`_sun-rdf
`
`
`
`Pharmacokinetics:
`
`365 mg/m2 every 3—4 weeks
`Dose:
`Mechanism of action: Covalent binding to DNA
`Metabolism:
`Drug is primarily excreted. Only
`a small fraction is metabolized
`After IV bolus,
`t1not = 12—24 min
`rl/ZB = 1.3—1.7 h
`‘1/2'Y =
`h
`Approximately 90% excreted in
`the urine in 24 h
`
`Elimination:
`
`Toxicity:
`
`Drug interactions:
`
`None described in detail; see
`cisplan'n
`1. Myelosuppression, especially
`thrombocytopenia
`2. Nausea and vomiting
`3. Nephrotoxicity at high doses
`and in patients with prior renal
`dysfunction
`1. Reduce dose in proportion to
`Precautions:
`reductions in creatinine clearance
`
`
`
`
`

`

`Platinum Analogues
`
`Figure 14-]. Cisplatin, the inactive isomer transplatin, and
`other platinum analogues of clinical interest. While most
`compounds have platinum in the (II) valence state, note the
`platinum(lV) derivatives ormaplatin and tetraplatin; also
`note the several analogues that contain a stabilizing cyclo-
`hexane group attached to the platinum molecule.
`
`CI
`
`NH3
`
`CISPLATIN fl
`
`Cl
`
`NH3
`
`
`
`marized in Fig. 14—2. Although it is possible that a chlo—
`ride ligand in cisplatin might in some cases be displaced
`directly in a reaction with a macromolecule, it is generally
`agreed that the more usual path is via an initial aquation
`reaction in which a chloride is replaced by a water mole-
`cule. The aquation reaction is driven by the high concen—
`tration of water and low concentration of chloride in the
`
`tissues. The aquated platinum complex can then react
`rapidly with a variety of strong binding sites.
`The equilibrium reaction in water (Fig. 14-2) may be
`of pharmacologic importance, not only because of the
`high reactivity of the aquated species but also because of
`the different ionic states that may affect the ability of
`particular species to penetrate through lipid membranes.
`The uncharged species,
`including the dichloro- and
`chlorohydroxy complexes, would be expected to pene-
`trate membranes most easily; these species are shown with
`a zero charge exponent in Fig. 14-2.
`The equilibrium and rate constants of the aquation
`reactions (reactions 1 and 2 in Fig. 14-2) are listed in
`Table 14-3. The values listed indicate that for 4.4 mM
`
`cisplatin in water at 37°C, the first aquation occurs with
`a half-time of 2.5 hours and would eventually reach an
`equilibrium ratio of 1:1 for the dichloro:chloroaquo
`species. Because of the slowness of the reaction and the
`presence of reactive constituents other than water, it is
`unlikely that this equilibrium would be approached in
`biologic systems.
`In blood plasma, the high chloride concentration of
`approximately 100 mM would keep cisplatin predomi-
`nantly in the uncharged and relatively unreactive dichloro
`form. This form may react to some degree with sulfhydryl
`groups of plasma proteins. The fi'ee dichloro form could
`enter cells by passive diffusion. In the cytoplasm, the
`relatively low chloride concentration of approximately
`4 mM would favor the aquation reaction, which would
`yield highly reactive species whose ionic charge may
`retard exit from the cell.
`
`Reactions with DNA
`
`Cisplatin has been noted to bind to all DNA bases, but
`in intact DNA there seems to be preferential binding to
`the N—7 positions of guanine and adenine?»5 This may be
`due to the high nucleophilicity of the imidazole ring,
`particularly at the N-7 position.” Cisplatin binds to RNA
`more extensively than to DNA and to DNA more than to
`protein when binding is assessed as moles of drug per
`gram of macromolecule.8
`In the reaction of cisplatin with DNA or other macro-
`molecules, the two chloride ligands can (usually after aqua-
`tion) react with two diEerent sites to produce cross—links.
`The bond distance between the chloride leaving groups of
`cisplatin is fixed at approximately 3 A5 as compared with
`the corresponding bond distance for alkylating agents,
`which can range between 7 and 10 A. A more complete
`description of various aspects of the chemistry of cisplatin
`can be found in several excellent recent reviews.4’9’1°
`
`“JV‘V
`
`G
`
`1
`
`tmflflfifl
`.1.“l»1
`
`
`
`333,};“iii
`
`t
`
`'33.
`is;‘ 74:1
`
`
`
`TRANSPLATINUM
`
`CI
`
`NH3
`
`flNH3
`
`01
`
`CI
`
`NH2\ | / c:
`/Pt\
`1
`C]
`
`Cl
`
`~~NH2
`
`°
`
`""3
`
`o \
`
`‘\3 \
`
`TETRAPLATIN
`
`CARBOPLATIN
`
`PLATINUM-DACH
`
`ORMAPLATIN
`
`OXALIPLATIN
`
`o‘c\
`
`CI
`
`a
`
`NH3
`
`NH2 H
`
`Cl
`
`NH2
`H
`
`NH2
`
`ell
`CE :Pt<
`
`NH;
`
`l
`c:
`
`cu
`
`c:
`
`o
`NH2\ /0 —c/
`/P1\
`l
`° _ C§O
`
`""2
`
`CH3
`
`
`
`
`NHz\m/°
`
`NH2/
`
`\0
`
`0
`
`j
`
`I- 73
`
`c 9
`
`Pcrhaps allow some platinum ligands to be dislodged
`fl0m tight binding sites_under physiologic conditions
`and may account for the biologic difl‘erences of various
`Platinum complexes.
`
`Reactions in Water and Biologic Fluids
`
`The pharmacologic behavior of cisplatin is in part
`determined by its reactions in water, which are sum—
`
`
`
`

`

`360
`
`CANCER CHEMOTHERAPY AND BIOTHERAPY
`
`
`7
`
`Figure 14-2. Aquation and hydrolysis equilibria of cisplatin (pKa values are from ref. 31 3). Note that reactions 3 and 6 are favored
`at physiologic pH and yield products that have a neutral charge and that theoretically could readily cross cell membranes. The rate
`constants tor the initial aquation reactions (1 and 2) are listed in Table 14-3.
`
`
`o
`
`+
`
`H3N\ ,0
`pl
`H3N/
`‘cn
`dichtoro
`
`,ow2
`1
`. H3N\
`‘—
`\CI
`HaN/
`chbtoaqm
`
`P!
`
`3
`
`pK.~7.4
`
`+ +
`
`,OH2
`.__2_.. H3N\
`w—-
`’Pt\
`H3N
`OH2
`
`
`
`
`
`+
`H
`“3” \Pt ’0“ —7-8 “WM/ONR ’NH3
`I \0/ \NH
`dimer
`
`+ +
`
`are consistent with earlier reports from Lippard’s group.
`Using an analysis based on gel electrophoretic mobility,
`reactivity with chemical probes, and molecular mechan-
`ics, Sip et al.13 have demonstrated that a centrally placed
`interstrand cross—link in a 22-mer DNA duplex bends the
`double helix by approximately 55 degrees toward the
`major groove. This results in exposing the cytosine
`residues but does not result in local denaturation within
`the flanking sequences.
`A similarly placed intrastrand d(GpG) adduct bends
`DNA to a lesser extent.14 The frequency of specific inter-
`and intrastrand adducts found in mammalian cells after
`exposure to platinum derivatives is shown in Table 14-4.
`
`
`
`Table 14-4. Types oi DNA Lesions Caused by
`
`Cis latin and Carbo lat'n
`P
`P
`l
`
`PERCENT
`OF TOTAL
`DNA
`DAMAGE
`
`DNA LESION
`
`N7— d(GpG)-intrastrand
`adduct
`
`N7- d(ApG)—intrastrand
`adduct
`
`N7- d(GpoG)—intrastrand
`adduct
`
`~60%
`
`~ 30%
`
`~10%
`
`
`
`N7- d(X)-d(X)-interstrand
`cross-link
`
`~ < 1%
`
`Important in
`gene-specific
`
`repair studies
`
`COMMENT
`
`Possibly lethal
`to cells
`
`Possibly lethal
`to cells
`
`Potential lethality
`unclear
`
`_‘l
`
`
`
`Using x-ray crystallographic techniques, Sherman
`et a1.11 studied the crystal and molecular structure of the
`intrastrand deG diammine-platinum adduct, which
`comprises approximately 60% of total platinum binding
`to DNA following in Vitro or in vivo drug exposures.
`These studies show that when cisplatin is covalently
`bound to the d(GpG) dinucleotide, the conformation of
`the dinucleotides largely depends on the stereochemical
`requirements for maintaining the planar platinum coor-
`dination relationship. Further, it appears that the forma—
`tion of an intramolecuiar hydrogen bond between a
`proton of the amine ligand and an oxygen atom of the
`terminal 51-phosphate contributes to the stabilization of
`the platinum dinucleotide complex.
`Kozelka and Chottard12 studied cisplatin binding to
`defined double-stranded decanucleotides and showed
`that cisplatin causes kinking of the helix by approximately
`60 degrees. The bend occurs toward the major groove
`and is associated with limited unwinding. These findings
`
`
`
`Table “'3' EqUilibfiUm and Rate Constants For
`Aquation oi Cisplafinalz—sis
`
`EQUILIBRIUM
`REACTION CONSTANT, (mM)
`(REFERTO
`FIG. 14-2)
`
`AT 25°C AT 37°C AT 25°C AT 37°C
`
`RATE OF FORWARD
`REACTION(t1/2, H)
`
`
` 1
`
`2
`
`3.6
`0.11
`
`4.4
`0.19
`
`7.7
`5.8
`
`2.5 —i
`
`

`

`Inwith:jgig-gum\aw"
`atiiiHRH
`
`—«‘hé‘!ktnhaltai‘5‘:
`E:-.‘.sWfififiéflmmaa
`
`i-
`
`5%.;
`its;
`
`''Jfi'glé
`'5
`
`Platinum Analogues
`
`METHODS OF MEASUREMENT
`
`Shortly after the observation that cisplatin reacts with
`DNA, methods began to be developed to measure the
`binding of drug to DNA and the production of particu-
`lar types of cisplatin-DNA lesions. There are now several
`methods available for measuring cisplatin-DNA adducts.
`The standard for the quantitation of elemental plat-
`inum is atomic absorption spectrometry (AAS).1° AAS
`utilizes high temperature to atomize molecules within a
`graphite chamber. The chamber is monitored by an opti-
`cal system that reads the absorbance of the atomic cloud
`generated. One metal can be distinguished from another
`by its absorbance at specific wavelengths. Platinum is
`measured at a wavelength of 265.9 nm. Recently, AAS
`instruments have been equipped with optional Zeeman
`background correction for nonspecific absorption.15
`Using Zeeman correction, one can now obtain highly
`precise and sensitive measurements of platinum bound to
`DNA after therapeutic levels of drug exposure.16
`An alternative method to assess cisplatin—induced
`DNA lesions in cells treated at pharmacologically relevant
`dosage is DNA alkaline elution.17 In this method, cellular
`DNA is prelabeled with a radioisotope such as [3H]
`thymidine. After exposure to drug, the cells are deposited
`on a filter surface and lysed with a detergent solution. An
`alkaline solution (pH 12) is then added. The high pH
`disrupts the hydrogen-bonded structure that holds the
`DNA strands together. The alkaline solution is pumped
`slowly through the filter, and the rate of elution of the
`single-stranded DNA is determined. Using appropriate
`modifications of this technique, it is possible to quantify
`DNA interstrand cross-links, DNA-protein cross—links,
`and DNA strand breaks, all ofwhich modulate the rate of
`DNA elution from the filter. The method detects inter-
`
`strand cross—links, which retard elution of DNA, as well
`as DNA strand breaks, which accelerate elution. The
`technique does not detect DNA intrastrand cross-links or
`DNA-drug monoadducts, which can be measured by
`HPLC methodsmil"
`
`A number of ELISA assays can assess drug binding to
`DNA in various settings. The first such assay was devel-
`oped by Poirier and colleagues”23 and has been used to
`Study tissues from human cancer patients as well as tissues
`from experimental animals. It utilizes polyclon‘al antisera
`Clicited in rabbits against an antigen ofintact calf thymus
`DNA modified with cisplatin. Although Poirier’s method
`quantitates an antigen that correlates well with cisplatin
`biologic efiect, it measures only a fiacfion ofthe covalent
`Platinum-DNA complexes measured by AAS.24 Subse-
`quent ELISA assays developed by Fichtinger-Schepman
`and colleagues“ use polyclonal antisera elicited against
`purified,
`isolated cisplatin-DNA adducts, and those
`developed by Lippard and colleagues26 utilize murine
`monoclonal antibodies elicited against cisplatin-modified
`iIltact calfthymus DNA. Fichtinger—Schepman’s assay can
`quantitate cisplatin-DNA lesions only after the DNA has
`
`been degraded and adducts have been purified by HPLC.
`Fichtinger-Schepman et al.27728 performed limited studies
`with material from human cancer patients, and it is
`unclear whether the need to subject isolated DNA to en-
`zymatic degradation and subsequent HPLC purification
`limits the application of this assay to the study of small
`numbers of samples. However, this assay correlates well
`with AAS measurements of plau'num binding to DNA.
`The determination of the nature of the DNA lesions
`
`formed by cisplatin after defined drug exposures was
`made possible by HPLC analyses of cisplatin-modified
`DNA after that DNA was degraded by enzymatic or
`chemical hydrolysis”,19 Generally, HPLC methods can
`readily assess the relative percentages ofspecific platinum-
`DNA lesions and have been used to measure the relative
`
`proportions of the deG and dApG diammineplatinum
`adducts after cisplatin exposures. However, these meth-
`ods do not determine the absolute amount of drug
`bound to DNA.
`
`Terheggen and colleagues”31 have developed an
`immunocytochemical technique that uses microdensi-
`tometry to assess relative cisplatin binding to DNA in
`rodent and in human tissues. Their assay utilizes a poly-
`clonal antiserum elicited by the method of Poirier. A
`“double peroxidase” approach is used to visualize the
`binding sites of the antiserum on the DNA. With this
`method, as few as several hundred cells are required to
`obtain an assessment ofthe relative level of cisplatin bind-
`ing to DNA, a threshold that is 5 to 6 logs less than the
`number of cells needed in the assays of Poitier and
`Fichtinger-Schepman. It is currently not clear if this assay
`correlates with AAS determinations. Others have adopted
`these methods to measure gene-specific repair.32
`Sancar and colleagues“,34 used the uvrABC excision
`nuclease complex from Escherichia 5011'. Such a method
`may prove useful in determining the persistence of cis-
`platin-DNA adducts in viable tissues of human cancer
`patients months after drug exposure.35
`Taq polymerase, used in PCR reactions to amplify
`DNA, is inhibited by cisplatin DNA adducts and can de-
`tect as few as one adduct per 2000 base pairs?”8 This
`sensitivity may not be adequate to measure adducts in
`clinical
`specimens,
`which
`reach
`levels
`of
`1 adduct in 10,000 to 100,000 base pairs.
`
`MECHANISM OF ACTION
`
`Most data are consistent with the hypothesis that the
`major cytotoxic target of cisplatin is DNA. The types of
`DNA lesions primarily responsible for the cytotoxicity
`and antitumor activity, however, are not clearly estab-
`lished. Analogues of cisplatin that are tumoricidal in
`experimental systems have chemical leaving groups in the
`cis configuration and have effects on DNA that are in
`most respects similar to those of the parent compound.
`
`
`
`

`

`362
`
`CANCER CHEMOTHERAPY AND BIOTHERAPY
`
`The cytotoxicity ofcisplau'n against cells in culture cor-
`relates closely with total platinum binding to DNA)",39 to
`interstrand cross-links,17’4° and to the formation of intra-
`strand bidentate N-7 adducts at d(GpG) and d(ApGYML42
`(Fig. 14-3). The relative percentages of these lesions have
`been studied after treatment of isolated DNA with cis-
`platin,“"19 as well as after treatment ofmalignant and non-
`malignant rodent cells”,43 and after treatment of human
`cancer patients.“44 HPLC analysis of degraded DNA
`indicates that the d(GpG) lesion makes up 60% to 70% of
`total platinum binding to DNA, the d(ApG) lesion is 20%
`to 30% of total platinum binding, and the remaining 10%
`is a mixture of various intrastrand d(GpoG) adducts, in—
`terstrand cross—links, and monoadducts.“"19’45 Analogous
`studies have been carried out in which intact cells were
`treated with cisplatin. In L1210 murine leukemia cells43
`and in Chinese hamster ovary cells,25 the relative percent-
`ages of adducts measured by HPLC analysis approximate
`the following:
`interstrand cross-links,
`less
`than 1%;
`d(GpG) intrastrand adducts, 60%; d(ApG) adducts, 30%.
`Composing the remaining 10% are a variety of DNA
`lesions, including DNA—protein cross-links, DNA mono-
`adducts, and d(GpoG) and d(ApoG) intrastrand
`adducts. In human leukocyte DNA after a Cisplatin intra-
`venous infusion?!"44 the relative percentages of the
`Cisplatin-DNA lesions formed in human leukocyte DNA
`were the same as those reported for L1210 cells, Chinese
`hamster ovary cells, and salmon sperm DNA.
`
`Several recent reports have focused on the inhibi-
`tion of DNA replication and specific enzymatic activities
`by different Cisplatin adducts. Pinto and Lippard45 com-
`pared the ability of Cisplatin and transplatinum to inhibit
`DNA synthesis using a single-stranded DNA template and
`DNA polymerase I. Cisplatin was much more effective at
`inhibiting DNA replication than was the trams-Ptal) iso-
`mer in this system. A d(GpG) adduct was very effective in
`inhibiting replication of DNA polymerase I.
`Intrastrand d(GpG)-Cisplatin adducts cause a marked
`distortion of DNA conformation. Lippard and col-
`leagues‘“48 have described two relevant alterations. In
`one study, it was shown that the d( GpG)-cisplatin lesion
`disrupts the local structure of the DNA, allowing the
`cytosine bases that normally are paired to guanines to
`become partially or totally unpaired, as indicated by their
`becoming accessible to binding by an antiserum that
`recognizes unpaired DNA bases.47 This disruption of
`structure requires bidentate binding by platinum because
`it did not occur in DNA treated with a monodentate plat-
`inum analogue. In another set of studies, Lippard’s group
`has shown that the intrastrand adduct introduces a sharp
`bend in the helix axis of the DNA and that this is a con-
`sequence of the geometry of the cis—Pt(II) complex
`bound to two adjacent guanine-N-7 positions.48 Lippard
`suggested that these alterations in local DNA structure
`that are caused by a Cisplatin intrastrand adduct may be
`responsible for the cytotoxic effect.49
`
`
`
`
`
`PM“
`t, “w
`M;
`Film1 : 5M"'ttlrr
`must?
`
`“nma
`
`
`
`-Nsa-d’l£451
`
`Figure 14-3. Biiunctional adducts of Cisplatin with DNA. Lesions indicated in panels A, B, and C represent different intrastrand
`adducts, which together account For >90% of total platinum binding to DNA. The lesion indicated in panel D is the interstrand cross-
`link measured by alkaline elution and accounts for less than 5% of total platinum binding to DNA. See text For discussion.
`
`d(GpG) Adduct
`
`d(ApG) Adduct
`
`H3N
`
`NH3
`\ /
`Pt
`
`/\
`5'— A—G _3'
`O
`O
`
`3'— T—C —5’
`B
`
`lnterstrand Crosslink
`
`H3N
`
`NH3
`
`d(GpoG) Adduct
`
`NH3
`
`HEN
`\W/’
`5'— G{x>e— 3'
`o
`o
`.
`
`3,—C—X—C—5'
`C
`
`
`
`
`

`

`Platinum Analogues
`
`When cisplatin binds to cellular DNA and/or RNA,
`such lesions may inhibit the fiinction of a number of
`enzymatic processes important to the cell. Such functions
`include inhibition of SI nuclease (an endonuclease spe-
`cific for single—stranded DNA)“; inhibition of restriction
`enzymes“; inhibition of DNA polymerase 1”; inhibition
`of RNA polymerase“; and inhibition of RNA transla-
`tion“,55 Synthesis of new DNA or RNA can proceed
`across cisplatin intrastrand cross—links, as mediated by
`DNA and RNA polymerases, respectively,“ although
`only 10% of adducts are bypassed. The d(GpG) lesion is
`the most inhibitory. In E. coli, the intrastrand d(ApG)
`adduct is five times more mutagenic than d(GpG).57 In
`this system the d(ApG) may not be an effective blocking
`lesion for either DNA or RNA synthesis.53 Guanine-rich
`sequences appear to be preferred targets for cisplatin
`attack and mutation.58 Interestingly, the type and extent
`of helical unwinding and bending are similar for the two
`intrastrand adducts.59
`The tmm—Pt(II) isomer of cisplatin is capable of bind-
`ing and cross-linking DNA but has much lower cytotox—
`icity and no antitumor activity. Many studies have inves-
`tigated the origin of the biologic difference between the
`cis— and tmm—Pt(II) isomers to elucidate the mechanism
`of action of cisplatin. Although both isomers produce in-
`terstrand cross-links in purified DNA, only the cis isomer
`(cisplatin) produces substantial interstrand cross-linking
`in mammalian cells.”40 The trans isomer, however,
`produces more DNA-protein cross-links than does cis-
`platin. This argues against DNA-protein cross-links as the
`primary cytotoxic lesions; however, this conclusion is not
`definitive because of possible differences between the cis
`and trans isomers in regard to the nature of the proteins
`bound and the locations of their binding in the genome.
`Ciccarelli et al.‘50 compared the actions of cisplatin and
`its trans isomer in SV40-infected green monkey CV—l
`cells. In this system, the trans isomer was much less
`potent than cisplatin in generating Pt adducts of SV40
`DNA. At equal extracellular concentrations, much less
`trans than cis isomer was bound to the viral DNA isolated
`
`from treated cells. The inhibition of SV40 DNA replica-
`tion, when gauged on the basis of amount of Pt bound to
`the DNA, was reported to be nearly equal for the two iso-
`mers. These data were interpreted as indicating a rapid
`removal of bound tram-Pt from the DNA oftreated cells,
`in contrast to little or no such repair in the case of
`cisplatin. Roberts and Friedlos,61 however, were not able
`to confirm the proposed rapid repair of tram-Pt DNA
`adducts and attributed the apparent
`reduction in
`I)t/DNA ratio not to a removal of Pt but to continued
`DNA synthesis in the trans—Pt treated cells.
`The reason for the high cytotoxicity of cisplatin and
`10W cytotoxicity of trans-PtaI) thus is still unclear. It may
`be that the interstrand cross-links, even though they con-
`stitute only a small fraction of cisplatin DNA adducts, are
`the major cytotoxic lesions because the trans isomer pro-
`duces virtually no interstrand cross-links in mammalian
`Cells. A second possibility is that the d(GpG) or d(ApG)
`
`intrastrand cross-links produced by cisplatin are much
`more cytotoxic than are the d(GpoG) intrastrand cross-
`links produced by the trans isomer. Tram-Ptal), because
`of its particular geometry, cannot form intrastrand cross-
`links between adjacent bases, but it can form intrastrand
`cross—links in which one base is skipped, as in d(GpoG).
`A third possibility is that cisplatin produces an especially
`cytotoxic type of DNA-protein cross—link different fi'om
`the trans isomer.
`
`Because cisplatin could react avidly with many accessi-
`ble sulfur and nitrogen sites on a variety of proteins, it is
`difficult to evaluate the possible importance of non-DNA
`targets in the mechanism of antitumor action or mecha-
`nism of toxicity. Platinum complexes bind extensively to
`sulfliydryl groups of plasma proteins, and it has been hy-
`pothesized that ligand exchange reactions with sulfhydryl
`groups of critical enzymes may be responsible for to