he Basic
`Science of
`
`IPR2017-01122
`
`Genentech 2137
`Genentech 2137 |
`Celltrion v. Genentech
`Celltrion v. Genentech
`IPR2017-01122
`
`

`

`The Basic Science of
`Oncology
`
`Second Edition
`
`Editors
`lan F. Tannock, M.D., Ph.D.
`Richard P. Hill, Ph.D.
`
`The Ontario Cancer Institute
`and
`The University of Toronto
`Toronto, Ontario
`Canada
`
`McGraw-Hill, Inc.
`Health Professions Division
`
`San Francisco Auckland
`St.Louis
`New York
`Bogota Caracas
`Lisbon
`London Madrid
`Mexico Milan Montreal New Delhi
`Paris
`San Juan
`Singapore
`Sydney
`Tokyo
`Toronto
`
`
`
`

`

`
`
`THE BASIC SCIENCE OF ONCOLOGY
`Copyright © 1992 by McGraw-Hill, Inc. All rights reserved.
`Printed in the United States of America. Except as permitted un-
`der the United States Copyright Act of 1976, no part of this pub-
`lication may be reproduced or distributed in any form or by any
`means, orstored in a data base orretrieval system, without the
`prior written permission of the publisher.
`
`1234567890 MAL MAL98765432
`
`This book was set in Baskerville by Beljan, Ltd.
`The editors were Jane E. Pennington and Lester A. Sheinis.
`The production supervisor was Roger Kasunic.
`The cover was designed by Carol Matsuyama.
`Malloy Lithographing, Inc., was printer and binder.
`
`ISBN 0-07-105407-3
`
`Library of Congress Cataloging-in-Publication Data
`
`The Basic science of oncology / edited by Ian F. Tannock, Richard
`P. Hill.
`. cm.
`Includes bibliographical references and index.
`ISBN 0-07-105407-3
`1. Cancer.
`2. Oncology.
`Richard P.
`[DNLM: 1. Neoplasms. QZ 200 B3115]
`RC261.B37 1992
`616.99’ 4—dc20
`DNLM/DLC
`for Library of Congress
`
`92-2884CIP
`
`I. Tannock, Ian F.
`
`Il. Hill,
`
`

`

`zs
`
`BOO
`Bal!
`oeek
`
`:
`
`||
`
`)
`
`|
`
`|
`|
`|
`
`|
`
`|
`
`|
`
`Contents
`
`About the Editors and Contributors hee GaSeEe ei
`
`Preface 2.0 eee bee be eee bebe eee een bub ybbbnbbb LEE EEE bbb EER EGS ix
`
`1. ‘Introduction: Cancer as a Cellular Discasesvvssar uns suger ners eee
`Richard P. Hill and Ian F. Tannock
`
`PART 1: CANCER CAUSATION
`
`4, Epidemiology: of Canceraaniesveiaie eeee
`Norman F. Boyd
`
`3. Methods of Genetic Analysis. 00.0000 cee ence een e ee es 23
`Jeremy Squire and Robert A. Phillips
`
`4. Geneétre ‘Basis of Cancereasusianeereewse sey or ee eA
`Jeremy Squire and Robert A. Phillips
`
`5. Oncogene«2...en ee eee eee ee ee ees 61
`Mark D. Minden and Anthony J. Pawson
`
`6. Viruses and: Camcmersrccsera cc 0
`Sam Benchimol
`
`i 5Baa 88
`
`Te
`
`(Gheimical CareiieSEESIS ecco wane osi ane eecsauecsecr qm amine: cecame ese as mem ucelucerce scceuatien sameeren ns 102
`Michael C’. Archer
`a-admition Carcinogenesis nese ieee Wea OE el 1D
`ff A. Michael Rauth
`
`PART 2: CANCER BIOLOGY
`
`®.. Propertics of Malignent Odlescsesecssnasr en ea) Ra eee ee
`Ronald N. Buick and Ian F. Tannock
`
`LO),
`
`Ae Dis
`
`lan F. Tannock
`
`(EZRESPOMTetreHOM fore
`
`cies
`
`cca ceecarcasnerpunanse
`
`cau
`
`ever aesesntg ese citupse tesa epee cetera eerie erent tems
`
`iet
`
`actor
`
`aeraS eo CNET 154
`
`BVASRSUaSS eS Sea acca i Sea pe wie Aa alse tte auc La O
`Richard P. Hill
`
`U2)
`
`SEaraeas" IA earnbeaSa oe sacszer
`Aaron Malkin
`
`perce sseereeaa
`
`npenasscea
`
`gm atcse cata gap ALCS Sar
`
`ae AES TESLA SIS TT STE 196
`
`13; Horriones: and, Catacer'yaisigcicte im csce, goo boase asad wea} gumsmomsracimemsnsmepcmonsensm t@eedumym nein
`DonaldJ. Sutherland and Betty G. Mobbs
`
`smcmcancnsacmonamemeane 207
`
`
`
`

`

`
`
`vi
`
`CONTENTS
`
`14,
`
`Immunology and Immunotherapy of Cancer: pura veepey epee eS eee eee
`Richard G. Miller and Ian F. Tannock
`
`ee 232
`
`PART 3: BIOLOGY UNDERLYING CANCER TREATMENT
`
`Cellular Basis of Radiotherapy.........
`Richard P. Hill
`
`. 259
`
`Experimental Radiotherapy sseeeess eae ee RS OE
`Richard P. Hill
`
`276
`
`Miblopical Properties of Agienmoer TMi 10ers caesar ee RE ET
`lan F. Tannock
`Pheomacology of Anvilcanmes Dreppececscsusesrenanocsnen eu un sv aRUaN RW CEES
`Charles Erlichman
`
`302
`
`317
`
`Eepernmential Chemiethenany sce. can anes eek eae ee RRS
`lan F. Tannock
`
`338
`
`Hyperthennuia: and Photadynamic UhetapyMaia
`Richard P. Hill
`
`Guide to Studies of Diagnostic Tests, Prognosis, and Treatment................
`Norman F. Boyd
`
`155
`
`16.
`
`ihe
`
`18.
`
`19),
`
`20.
`
`21,
`
`Glossary eee a aS
`
`|Xe
`
`403
`
`360
`
`So
`
`O90
`
`

`

`
`
`19
`
`Experimental Chemotherapy
`
`lan F. Tannock
`
`
`
`19.1 INTRODUCTION
`19.1.1 Current Status of Chemotherapy
`19.1.2 Therapeutic Index
`19.1.3 Relationship Between Tumor Remission and Cure
`
`19.2 DRUG RESISTANCE
`19.2.1 Mechanisms of Drug Resistance
`19.2.2 Genetic Basis of Drug Resistance
`19.2.3 Resistance to Methotrexate
`19.2.4 Multiple Drug Resistance
`19.2.5 Topoisomerases and Drug Resistance
`19.2.6 Glutathione and Drug Resistance
`
`19.3.3 Directed Drug Delivery
`19.3.4 Drugs That Cause Differentiation
`19.3.5 Anti-angiogenesis
`19.3.6 Drugs for Hypoxic and Nutrient-deprived Cells
`19.3.7 Higli-Dose Chemotherapy with Bone-Marrow
`Transplantation
`
`19.4 TREATMENT WITH MULTIPLE AGENTS
`19.4.1 Influence on Therapeutic Index
`19.4.2 Synergy and Additivity
`19.4.3 Growth Factors and Other Modifiers of Drug Activity
`19.4.4 Radiation and Drugs
`
`19.3 EXPERIMENTAL APPROACHES TO IMPROVEMENT
`OF CHEMOTHERAPY
`19.3.1 Adjuvant Chemotherapy
`REFERENCES
`19.3.2 Antimetastatic Drugs
`
`
`19.5 SUMMARY
`
`19.1 INTRODUCTION
`
`19.1.1 Current Status of Chemotherapy
`
`In currentclinical practice chemotherapy is used in the
`following ways: (a) as the major curative modality for
`a few rare types of malignancies, such as Hodgkin’s dis-
`ease and other lymphomas, acute leukemia in children,
`and testicular cancer in men;(b)as palliative treatment
`for many types of advanced cancers; (c) as adjuvant
`treatmentbefore or after local treatment (usually sur-
`gery) for primary disease, with the aim of eradicating
`occult micrometastases; and (d) in combination with
`other modalities (usually radiotherapy) in an attempt to
`improve their therapeutic effects.
`Improvements in clinical chemotherapy have de-
`pended largely on the use of drugs in combination.
`Some drugs have been combined because there is a the-
`oretical or experimental basis for expecting synergistic
`interaction through their known mechanismsof action
`either at the molecularlevel or because of their comple-
`mentary effects on cell-cycle kinetics. Synergy does not
`lead, however, to therapeutic benefit unless the interac-
`tion between drugs is tumorspecific (section 19.4.1).
`
`The most important factors underlying the successful
`use of drugs in combinationare (a) the ability to com-
`bine drugsat close to full tolerated doses with additive
`effects against tumorsandless than additivetoxicities to
`normal tissues, and (b) the expectation that drug com-
`binations will include at least one drug to which the tu-
`moris sensitive.
`
`19.1.2 Therapeutic Index
`
`All anticancer drugs havetoxicity as well as antitumor
`effects, and toxicity to normal tissues limits the dose of
`drugs that can be given to patients. Therelationship be
`tween probability ofa biological effect of a drug and ad-
`ministered doseis usually described by a sigmoid curve
`(Fig. 19.1). If the drugis to be useful, the curve describ-
`ing probability of antitumoreffect (e.g., complete clin-
`ical remission) must be displaced toward lower doses 8
`comparedto the curve describing probability of major
`toxicity to normal tissues (e.g., myelosuppression lead-
`ing to infection). Therapeutic index (or therapeutic 14”
`tio) may be defined from such curvesasthe ratio of the
`doses required to produce a given probability of tox
`
`338
`
`uffr
`
`Wie
`
`ekJui
`
`

`

`EXPERIMENTAL CHEMOTHERAPY
`
`339
`
`
` a) Anti-tumor,
`
`effect b) Normal tissue
`
`| |
`
`1.0
`
`"
`
`a “surgically confirmed complete remission” may be
`compatible with the presence of a large numberof tu-
`morcells. Tumor cure requires eradication of all tumor
`cells that have the capacity for tumor regeneration. The
`2
`toxicity
`
`proportion of such stem cells amongthose of the tumor
`@ 05
`population is unknown(see chapter9, section 9.2), but
`=
`Therapeutic index
`
`a
`|
`_
`TD-05
`clinical and even surgically confirmed complete remis-
`||=©9)-Ep-50
`sions are compatible with the presence ofa substantial
`
`residual population of tumorstem cells. Attainment of
`f
`complete remission is but a small step toward tumor
`cure.
`ED-50 TD-05 TD-50
`Dose of Drug
`
`For many drugstherelationship between cell survival
`and doseis close to exponential, so that a constantfrac-
`tion ofthe cells (rather than a constant number)is killed
`by a given doseof drug (chapter 17, section 17.2.4).
`Drugsare usually given in sequential courses, with dos-
`age and schedule limited by normal-tissue tolerance.
`Somerepopulation of tumorcells may take place be-
`tween courses, so that the numberof tumorcells in a
`drug-sensitive tumor may change with time during a
`course of chemotherapy, asillustrated in Figure 19.2.
`In this example, each course ofdrug kills 90% ofthe tu-
`morcells, and starting from a large (100 g) tumor, com-
`plete clinical remission is achieved after three courses.
`Note thata furthersix to ten courses (depending on the
`prevalence of tumorstem cells) would be required to
`achieve cure. Realization of the need to continue aggres-
`sive treatment during complete remission, demonstrated
`originally by the experimental work of Skipper and his
`
`Figure 19.1. Schematic relationships between dose of a drug
`and a, the probability of a given measure of antitumoreffect,
`andb, the probability of a given measure of normal-tissue tox-
`icity. The therapeutic index might be defined as the ratio of
`doses to give 50% probabilities of normal-tissue damage and
`antitumoreffects. However,if the endpoint for toxicity is se-
`vere (e.g., sepsis due to bone-marrow suppression), it would
`be more appropriate to define the therapeutic index at a
`lowerprobability of toxicity (e.g., TD-05/ED-50).
`
`icity and antitumor-effect. Therapeutic index in Fig-
`ure 19.1 might be represented by the ratio of the 5%
`level of probability of severe toxicity (sometimesreferred
`to as Toxic Dose-05 or TD-05), and the 50% probabil-
`ity of antitumoreffect(i.e., effective dose 50 or ED-50).
`Anystated levels of probability might be used, and the
`appropriate endpoints of tumor response andtoxicity
`will depend onthelimiting toxicity of the drug,the in-
`tent of treatment(i.e., cure versus palliation), and on
`whethertreatmentis given to a patient or an experimen-
`tal animal. However, dose-response curves similar to
`those of Figure 19.1 have been defined rarely for drug
`effects in man.
`Improvementin the therapeutic index is the goal of
`experimental chemotherapy. The concept emphasizes
`that any modification in treatment that leads to in-
`creased killing of tumorcells in tissue culture or animals
`must be assessedforits effects on critical normal tissues’
`prior to therapeutictrials.
`
`19.1.3 Relationship Between Tumor
`Remission and Cure
`
`For most solid tumors the limit of clinical and/orradio-
`logic detection is about 1 g of tissue (~10° cells). If
`therapy can reduce the numberof malignantcells below
`this limit of detection, the patient will be described as
`being in complete clinical remission. Surgical biopsy of
`sites that were known to be previously involved with tu-
`mor maylowerthelimit of detection, but a pathologist
`is unlikely to detect sporadic tumorcells presentat a fre-
`quencyofless than 1 in 1000 normal cells, so that even
`
`
`
`Numberoftumorcells
`
`
` Limitofclinical detection —
`
`
`“Complete remission’
`
`
`
`3Tumorweight(g/
`
`lo
`
`10° Treatments t
`
`
`}
`Response
`
`t
`
`}
`t
`Drug resistance
`
`
`
`
`r
`
`Tumor,
`cur
`
`Time (months)
`
`Figure 19.2.Illustration of the relationship between tumor
`remission and cure. In this hypothetical example, treatment
`of a human tumorstarts whenit has 10"cells (~100 g), and
`each treatment, given at monthlyintervals, kills 90% of the
`cells present. This course of therapy leads to complete disap-
`pearanceofclinical tumor. Drugresistance then develops, and
`the tumor grows despite continued treatment. Note that
`despite the attainment of a complete clinical response there
`are always at least 10° viable cells present, and that the
`reduction in cell numbersis small compared to that required
`
`for cure.
`
`

`

`
`
`340
`
`THE BASIC SCIENCE OF ONCOLOGY
`
`led to success in the treatment of
`colleagues (1964),
`acute lymphoblastic leukemia in children, and subse-
`quently to cures in other tumors such as lymphomas.
`Unfortunately, for most solid tumors, a drug-resistant
`subpopulation emerges andleads to relapse, as shown
`in Figure 19.2.
`
`19.2 DRUG RESISTANCE
`
`19.2.1 Mechanisms of Drug Resistance
`
`Manytypesof cancer that occur commonly in man (e.g.,
`colon cancer, lung cancer other than small-cell type)
`show only infrequent responsesto treatment with anti-
`cancer drugs. This resistance to chemotherapy may be
`influenced by such factorsas the proliferative state of
`the cells (chapter 17, section 17.2) and vascular access
`and penetration ofdrugs into tissue (section 19.3.6), but
`the most importantfactor is the intrinsic resistance of the
`tumorcells to available anticancer drugs. Other human
`tumors (e.g., breast cancer or small-cell cancer of the
`lung) often respondto initial treatment, but acquired re-
`sistance to further therapy usually prevents drugtreat-
`ment from being curative. Thusintrinsic and acquired
`drug resistance are the major factorsthat limit the suc-
`cessful use of chemotherapy. A wide range of metabolic
`or structural properties of cells may lead to drug re-
`sistance, and some of the mechanisms underlying in-
`trinsic or acquired drug resistance are summarized in
`Table 19.1.
`Alkylating agents and cisplatin cause cellular dam-
`age by binding with DNA, leadingto cross-linkages
`and breaks in DNAstrands(chapter 18, section 18.3).
`Cells may be resistant to these drugs through a num-
`ber of mechanisms, including decreased cellular up-
`take, reduced drug activation, binding of alkylating
`species by sulfhydryl compoundssuch as glutathione
`(section 19.2.6), increased removal of drug adducts
`from DNA,andincreased repair of DNA (see chap-
`ter 4, section 4.4 and chapter 17, section 17.2). Cross-
`resistance to several chemically unrelated drugs has
`been observed for naturally occurring and semisynthetic
`compoundssuch as doxorubicin and etoposide, due to
`common mechanisms of stimulated drug efflux from
`cells (section 19.2.4), or to decreasedactivity of the en-
`zyme topoisomerase II that allows conformational
`changes in DNA(section 19.2.5).
`‘A number of mechanisms mayalso lead to resistance
`to antimetabolite drugs. These mechanisms include im-
`paired drug transport into cells, overproductionor re-
`duced affinity of the molecular target, stimulation of
`alternative biochemical pathways, and impaired activa-
`tion or increased catabolism of the drug. The most
`widely studied of the antimetabolites is methotrexate,
`
`Ao
`
`Table 19.1. Probable Mechanisms Associated with
`Resistance to Some Commonly Used
`Anticancer Drugs
`
`
`
`Mechanism Drugs
`
`Increase in proficiency Alkylating agents, cisplatin
`of repair of DNA
`Decrease in cellular
`uptake or increase in
`efflux of drugs
`
`Cisplatin, doxorubicin, etoposide
`melphalan, 6-mercaptopurine,
`methotrexate, nitrogen mustard,
`vinblastine, vincristine
`Methotrexate
`
`Increase in levels of
`“target” enzyme
`Alterations in “target”
`enzyme
`Decrease in drug
`activation
`
`5-Fluorouracil, 6-mercaptopurine,
`methotrexate, 6-thioguanine
`Cytosine arabinoside, doxorubicin,
`5-fluorouracil, 6-mercaptopurine,
`6-thioguanine
`Bleomycin, cytosine arabinoside,
`6-mercaptopurine
`Cytosine arabinoside
`
`Increase in drug
`degradation
`Alternative bio-
`chemical pathways
`Inactivation of active
`metabolites by binding
`to sulfhydryl
`compounds
`Amsacrine, doxorubicin, etoposide
`Decrease activity of
`topoisomerase IT
`
`Alkylating agents, cisplatin,
`doxorubicin
`
`and the varied mechanismsofresistance to this agent
`will be described as an example for other antimetabo-
`lites in section 19.2.3.
`
`19.2.2 Genetic Basis of Drug Resistance
`
`Thefollowing evidence suggests that manytypes of drug
`resistance are genetic in origin.
`
`1. Characteristics of drug-resistantcells (i.e., their phe-
`notypes) are often stably inherited in the absence of
`the selecting drug.
`2. Drug-resistant cells are spontaneously generated
`with a rate thatis consistent with knownrates of ge-
`netic mutation.
`!
`3. Generation of drug-resistantcells is increased by €%-
`posure to compounds(e.g., ethyl methane sulfonate)
`that induce mutation. This property has been used
`to generate andselect a large number of drug-resis-
`tant mutants that have been usedto study drug-resis-
`tant phenotypes (Fig. 19.3).
`4. Altered gene products have beenidentified in some
`drug-resistantcells, and someof the genes have bee®
`cloned and sequenced.
`
`

`

`
`
`EXPERIMENTAL CHEMOTHERAPY
`EXPERIMENTAL CHEMOTHERAPY
`
`341
`
`———_—
`
`Mutagen LD MutagenDD DP
`
`—_— —
`
`Chinese Hamster
`Ovary Cells
`
`Culture in low
`concentration of
`drug
`
`Culture surviving
`cells in intermediate
`concentration of
`drug
`
`Culture surviving
`cells in high concentration
`of drug
`
`Setect clones
`
`|
`
`Selectclones
`
`|
`
`|
`
`
`Cells with increased levels of drug resistance
`
`Figure 19.3. General method used for stepwise selection of drug-resistance mutant
`cells.
`
`5. Some drug-resistant phenotypes have been trans-
`ferred to drug-sensitive cells by transfer of genes,
`using techniques described in chapter3, section 3.3.9.
`
`Atleast two types of genetic lesion may convey drug
`resistance: point mutation and gene amplification. Drug
`resistance due to point mutation may be expected to oc-
`cur in a single step that is not critically dependent on
`the concentration of drug usedto select for resistance,
`and the degree of resistance should be quite stable. One
`of many well-characterized examples is mutation in the
`gene that codes for the enzyme HGPRT(hypoxanthine
`guanine phosphoribosyl transferase), which is neces-
`sary for activation of the drug 6-thioguanine. In con-
`trast to point mutation, gene amplification occurs in a
`stepwise manner, and drug-resistant cells are selected
`more rapidly by exposingcells to graded increases in the
`concentration ofthe selecting drug. The best-character-
`ized examples (Schimke, 1984; Bradley et al., 1988) are
`amplification of the gene encoding dihydrofolate reduc-
`tase (DHFR), the target enzyme of methotrexate (sec-
`tion 19.2.3), and amplification of the mdr genes that
`encode P-glycoprotein (section 19.2.4).
`The frequency of occurrence of drug-resistantcells in
`a cell culture or tumor population may be estimated by
`plating the cells after exposure to the drug at a pharma-
`cologically achievable concentration and counting the
`resultant colonies. Typical spontaneous frequencies for
`the incidence of a variety of stable drug-resistant phe-
`notypesare in the range 10-’-10~°. These frequencies
`are consistent with origin by spontaneous mutation, and
`definitive evidence for this origin has been obtained for
`a few drugs by performing a fluctuation test —a test de-
`vised by Luria and Delbruck (1943) to distinguish be-
`tween (a) mutation and selection and (b) induction by
`a drug, as the major cause of bacteriophageresistance
`
`in bacteria. The fluctuation test has been performed
`using only a few drugs (e.g., ouabain, L-asparaginase,
`and hydroxyurea), but has established mutation and se-
`lection as one major mechanism for generation of drug-
`resistant mutants.
`At least two other mechanisms might contribute to
`the generation of genetically based drug resistance.
`Manyanticancer drugs (especially alkylating agents) are
`themselves mutagenic: treatment with such drugs may
`be expected to increase the rate of generation of mutant
`cells and to accelerate the developmentofresistance to
`themselves and to other drugs. Also, drugs that cause
`cell-cycle-progression delay during DNA synthesis may
`allow an increased rate of gene amplification. Thus,
`treatment with some drugs maystimulate the develop-
`ment of drug resistance.
`Second, thereis evidence that different populations of
`cells may interact to influence the drugsensitivities of
`each other (e.g., Heppner, 1984). This influence may
`involve sharing of drug metabolites or increased activa-
`tion of a drug by oneofthecell populations. Cell fusion,
`the transfer of DNA between cells, or the uptake of
`DNAfrom dead cells are also potential mechanismsthat
`could lead to horizontal transmission of drug resistance
`between mammalian cells.
`Generation of drug-resistant mutants amongcells in
`human tumors has implications for planning optimal
`chemotherapy. Goldie and Coldman (1984) have dem-
`onstrated that the probability of there being at least one
`drug-resistantcell in a tumor population is dependent
`on tumor size (Fig. 19.4). This probability increases
`from near zero to near unity over a small range of tu-
`morsizes (~6 doublings) with the critical size depend-
`ing on the rate of mutation to drug resistance. This
`effect, and others, imply a greater chanceofcure if ther-
`apy is begun early, when only microscopicfoci of tumor
`
`
`
`

`

`
`
`342
`
`THE BASIC SCIENCE OF ONCOLOGY
`
`oON Probabilityofatleastone
`
`drug-resisantcell oa
`
`fo
`‘waa rate
`
`=10°6
`
`——®
`
`Oo
`
`3
`
`°@
`
`°a
`
`2°B
`
`ae=:
`
`i
`
`10*
`
`107
`Numberofcells in tumor
`
`10°
`
`10°
`
`Figure 19.4. Probability that there will be at least one drug-
`resistant cell in a tumor containing varying numbers of cells,
`based on rates of mutation of 10-® (open symbols) and 107+
`(closed symbols) percell per generation. Note that this prob-
`ability increases from low to high values overa relatively short
`period in thelife history of the tumorand that drug-resistant
`cells are likely to be established prior to clinical detection.
`(Adapted from Goldie and Coldman, 1984.)
`
`cells are present (see section 19.3.1). The Goldie-
`Coldman modelalso predicts a better therapeutic effect
`when two equally effective and non-cross-resistant
`drugs are alternated, rather than given sequentially,
`since this minimizes the emergenceofcell populations
`that are resistant to both drugs.
`Although drug-resistant phenotypes in many types of
`cell have been shownto be due to mutation or amplifi-
`cation of genes, this could be due in part to methods
`used for selection of drug-resistant cells. Exposure of
`cells to mutagens, followed byselection in high concen-
`trations of drug (Fig. 19.3) may predisposeto selection
`of cells with genetically based drug resistance. Exposure
`ofcells to lower concentrations of drugs, without prior
`exposure to mutagens, often leads to cells that show
`transient resistance to drugs; indeed, transient resis-
`tance of somecells in the population may occur spon-
`taneously, without prior drug exposure (Cillo et al.,
`1989). Mechanisms underlying drugresistance that is
`unstable may includethe transient amplification ofgenes
`or changesin patterns of DNA methylation or of other
`factors that influence gene expression (sometimesre-
`ferred to as “epigenetic”). Such changes in drug sensi-
`tivity might be expected to occurin vivo,since thecells
`are exposed to relatively low concentrations of drugs
`during cancer treatment. In one study, repeated treat-
`ment of tumor-bearing mice with alkylating agents led
`to high levels of resistance of the tumorin vivo, but no
`significant resistance was observed whencells derived
`from the tumors were exposed to the same drugsin cul-
`ture (Teicheret al., 1990). It was proposed that changes
`in the tumorcells led to changes in drugdistribution in
`
`
`
`vivo. Clinically important drug resistance is probably
`due to both genetic and nongenetic mechanisms.
`
`19.2.3 Resistance to Methotrexate
`
`Resistance to methotrexate may occur by several mech-
`anisms (Bertino et al., 1981; Fig. 19.5). Methotrexate
`is transported across cell membranes both by passive
`diffusion and by an energy-dependentactive transport
`system. Drug-resistant cells may arise that have im-
`paired transport of methotrexate into thecell, probably
`due to point mutation (e.g., Flintoff et al., 1976).
`Transport-deficient cells also show a decrease in poly-
`glutamation of intracellular methotrexate (Frei etal.,
`1984); primary defects in polyglutamationorin intra-
`cellular binding could lead to decreased uptake of meth-
`otrexate in the drug-resistant mutants.
`A second type of mutation maylead to production of
`variant forms of DHFR,the target enzyme for metho-
`trexate. Variant enzymes have been foundto retain ad-
`equate function for reduction of their normal substrate
`(dihydrofolate), but to have a greatly decreasedaffinity
`for methotrexate (e.g., Goldie et al., 1980).
`The most common mechanism leading to methotrex-
`ate resistancein cell lines and experimental tumors that
`are exposed to increasing concentrations of the drugis
`overproduction of DHFR from amplified genes. This
`process has been extensively characterized by Schimke
`and hiscolleagues (e.g., Schimke, 1984) and their ma-
`jor findings are:
`
`1. Selection for methotrexate resistance in mammalian
`cells by stepwise increase in drug concentration in
`
`Seee
`
`if
`
`Decreased formation
`of polyglutamates
`Decreased transport
`
`i
`Variant forms o!
`of MTX into cell
`
`“oH FR with low
`
`affinity for MTX
`
`
`
`|
`Methotrexate
`Polyglutomate
`
`
`(MTXKw ) formation
`
`
`Decreased binding
`of MTX to its target
`enzyme OHFR
`
`
`
`Artplitied genes
`for DHFR on
`chromosomes or
`double minutes
`(leads to large
`increase in OHFR
`copies)
`
`Figure 19.5. Probable mechanisms underlying cellular resis-
`tance to methotrexate.
`
`

`

`EXPERIMENTAL CHEMOTHERAPY
`
`343
`
`the medium may lead to as many as 100-1000 cop-
`ies of the DHFR gene, and to high levels of drug
`resistance.
`2. Highlevels of resistance to methotrexate cannot be
`obtained in a single-step selection process. Gene am-
`plification occurs in small steps.
`3. Resistance occurs from overproduction of the normal
`enzyme(although overproduction of variant forms of
`DHFRhasalso been observed).
`4, Drugresistance and amplified genes may be either
`stable or unstable whencells are passaged in the ab-
`sence of the drug. Stable amplification is usually as-
`sociated with chromosomal location of the genes,
`seen as homogeneously staining regions in stained
`chromosomepreparations (chapter 4, section 4.3.1)
`Unstable amplification is usually associated with lo-
`cation of the genes in extrachromosomal chromatin
`structures known as double minutes. Both locations
`may be evident duringselection for drug resistance.
`5, Gene amplification appears to take place by multiple
`replication ofthe DHFR gene(andflanking sequences)
`during the S phase of the cell cycle. Interruption
`of DNA synthesis in synchronized cells by drugs
`such as hydroxyurea, or by transient exposure to
`hypoxic conditions (Rice et al., 1986), leads to an in-
`crease in the rate of gene amplification. Other anti-
`cancer drugs mayincrease resistance to methotrexate
`by gene amplification.
`6. Gene amplification and overproduction of DHFR
`have been observed in cells from human tumors
`treated with methotrexate.
`
`Table 19.2. Drugs Associated with Cross-Resistance and
`Increased Expression of P-Glycoprotein
`
`Etoposide (VP-16)
`Actinomycin D
`Mitoxantrone
`Colchicine
`Puromycin
`Daunorubicin
`Vinblastine
`Doxorubicin
`Vincristine
`(Adriamycin)
`
`Epirubicin Vindesine
`
`that are selected for resistance to an individual drug of
`the group are cross-resistant to other membersofthe
`group but usually show the highest degree ofresistance
`to the drug that was used for selection. Thus, this mech-
`anism of drug resistance shows some degreeofspecific-
`ity. Multidrug-resistant cells do not usually show major
`cross-resistance to antimetabolites or to alkylating
`agents, and increasedsensitivity has been reported to a
`few drugs, including steroid hormones.
`When membraneproteins from multidrug-resistant
`cells are separated by electrophoresis, they are found
`to contain a glycoprotein whose molecular weight is
`about 170 kDa, which has been termed P-glycoprotein
`(Fig. 19.6).
`A direct correlation has been found between the con-
`tent of P-glycoprotein in the cell membraneandthe de-
`gree ofresistance to the selecting drug. Drugresistance
`can be transferred to drug-sensitive cells by transfection
`
`Sf
`4
`x ate&
`eT eeég
`
`Although gene amplification has been studied most
`extensively in relation to methotrexate, there is increas-
`ing evidence for the importance of this mechanism in
`determiningresistance to several other drugs, including
`5-fluorouracil and the multiple drug-resistance pheno-
`50
`> > aatee
`
`type described below.
`*&ee
`
`ee | P-glycoprotein
`<
`
`Origin
`pa
`100
`
`19.2.4 Multiple Drug Resistance
`
`Several investigators have observed that cells which are
`selected for resistance to one of a group of drugs (Ta-
`ble 19.2) may show cross-resistance to each of the other
`drugs in the group (for reviews, see Bradley et al., 1988;
`Endicott and Ling, 1989). Thus, Chinese hamster ovary
`(CHO)cells selected for resistance to colchicine after
`prior treatment with a mutagen were found subse-
`quently to be cross-resistant to several anticancer drugs,
`including anthracyclines (e.g., doxorubicin) and vinca
`alkaloids (e.g., vinblastine), and to other drugs such as
`puromycin, an inhibitor of protein synthesis. Many of
`these drugs have quite different chemical structures, but
`most of them are derived from natural products. Cells
`
`>
`=
`
`L
`
`10
`
`Front
`
`Figure 19.6. Increased levels of P-glycoprotein in increasingly
`drug-resistant mutant Chinese hamster ovary cells derived
`from wild type (Aux B1 cells) by stepwise selection (as indi-
`catedin Fig. 19.3). Membrane components were separated by
`SDSgel electrophoresis, transferred to nitrocellulose paper,
`and stained using a radiolabeled hetero-antiserum.(Reprinted
`with permission from Ling, “Genetic basis of drug resistance
`in mammaliancells.” In: Bruchovsky, Goldie (eds.). Drugs and
`HormoneResistance in Neoplasia. Copyright 1982, CRC Press,
`Boca Raton,FL.)
`
`
`
`

`

`
`
`344
`
`THE BASIC SCIENCE OF ONCOLOGY
`
`of DNA from drug-resistant cells or with cloned genes
`(knownas mdr, or multidrug resistance genes) that en-
`code P-glycoprotein, demonstrating a cause-and-effect
`relationship between the presence of P-glycoprotein and
`drugresistance. P-glycoprotein is present at low concen-
`tration in the cell membraneof drug-sensitivecells, and
`drug resistance is most often due to amplification of mdr
`genes with a consequent increase in the production of
`P-glycoprotein. This effect is analogous to the amplifi-
`cation of the DHFRgene, which leadsto resistance to
`methotrexate (section 19.2.3). In some multidrug resis-
`tant cells, there is increased transcription of mdr mRNA,
`without amplification of mdr genes.
`The mechanism underlyingthis type of multidrugre-
`sistance is reduced accumulation of drugsinside thecell.
`Studies using isotopically labeled drug (Fig. 19.7), or
`fluorescent compoundssuch as doxorubicin, have dem-
`onstrated a decrease in the net rate of uptake, and more
`rapid efflux, of drug from drug-resistant cells. There is
`evidence in somecells that P-glycoprotein may act as a
`permeability barrier to drug uptake, but its major func-
`tion appearsto be as an energy-dependentefflux pump
`which extrudesa variety of unrelated compounds from
`cells (Bradley et al., 1988).
`Genetic analysis has demonstrated that P-glycopro-
`teins are produced by a small family of homologous mdr
`genes, which have been cloned and sequenced from hu-
`man, hamster and murinecells (Gottesman and Pastan,
`1988; Raymond and Gros, 1989; Endicott and Ling,
`1989). Two homologous linked genes (mdr1 and mdr2
`have been located to chromosome7 in humancells, but
`only the protein product of mdr1 is associated with drug
`
`resistance. Three mdr genes have been identified in
`mouse and hamstercells, two of which appear to encode
`a functional P-glycoprotein. The sequences of the hu-
`man and rodent genes show about 80% identity, and
`each gene contains two homologoushalves, suggesting
`that they were produced originally by duplication of a
`’ smaller gene. Small variations in sequenceleadto pref-
`erential resistance to different drugs (Safaet al., 1990).
`Sequencingofthe geneshas allowed determination ofthe
`amino acid composition of the correspondingproteins,
`andthe location ofhydrophilic and lipophilic components
`has allowed modeling ofthe orientation of P-glycopro-
`tein in the cell membrane(Fig. 19.8). This modelsug-
`gests that there are 12 transmembrane domains, with 2
`internal ATP-bindingsites. P-glycoprotein hassignifi-
`cant homology with a number of ATP-bindingtransport
`proteins found in bacteria, suggesting a general role for
`this type of protein in the energy-dependentefflux of a
`variety of agents from cells.
`Use of monoclonal antibodies against P-glycoprotein,
`or of specific genetic probes, that can bind to mdr
`mRNAin northern blots (chapter 3, section 3.3.4), has
`allowed quantitation of P-glycoprotein, or its mRNA,
`in cells from a variety of tumors and normal tissues.
`P-glycoprotein is present at high levels in the normal hu-
`man kidney and adrenal gland; at intermediate levels in
`lung,liver, colon, and rectum; andatlow levels in most
`other tissues (Fojo et al., 1987). Immunohistochemical
`techniques have shownthat P-glycoproteinis localized
`to the surfaces o