Document XzeN9mB184jbMEKMR0E3eRDOw

DownloadSearchable PDFRandom document
Cui~r,rc.ivPrrnrrplcs C.Priirllc?01Omoloqy, Fourth Edrlrun. editcd by Vincent T. VeVila. Jr , Samuel Hullman. Steven A. Rosenberg. J 6. Lippincort Co.. Philadelphia Z 1993 Peter G. Shields Curtis C. Harris 1 1CHAPTER Principles of Carcinogenesis: Chemical Chemical carcinogenesis has its roots in the epidemiology of cancer.' In 1759,John Hill reported that tobacco snuff caused oral cavity cancers.2Soon after, Percival Pott published his findings that working as a chimney sweep, exposure to soot, and poor hygiene led to scrotal ~ a n c e rO. ~ther associations followed, such as bladder cancer with aromatic amine^,^ benzene with l e ~ k e m i aa,n~d lung cancer with tobacco Epidemiologic relations were also explored in the laboratory. As early as 1918and in reaction to Pott's findings, Yamigawa and Ichikawa reported that coal tar could cause skin cancer in laboratory animals.8 which was followed by similar findings for individual chemicals. This chapter presents some of the basic principles of chemical carcinogenesis, focusing on topics that are especially relevant to the understanding of human carcinogenesis. The molecular epidemiologyof human cancer, an emerging field in cancer research, is explored. Reviews of the research accomplishments in the field of chemical carcinogenesis are available and contain more extensive bibliographie~.'~~~'~ MULTISTAGE CARCINOGENESIS Carcinogenesis is a multistage process driven by genetic damage and epigenetic changes (Fig. 11-1). The traditional view of carcinogenesis is derived primarily from studies of animal models, but more recent studies rely on the molecular analysis of cancer-related genes in human cells.' Tumor initiation begins in cclls through mutations from exposure to carcinogens. These mutatcd cclls m:iy have an sltered rcsponsivcness to their iiiic~loc.n~irontiicn'tand a selective growth advantage 200 when compared with the surrounding normal cells. The tumor-initiated cells also may have a decreased responsiveness to the intercellular and intracellular signals that maintain their architecture and regulate homeostatic growth. For example, initiated cells may be less responsi\ e to negative growth factors, terminal cell differentiation. or programmed cell death. Selective clonal expansion of the initiated cells may also occur by physical perturbation of the normal microenvironment (e.g., wounding of mouse skin or partial hepatectomy in ro- dents), chemical agents (e.g., phorbol esters on mouse skin or rat liver and phenobarbital). microbial agents (e.g., influenza virus enhancement of rodent lung carcinogenesis or hepatitis virus in human liver carcinogenesis), or other inflammatory processes. Tumor promotion results in further selective clonal expansion and proliferation of the initiated cells, thereby enhancing the probability of additional genetic damage through endogenous mutations or DNA-damaging agents. During tumor progression. malignant cells continue to exhibit progressive phenotypic changes'' and genomic instability, including gene amplification, chromosomal aber- rations, and altered gene expression.'? I3 The classic view of two-stage carcinogenesis, in which tumor initiation (mutation) is folloi\ ed b! tumor promotion (epigenetic changes). has been conceptually important but is too simplistic. There may be six or more independent mutational events.'" l5 Furthermore, chemical carcinogens may be geno- ''toxic, nongenotoxic. or cause epigenetic effect^.'^." Dose- response relations may be linear or nonlinear." Endogenous mutagenic mechanisms, such as DNA oxy-radical damage. depurination, polymerase infidelity, and deamination of 5methylcytosine also contribute to carcinogenesis.20-13None- i i \nitiation Promotion Conversion 0 Defects in Growth Control Selective \ I l l l / / \ l U l / t ( , i t ( / / / ( I [ / ( `/ , I \ Progression , .- \ NORMAL :ELL - -R U S 0 Activation of Protooncogenes Inactivation of Tumor Suppressor Genes 0 Inactivation of Antimetastasis Genes CLINICAL theless. a debate coniinues regarding the relati\,oimportance \shereas quantitati\,e differences are partly responsible for of endogenous vcrsus esogenous mutagenic e\.ents and the interspecies, interindi1,idual. and intertissue responses to value of animal bioassays or short-term mutagenic assays for carcinogcn~.~' the assessment of human cancer Societal and reg- The metabolism of any individual carcinogen can be com- ulatory decisions crucial to public health are at issue in this plex because i t can be a substrate for several enzymes. For debate. example. significant progress has been achieved in elucidating the metabolic activation and detoxification pathways for poly- cyclic aromatic hydrocarbons (PAH). such as benzo[a]pyrene CARCINOGEN lIETABOLIS.\.1 AND DNA DAMAGE (BP).33 These compounds are composed of fused benzene rings that are essentially water insoluble but are readily ab- Most chemical carcinogens require metabolic act]\ atim sorbed through the lungs and gastrointestinal tract. They are through the generation of highly reactive electrophiles. These commonly found as combustion products of fossil fuels (e.g., form DNA adducts b! co\alently binding to nucleic acids. ex- coal, wood, diesel exhaust) and vegetable matter. Conse- erting a promutagenic effect Metabolic activation is generally quently, PAHs are common environmental pollutants. BP is catalyzed by cytochrome P450 enzymes through oxidation metabolically activated by phase 1 enzymes forming a reactive Cytochrome P450 genes are continually being identified; or- diolepoxide (Fig. 11-2). Initially, CYPl A1 and epoxide hy- ganisms can have over 200 distinct P450 enzymes." A pro- droxylase catalyze the conversion fit` BP to a dihydrodiol. Then, posed nomenclature for individual genes uses the prefix CW cytochrome CYP3A4 converts t!iis product to the diolepoxide followed by an arabic numeral, letter, and numeral that in- (BP-7,8-diol9,1O-epoxide)A.long this pathway, intermediates dicate the family, subfamil!. and gene number, respectively.28 may be detoxified by conjugation, oxidation, or reduction and -r' The families designated CYPI, CYP2, CYP3, and CYP4 are pri- then excreted in urine or feces. marily responsible for metabolism of foreign c h e r n i ~ a l s . ~ ~ Family and subfamilies are determined by the percentage of gene sequence homolog). The ability to isolate and clone hu- CRITICAL DNA TARGETS: PROTOONCOGENES man cDNA has allowed the determination of substrate spec- AND TUMOR SUPPRESSOR GENES ificityfor individual P430 enzymes, in many cases confirming earlier work using purified rodent protein^.'^ Protooncogenes.are normal cellular genes that can be inap- The ability to metabolize carcinogens varies widely among propriately activated to cause dysregulation of cell growth and animal species.27Cytochrome P450 enzymes also vary among differentiation, which increases the probability of neoplastic different tissues within and among specie^.^" 3' Although in- transformation. They can be activated by carcinogens through ferspeciesdifferences have long been known to be quantita- nucleotide base substitutions, chromosomal translocations, tive, the metabolic actir ation pathways are qualitatively sim- and gene amplification. Among the best-studied protoonco- h.T3he1se observations support the qualitative cxtrapolation genes is the RAS fami!y. RAS protein products are involved Of carcinogenesis data from laboratory animals to humans, in signal transduction pathways initiated by growth factors ' 202 Prinriples o ~C' arcinogenesis: Ciieinical Urtnown Diol Protein RNA tGsH-. / /7'8 GSH-BP MEtdbolite Epoxide II Adducts Adducts P4501A1 \F~P450111A4 0 N Mutation f (G -T Transversion) P lNEBP 7 3 Epoxide-( - ) t - 7 , 8 - 0 i o l v Diol Epoxide Reductase MFO(s) Phenols Quinones -.D/TSiol Sulfate Diol THe2trYo/Yk4:F'$ Triol Glucuronide Epoxides GSH-BPDE N` Benzo(a)pyrene-GuanineAdduct FIGCRE 1 1-2. Metabolic activation, deactivation, and formation of DNA adducts for benzoia]p!~ene (BP). and hormones at cell membrane receptors. In several experimental systems, activation is associated with tumor forma- t i ~ na,n~g i~o g e n e s i ~a,n~d~m e t a ~ t a s i sM. ~u~tation of RAS pro- tooncogenes have been observed in several types of human cancers.37 Base substitutions occur nonrandomly at codons 12, 13, and 61 in vitro and in vivo on exposure to such agents as PAHs and radiation ?4.38-40 Mutation of the Ha-RAS protooncogene is an early event in rodent models of skin and mammary c a r c i n o g e n e ~ i s . ~I~n, ~a'ddition, the v-Ha-RAS transgene can substitute for the initiation step in mouse skin carcinogenesis in transgenic In contrast to protooncogenes, tumor suppressor genes are normal cellular genes that can be inappropriately inactivated to cause dysregulation of growth and differentiation pathways. This also increases the probability of neoplastic transformation. Tumor suppressor genes perform different functions (Table 11-1). Although the loss of suppressor gene function is dominant in carcinogenesis, the inheritable trait is recessive; loss of function for both alleles is the basis of the two-hit hypothesis for c a r c i n o g e n e ~ i s .A~ l~te,r~n~atively, suppressor genes can be inactivated by genomic i m ~ r i n t i n gi,n~cr~eased proteolytic digestion of gene product^,^' or other dominant negative mechanism^.^^ Based on target size theory, the inactivation of a single allele of a tumor suppressor gene (in which mutations at multiple sites may cause inactivation) should have an intrinsically higher probability than activation of a protooncogene (in which only a few mutations at specific codons can cause activation). The requirement for inactivation of both alleles of a tumor suppressor gene counterbalances this probability except in those cases in which an inactivated allele is inherited (e.g., familial polyposis coli, Li-Fraumeni syndrome, Wilms' tumor, or retinoblastoma) or the process TABLE 1 1 - 1 . Examples of Functions of Putative Tumor Suppressor Genes Induce terminal differentiation hlaintain genomic stability - Trigger senescence Regulate cell growth Signal transducers of negative growth factors Regulators, e g . , PTPase-7, of tyrosine kinases - Induce proteases Induce programmed cell death Alter DNA mcthylasc activity hlodulatc 1iistocotnp;ltibility antigcns Rcgulutc angiogcncsis F x i l i t a t c ccll-ccll c . o m ~ n ~ r n i c ; ~ t i o n of inactivation is by a dominant negative mechanism as proposed for the p53 tumor suppressor gene.48.49Ionizing radiation, carcinogenic hormones, metals, aldehydes, and fibers such as asbestos tend to cause gross chromosomal abnormalitie~.~TOherefore, tumor suppressor genes may be targeted by these carcinogens more readily than protooncogenes. Chemical carcinogens frequently cause promutagenic DNA adducts and chromosomal abnormalities. f !*' :* MUTATIONAL SPECTRUM The mutational spectra of endogenous and exogenous carcinogens are largely responsible for actix ating protooncogenes and inactivating tumor suppressor genes. Studies using prokaryotic, simple eukaryotic, and site-specific mutagenesis assays have suggested that carcinogenic agents produce a fingerprint of DNA adducts and mutations. Similar evidence exists in eukaryotic studies such a s esogenous gene insertion by shuttle vector" " or endogenous gene ana1y~is.jT~hese can become mutated at specific loci to produce detectable phenotypic changes (adenine phosphoribosyl transferase [APRT].dihydrofolate reductase [DHFR].and hyposanthineguanine phosphoribosyl transferase [HPRT]) However, these assays may underestimate mutational frequency because gene deletions, chromosomal nondisjunction. and frameshift mutations cause loss of other genes essential for cell sumival. In addition, certain mutations ma! cause clonal expansion or inhibition of the mutant cell. funher complicating the interpretation of the mutational spectrum. A comparison of the spontaneousl? occumng mutational spectrum at the APRT locus in Chinese hamster o\ ary (CHO) cells with spectra induced b? ionizing radiation. ultra\ iolet radiation. or benzo[a]pyrene diolepoxide is shown in Table 11-2. The mutational spectrum of ultra\iolet light is consistent with mutagenesis models of promutagenic cyclobutane and pyrimidine-pyrimidone photoproducts. Ionizing radiation, in contrast, most frequently causes deletions. although specific point mutations are also obsen ed. Benzo[n]pyrenediolepoxide, which binds to the 2-amino group of deoxyguanosine, produces predominantly G:C-T .Itrans ersions, a similar mutational spectrum for benzo[a]p!rene diolepoxide has been obsened in human cell assa)s " The HPRT locus is of particular intcrest because of the potential to compare mutational spectra in cultured cells with those in human lymphocytes of persons esposed in vivo to en1ironniental carcinogen^.^' The molccular analysis of mutationally activated RAS protooncogenes in animal models suggests that mutational spectra ;* , TABLE 11-2. Percent Distribution of Mutations in the Endogenous Adcninc phosphoribosyi Transferase Locus in Chinese Hamster Ovary Cells Mutagen Trarisitiom G:C .-c.i:T A:T + G:C G:C + Tr.4 Tmiisiwsiows G:C -+ C;C: :\:T-+ C':G Spontaneous 71 0 13 3 0 Ultraviolet radiation 61 2 5 10 5 Ionizing radiation 19 6 6 6 19 Benzo[a]pyrene diolepoxide 5 0 62 14 0 ;l:'r-t 7':d 7 12 13 9 D ~ / ~ ~ o ? LI .~sz s ~ ~ ~ ~ o ~ I s 60 22 31 0 55 reflect DNA-adduct formation For example, mutations found in activated RAS protooncogenes of rodent tumors after AInitroso compound exposure are predominantly G:C-A T base substitutions; these are Iihelp due to methylation of deoxyguanosine at the O6position followed by mispairing with thymine during DNA synthesis.j6Although there are several guanine residues in R4S codons that would generate a transforming protein if substituted with adenine, the animal experiments have re! ealed that the mutations occur overwhelmingly at only certain mutation sites (codons 12, 13, and 61). Unexplained mutational specificities have been observed in other experimental s! stems and may reflect the differences in the spectrum of promutagenic carcinogen-DNA adducts, specificity of DNA repair enzymes, or resultant amino acid changes that are silent or nonlethal 57-59 The mutational spectra of activated R4S protooncogenes in tumors and preneoplastic lesions of laborator? animals may be instructive in the interpretation of mutations in human cancers. The spectra of Ki-RAS protooncogene mutations in human adenocarcinomas vaq according to tissue site, although the base substitution is not always specific.-For example, G:C+T:A transversions occur in lung tumors, which can be caused by PAH exposure, endogenous mutagens (8-OH-deoxyguanosine caused by oxy-radical damage), DNA depurination, or polymerase infidelity.6061 The p53 tumor suppressor gene is ideally suited for analysis of mutational spectra. First, p53 is well conserved in evolution and the DNA sequence of 5 domains are more than 90% homologous among humans and rodents.62Second, p53 is mutated in diverse types of human cancer.63Third, a wide spectrum of mutational types and codon sites has been observed that presumably define regions of the p53 protein likely to be essential for tumor suppression, cell-cycle control, and interactions with cellular and viral proteins. Most mutations in human tumors occur in the evolutionarily highly conserved domains in exons 5 to 8 of the p53 gene.63 Moreover, the missense mutations are predominantly transitions at G:C base P a ,and 95% of the amino acids are entirely conserved in mouse, rat, monkey, and human. Mutational spectra also vary among cancer types (Table 11-3).The G:C4A:T transitions are most frequent in colon tumors and 68% occur at CpG dinucleotides. These findings are consistent with endogenous mutational mechanisms due to deamination of 5-methylcytosine residues. More than half of the p53 mutations in colon tumors are at hotspot codons 175, 248, 273, or 282, each of Which is a CpG dinucleotide. In contrast to colon tumors, CpG dinucleotides are less frequently found at mutation sites of other human cancers. For example, G:C-T:A transversions are seen commonly in breast and lung cancers but not in colon tumors. CpG-TpG mutations occur with intermediate fre- quency in esophageal cancers, but 36% of these mutations are at A:T pairs, which may be due in part to DNA depurination or exposure to chemical carcinogens such as urethane (a con- taminant of certain alcoholic beverages) or acetyladehyde (a metabolite of ethanol). The most striking p53 mutational spectrum is found in he- patocellular carcinomas from Qidong, People's Republic of China64and southern Africa.65Eleven out of 12 base substi- tution mutations in 26 tumors \ZYC ?: :I-$, third base position of codon 249, and all but one I:( T:A transversions. One additional mutation at codon , s also a G:C-+T:A transversion. These tumors were from patients who live in geographic areas where aflatoxin B, and hepatitis B virus are major risk factors for liver cancer. Aflatoxin B1,a product of mold that grows in crops, forms promutagenic adducts on deoxyguanosine and induces primarily G:C-T:A transver- sions and G:C-A:T transitions in experimental systems.66 Analysis of liver tumors from geographical areas where af- latoxin B, is considered not to be a significant rjsk factor will help determine whether exposure to this carcinogen or an- other coincident carcinogen may be responsible for these p53 mutations. CHEMICAL-VIRAL INTERACTIVE EFFECTS Interactive effects of chemicals, viruses, physical agents, and host factors have been observed (Table 11-4). One example is the relation of aflatoxin B, and hepatitis vjrus to the development of hepatocellular carcinoma. Hepa:itis B virus (HBV) and more recently hepatitis C virus have been linked to primary hepatocellular carcinoma.67Geographic location of HBV carriers indicates that factors such as aflatoxin Bl and alcoholic beverages are also important. Another example is in uranium miners, in whom the coexposure of tobacco smoke and radon significantly acts to increase risk of lung cancer.68 Occupational asbestos exposure and tobacco consumption also act synergistically toincrease the incidence of bronchogenic car~ i n o m a . ~In' the laboratory, in vitro DNA strand breaks increase with the combined exposure.70 Moreover, tobacco smoke and asbestos, through their ability to induce inflammation and lipid peroxidation, cause oxidative DNA damage.7'.72Another consideration for the interactive effect is the ability of asbestos to adsorb PAHs and act as a carrier for the ~arcinogen.~~ 204 Principles of Carcinogenesis:Chemical TABLE 11-3. Examples of Human Cancers With Base Substitution Mutations in the p53 Gene Worldwide Number of Mutations Detected in BCuarndceenr (Rank)182 Cuncer LTiunmesoranCdell MHouttsaptoiotsnal Tumors' (Codon) A:T GrC C p C + TpG Mutations G:C+ ArT G:C+ T:A G : C 4 C:G 1 2 3 4 7 8 11 9, 12 - - - Stomach Lung Breast Colon Esophagus Liver Bladder Leukemia and lymphomas Skin Sarcocias Brain ovary 6 24 70 273 7 63 63 15 48 48 175, 248, 10 38 272,282 38 12 26 22 249 2 20 15 3 12 84 175, 213, 26 58 248,272, 282 16 0 16 12 0 12 24 273 4 20 14 3 1 1 2 14 14 30 7 0 5 35 1 7 9 1 31 22 32 27 12 36 0 16 10 3 16 72 45 6 llt 4 82 17 2 61 0 9 9 2 0 1 3 7 1 2 1 4 * Mutation screening by an RFLP analysis of one site is not included. t Includes tandem double mutations. ASSESSMENT OF CANCER RISK IN HUMANS Our understanding of carcinogenesis and risk to human health comes from experimental models and methods including mutagenesis assays, mammalian cell-culture experiments, animal studies, classic epidemiology, and molecular epidemiology. The usefulness of each method can be contrasted with its limitations (Table 11-5).The evaluation of an individual patient must rely on an accurate history, physical examination, and research data. The latter is often beyond the scope of the practitioner, but methods of indiiidual cancer risk assessment have been proposed and resources are a~ailable.~S' ome known or potential human carcinogens are indicated in Table 11-6. Short-term assays for mutagenesis provide quick and inexpensive screens for potential carcinogen^.^^,'^ Among the most widely used and sensitive is the Ames' assay. in which frameshift mutations or base-pair substitutionsare measured in Salmonella typhimurium ba~teria.~T'he Ames' assay also TABLE 11-4. Interactive Effects of Carcinogens, Viruses, and Host Factors Type Example Associated Tumor Chemical-chemical Viral-chemical Physical-chemical Chemical-host Physical-host Viral-host Tobacco smoke and alcoholic beverages HPV and tobacco smoke EBV and N-nitrosamine HBV and aflatoxin B, Asbestos and tobacco smoke Radon and tobacco smoke PAH and CYP2D6 Tobacco smoke and CYP2D6 Asbestos and CYP2D6 Sunlight and xeroderma pigmentosum Radiation and RB-deficient genotype EBV and X-linked immunodeficiency syndrome Otolaryngeal. esophageal Cervical Nasophanngeal Liver Lung Lung Lung Lung Lung Skin Osteosarcoma Lymphoma HPV. human papilloma virus; ERV. Epstein-Barr virus: HBV, hepatitis B virus; CYP2D6. crytochrome P-450 CYP2D6 metabolic phenotype determined by debrisoquine sulfate administration and measuretncnt of urinary metabolites; PAH, polycyclic aromatic hydrocarbon; RB, retinoblastoma susceptibility genc. _* T.4BLE 11-5. Testing for Carcinogenicity Method Ad zlantaqes ~ Disadzu 72tayes In vitro testing Economical Uncertain in vitro to in vivo , Rapid results extrapolations Human cells can be used Frequent false-positives and false- negatives Mutagenicity is not carcinogenicity Substantial interlaboratory variation .Animal bioassay More predictive of human experience than short term tests Elucidates species differences Expensive Doses are higher than those experienced by humans Uncertain animal to human extrapolation Clissic epidermiology Direct measurement of human experience Covariables examined Dose-response data lnsensitive Does not prove causation Unknown confounding variables Slolecular epidemiology Measures internal dosimeter of exposure and genetic predisposition Identifies risk in an individual Early stage of development and validation TABLE 11-6. Selected Known or Potential Human CarcinogensIB3 Known or Potential Carcinogen * Target Organs Known or Potential Carcinogen* Chemical Aflatoxin 4-Aminobiphen! 1 Arsenic Asbestos Benzene Benzidine Benzo[a]pyrene Beryllium bis(chloromethy1)ether Cadmium Chromium coal tar and pitch Nickel compounds vinyl chloride Minerd oils Mustard gas %-Napthylamine - and Diet with tobacco moke beverage salted, pickled foods iet Liver Bladder Skin, lung Lung, pleura Bone marrow Bladder Skin, lung Lung Lung Prostate, lung Lung Skin, lung, bladder Lung, nasal cavity Liver, lung, brain Skin Lung Bladder Skin Skin, lung Oral Multiple sites Oral, esophageal, pancreas, liver Gastrointestinal Colon, mammary Industryt Aluminum production Auramine manufacture Boot and shoe manufacture Coal gasification Coke production Furniture and cabinet manufacture Iron and steel founding Hematite mining (radon) Isopropyl alcohol manufacture Magenta manufacture Paints Rubber industry Viral Epstein-Barr virus Hepatitis B Hepatitis C Human immunodeficiency Human T-lymphocyte Human papilloma hogenic agents or activitiesin this table are classified by the International Agency For Research er as recognized human carcinogens. The list is not all inclusive. to only some occupations within an industry. tis L,Aitio A, \I'ilbourn J , Shuker L., Human carcinogens so far identified. Jpn J Cancer Res Target Organs Lung, bladder Bladder Bone marrow, nasal sinus Skin, lung, bladder Skin, lung, kidney Nasal Lung Lucg Nasal sinus Bladder Lung Bone marrow, bladder Nasopharynx Liver Liver Lymphatic Lymphatic, bone marrow Cervical 206 Primiples oJ' Cnrciiiogenesis: Chemicul has been used as a biomonitor in hurnans. Urine from cigarette in one method are generally positive with other methods, al- smokers, for example, is mutagenic:."Other short-term assays though 100%concordance does not exist. Sensitivity for short- use Chinese hamster ovary cells, mouse lymphoma cells, V79 term assays is high but specificity is OW.^^.^' For example, cells, or rat hepatocytes to measure forward mutations, sister chemicals that are predictive of reactivity by chemical struc- chromatid exchanges, and unscheduled DNA ~ y n t h e s i sA. ~l-~ ture are commonly mutagenic, but 84% of tested carcinogens though short-term assays are useful in identifying potentially and 66%of noncarcinogens are m u t a g e n i ~ . ~C~h,e'm~icals that carcinogenic compounds in the respective cell system, the are not predictive of reactivity and are nonmutagenic are car- same sensitivity makes the results difficult to extrapolate to cinogenic less than 5% of the time. The concordance from humans; positive results might be unique io the strain. Factors animal to human experience is wide ranging (5% to 70%), such as metabolism, repair, and exposure cannot be assessed. although carcinogens that are more potent in one species tend Laboratory animal studies provide an important source for to be more potent in others.85." the identification of potential carcinogens in humans, mostly The physician can look to various regulatory, governmental, because few better alternatives exist. These animal bioassays or review organizations for extensive evaluation of the sci- are also expensive and time-consuming, preventing the testing entific literature. Some organizations generate documents re- of a large number of c h e m i ~ a l s . ~A~s.r'e~commended by the porting the findings of a panel of experts who critically review National Toxicology Program, carcinogenicity bioassay studies the scientific literature, whereas others simply summarizedata should use lifetime exposures in rats and mice with maximally from other organizations. Several lists of carcinogens have tolerated doses (MTD), that is, those not producing clinically also been published, but the evaluations of the literature and evident toxic effects. T o infer that a carcinogenic effect is the definitions can vary greatly among organization^.'^ The present in laboratom animals, dose-response relations, overall physician should be aware of the purposes and goals of an mortality rates, and consistency with data from other species organization when requesting its information. Quantitative must be examined. The limitations in these experiments in- risk assessment methods are used by regulatory agencies to clude the routine use of the MTD that potentially increases estimate the risk to a population exposed to a particular car- cell replication and endogenous mutations; interspecies and cinogen at a specific dose. Risk assessments serve public interstrain differences; use of rodents known to have high health interests as they attempt to predict the frequency of spontaneous rates of cancer; an inability to account for met- cancer in a population before epidemiologic investigations abolic differences between high- and low-dose exposure; and can be performed and advese outcomes occur. Several math- difficulty in interpreting data from doses that commonly ex- ematical models rank relative risks and suggest regulatory ceed those experienced by human^.^^-^^ exposures limits. Risk assessments include four general Human investigations provide the most relevant data re- steps." garding human risk. Classic epidemiology measures the incidence or prevalence of disease in human populations. Epidemiologic studies have identified previously unknown risks such as asbestos-related pleural mesothelioma, benzeneinduced leukemia, and bladder cancer in dye workers. Epidemiologic methods by themselves do not demonstrate causation. The assessment of causation can be aided by Sir Austin Bradford-Hill'sproposed criteria, whch consider the strengths of an association, consistency, specificity, temporality, biologic gradients, biologic plausibility, coherence, and anal~gies.`~ Study design and controlling for confounding variables are important determinants of a true association. For example, assessing cancer mortality rates from death certificates can be unreliable because the certificate diagnosis frequently is not accurate. Loss of persons in follow-up, inappropriate choice of control populations, ascertainment bias secondary 1. Hazard assessment, which qualitativelyreviews scientific literature to determine i\-hethera hazard might exist 2. Dose-response assessment. which evaluates the doses used in scientific studies and relates them to human exposures 3. Exposure assessment. ij-hich examines a population thought to be at risk with regard to the quantity, duration, and routes of exposure. Secessary information includes quantitative measurements of relevant media such as soil, water, air, or body fluids. 4. Risk characterization. which incorporates the above information and evaluates che assumptions and the uncertainties to estimate risk. At the conclusion. an incidence of cancer will be predicted. such as 1 additional person in 1 million persons. to an unrelated cancer cluster, and "healthy worker effects" Each of these steps requires assumptions that are open to may also influence outcomes. The latter effect can be impor- debate, such as subjective evaluations of the literature, ex- tant if an occupation impacts on risk factors, for example, by trapolations from laboratory animals to humans. and math- not allowing tobacco consumption at the workplace. It is also ematical methods. It is tempting to extrapolate conclusions important to examine existing exposure data and dose- of a formal risk assessment to risk for a given individual. es- response relations. Finally, statistical methods, beyond the pecially if that person is a cancer patient. Such an exercise scope of this chapter, must be carefully chosen to determine is inherently flawed because risk assessments calculate risk whether a finding is attributable to chance. in a group of people and do not assess coexposures or inher- Thc concordance between different methods for inferring itable variables that may increase or decrease an individual's human cancer risk is variable. In such cases as aflatoxin B,, risk of cancer. Separately. risk assessments must be reeval- mustard gns, radon. and vinyl chloride. animal testing rcvcaled uatcd when expcrimental or epidemiologic data become a carcinogenic risk hclore hunilin cpidcmiologic evidence \vas a\.ailablc. One example is dioxin exposure. whose laboratory av:iil:ililc.H Cotnp;iring DNA rcactivity based on chemical aniriial carcinogenicity data led to strict regulation in humans s!ritCtiiw. t i ~ i i t : i ~ c t i i ~cia~rcyi,nogcnicity in laboratory anilnals and with considerable cost. However, recent data have led atid c~)itl(.tiiiolol:\.nio5t chcmicals thnt have positive results the Environmental Protection Agency to review its policy.Bg EECULAR EPIDEMIOLOGY HUMAN CANCER ecular epidemiolog! is a niultidisciplinary field that seeks explore cancer risk through molccular genetics and biochemical methods.g0In contrast to classic epidemiology. u hich identifies cancer risk in populations, molecular epidemiology explores cancer risk in tndn iduals. This strategy includes exposure assessments b! measuring biologicallv effectil e doses of carcinogens in target tissues (or surrogate tissues) and an analysis of host susceptibilit: factors, within the frameis ork of well-designed epidemiologic studies. Among the bcst-studicd dic%tar\carcitic~gc.:~ir~<~, i l l ( , . ! l l c ~ toxins produced by ..\spc,r:yilliisJlc/~a,~n~ds . I ~ p ~ t q / p/ /i~rt/rssiliWS. Thcse molds arc contaminants of corn. pcanuts. sor:hum. and ricc. 7'hc risk of hcpatoccllular carcinoma corrclatcs \\it11 thc degrcc of contamination by geographic rcgion."" Adduct I c \ d s also corrclate \\it11 regional exposure and are inversely correlated with residence in industrialized countries.""."" N-nitrosamines also are dietary (and tobacco smoke) carcinogens whose adducts can be measured in DNA."'-'26 In this case, adduct levels are elevated in Chinese persons with esophageal cancert2'and in Japanese persons with liver cancer,'" confirming the risk associated with these adducts and their parent N-nitroso compounds. EXPOSUREASSESSMEST Carcin0gen-DN.Aadducx. somatic gene mutations. and cy- togenetic changes can be measured in the DNA of target cells or in surrogate blood The observation that CUcinogen-DNA adducts formed in cultured human tissues are generally the same ajthose found in experimental animal cancer models has encourased investigators to search for DNA adducts in humans. Carcinogen-DNA adducts are the result of exposure, absorption. metabolism, and DNA repair. The amount of a carcinogen that reaches the target DNA reflects the biologically effective dose and can be distinguished from the relevance of ambient air or other environmental mea- surement of exposure. .A l x i e t y of assays are available to identify carcinogen-DX.4 adducts in human tissue^.^^.^' Studies using laboratory animals generally demonstrate a relation be- tween dose. adduct level. and c a r c i n o g e n i ~ i t yI.t~is~ ~ex~-~ pected that human epidemiologic studies will bear a similar relation. The most important lifestyle risk factor in carcinogenesis is tobacco smoke e~posure.~Doue to aggressive advertising and the addictive nature of cigarettes, tobacco smoke has be- come the major cause of ca'ncer. Of growing concern is the documentation that passive exposure to tobacco smoke also increases the risk of lung ~ a n c e r . ' ~Pu. t~at~ive adducts have been correlated with consumption by the 32P-postlabelingas- say in human alveolar lavage cells,Io2 and pla- centa,lo3but not in or oral mucosa.1o5Tobacco- specific N-nitrosamines are potent carcinogens in laboratory animals.96Levels are higher in secondary rather than main- stream smoke, highlighting the role of passive smoke expo- sure. Hemoglobin adducts are increased in smokers over non- smokers, whereas even higher levels are reported in snuff dippers.'% Nitrosamine-related adducts have also been found in the human lung.lo7.1s Urine also can be used as a biomarker for tobacco smoke exposure because it is mutagenic in smok- ers. However, this cannot be solely attributed to tobacco be- Cause dietary heterocyclic amines also are found in the Urine.'m ..f PAHs, associated with an increased risk of lung and skin can cause adducts found in the white blood cells of Persons exposed to coke oven emissions, tobacco smoke, and areas (e.g., from industrial p ~ l l u t i o n ) . " ~ -D" i~etary ex- posure resulting from the overcooking of meats and fish also results in elevated adduct levels.' l7 Adducts also have been found in human lung and placental samples.''EWide inter- individual variations in levels have been noted, presumably reflecting variable cancer risk in individuals. INTERINDIVIDUAL VARIATION AND HOST SUSCEPTIBILITY FACTORS No one supposes that all the individuals of the same species are cast in the v e q same mould. These individual differences are highly important for us. . . . Charles Darwin, The Origin of Species, 1859 Metabolic Actiuatiori and Deactivation Physicians and scientists have repeatedly recognized person- to-person differences in behavior, morphology, and risk of disease. Such interindividual differences reflect inherited and acquired factors. For example, inherited differences in sus- ceptibility to physical or chemical carcinogens have been ob- served, including an increased risk of sunlight-induced skin cancer in people with xeroderma p i g m e n t o ~ u m ,b' ~la~dder cancer in dye stuff workers with a poor acetylator pheno- type,'28and bronchogenic carcinoma in tobacco smokers who have an extensive debrisoquine hydroxylator phen~type.'~I3'O Because most chemical carcinogens require metabolic acti- vation to exert their oncogenic effects, interindividual varia- tion in metabolism is considered to be an important deter- minant of cancer s u s ~ e p t i b i l i t y . ~ ~ PAH metabolism is probably the best example of the effects of interindividual variation and cancer risk. Several thousand- fold interindividual variation has been observed in lung.32Pla- cental aryl hydrocarbon hydroxylase (AHH) activity, which is under direct genetic control, can be induced by maternal exposure to environmental carcinogens (e.g., tobacco smoke or dietary factors) .131.132 Higher AHH activity is generally cor- related with higher adduct leve:.I0' The induction process itself may have a genetic c ~ m p o n e n t , a'n~d~inducible activity is higher in cultured lymphocytes from lung cancer cases com- pared with The ability to form benzo[a]pyrene diolepoxide-DNA adducts was higher in lung cancer patients than in ~ o n t r o l s . ' ~A~h. i'g~hl~y inducible allelic variant of CYPlA1 has been identified in h ~ r n a n s . ' ~A~r,e's~tri~ction fragment length polymorphism for CYPlA1 has also been described that is associated with increased risk of tobacco smoking associated lung cancer in a Japanese s t ~ d y . ' ~ ' CYP2D6 activity, another P450 cytochrome, is polymorphic and has also been linked to lung cancer CYP2D6 hydroxylates xenobiotic antihypertensives (including debri- soquine), antidepressives, and a carcinogenic tobacco-specific N - n i t r o ~ a m i n e .A' ~n~individual's polymorphic phenotype is inherited in an autos3mal recessive manner. The rate of 4- hydroxylation of debrisoquine varies several thousand-fold - 208 Priiiciples or Corciiiogenesis: Chemical among people. Lung, liver, and adlmced bladder cancer patients are more likely to have the er.1ensive hydroxylator phenotype when compared with noncancer controls.'29."1.14T2 he increased risk in lung cancer is found primarily for histologic types other than adenocarcinoma of the lung and increases in persons who are occupationally exposed to high amounts of asbestos or PAHs.'~'.'~~ Acetylation of carcinogenic aromatic amines has been proposed as a cancer risk factor.144The Kacetylation polymorphism is controlled by two autosomal alleles at a single locus in which rapid acetylation is the dominant trait and slow acetylation is recessive. The slow acetylator phenotype has been linked to occupationally induced bladder cancer in dye workers exposed to large amounts of N-substituted aryl compounds.128 In contrast, the rapid acetylator phenotype is more commonly found in cases from two studies of colon cancer145.14b6ut not in a n ~ t h e r . ' ~W' hether this association is due to metabolism of a carcinogenic aromatic amine in the colonic epithelium is not known. Glutathione S-transferases (GST) are multifunctional proteins that catalyze the conjugation of glutathione and electrophiles, including the ultimate carcinogenic metabolite of b e n z o [ ~ ] p y r e n e . ~ ~T"h.e' ~th~ree isoenzymes of GST (a,p, and T) vary in their substrate specificity, tissue distribution, and activities among individual^.'^^ Expression of GST-k is inherited as an autosomal dominant trait'" and individuals with low GST-/Lactivity may be at a greater risk for lung cancer caused by cigarette smoking.152 malian DNA repair genes and their molecular mechanisms provide an opportunity for i n ~ e s t i g a t i o n . ' ~I~n.a' ~dd~ition to finding severely depressed excision repair rates in xeroderma pigmentosum cells (e.g., complementation group A), a fivefold variation among individuals in unscheduled DNA synthesis induced by UV exposure of lymphocytes in vitro has been found in the general p ~ p u l a t i o n . 'A~s~ignificant reduction in unscheduled DNA synthesis induced in vitro by N-acetoxy-2acetylaminofluorene has been observed in mononuclear leukocytes from individuals with a history of cancer in first-degree relatives compared with those without a family h i ~ t o r y . ' ~ ~ . ' ~ ~ Interindividual variation has been noted in the activity of 06-alkyldeoxyguanine-DNAalkyltransferase; this enzyme removes the adduct 06-deoxyguanine caused by N-nitrosamine exposure. It is a suicide protein that transfers adducted alkyl groups to itself and becomes irreversibly inactivated. Cell cytotoxicity and tumor-cell resistance are negatively correlated with levels of this enzyme."* \Vide variations in DNA repair activity have been observed in different types of tissue^,'^^.'^^ and cells may have lower repair rates after terminal differentiati011.I~'The activity of this DNA repair enzyme is inhibited by certain aldehyde~'~a'nd alkylating cancer chemotherapeutic agents."' A decrease in this DNA repair activity has been observed in fibroblastsfrom patients with lung cancer compared with donors with melanoma or noncancer con- -t r o l ~ . "Th~erefore, acquired or inherited deficiency in 06- alkylguanine-DNA-alkyltransferasemay be a cancer risk fattor in tobacco smokers. Protooncogenes and Tumor Suppressor Genes Genetic differences in protooncogenes and tumor suppressor genes can also be predictive of cancer risk. Inheritance of a germline mutation in the tumor suppressor gene p53'54.155 and perhaps RBIs6 predisposes to the del elopment of cancer. It has been found that inheritance of rare alleles for H-RAS, detected by a restriction enzyme digest and Southern blot analysis, is found more frequently in lung cancer cases than c ~ n t r o l s . " ~ .T' ~h~ese alleles may provide an unstable site where recombination or amplification occurs. The relation of this polymorphism to lung and other cancers has been extensively r e v i e ~ e d . ' ~A'nalysis of the L-JJYC protooncogene also demonstrates a polymorphic restriction enzyme site. The site present allele has been detected more frequently in persons with soft tissue sarcomas and gastric and lung cancer, although not consistently in lung cancer.160-16T2he association also suggests a worse prognosis at the time of diagnosis in renal cell cancer patients. DNA Repair Rates DNA repair enzymes modify carcinogenic damage by the removal of DNA adducts. Studies of cells from donors with xeroderma pigmentosum have been particularly important in expanding our understanding of DSA excision repair and its possible relation to risk of cancer."' DSA repair rates. but not fidelity. can be determined by measuring unscheduled DNA synthesls and removal of DX.4 adducts. Substantial interindi\kI~idvariations i n DNA repair rates have been observed."'.' Thc fidclit! o f DNi\ repair also may vary among indi\ iduiils. ;itid rcccii[ ad\.nnccsin the idcntification of mani- CARCINOGENICITY OF CHEMOTHERAPY The carcinogenic effect of some chemotherapeutic agents in humans has been well documented and supported by laboratory animal studies (almost an 85% concordance), and re\ ieu s with extensive bibliographies are a>ailable.'74-17T6he International Agency For Research on Cancer, through working groups of scientists and ph! sicians, has identified about 20 single agents or combination chemotherapy regimens for M hich there is sufficient evidence of a carcinogenic effect in humans. and about 50 others in N hich a carcinogenic effect is suspected (Table 1l-7).lT4I-- The identification of a carcinogenic effect does not preclude its use for treatment in patients The decision process depends on the specific risks and benefits, considering the prognosis and life expectancy of the patient with and without treatment. Carcinogenic chemotherapeutic agents generally exhibit target organ specificity. The bone marrow, lymphatic system, and urinary bladder are most commonly affected. As with other carcinogens, genotoxic agents generally act by forming promutagenic DNA adducts and cro.-s"slinking,"' although chromosomal aberration also occurs ' * P130 metabolic activation is not necessarily required. Nitrogen mustard-type compounds (including nitrogen mustard. chlorambucil, melphalan, and chloronaphazine) have high]) reactn e electrophilic centers that react with DNA to form adducts \\ithout metabolic acti1 ation. Cyclophosphamide, in contrast, is a nitrogen mustard thdt undergoes cytochrome P330 metabolism in the liver, \\ here it is converted to acrolein and 4-hydroxycyclophospharnide. This latter compound breaks down spontaneously to phosphoramide mustard, \\hich then reacts with DNA. TABLE 1 1-7. Selected Pharmaceutical Agcnts That .4rc Known or Potential Human Carcinogens* Phenacetin Chloramphenicol Doxorubicin .4zacitidine Chloronaphazine Bischloroetti! lnitrosourea Busulphan Chlorambucil Chlorzotocin Cisplatin Cyclophosphamide hlelphalan Nitrogen mustard Procarbazine Streptozocin Thiotepa Arsenic salts Coal tars 8-Methosypsoralen 8r UVA Testosterone Nonsteroid estrogen Estrogen replacement .4zatiiioprine Cyclosporine Llrinan- tract. nasal -t Mammary, skin Bone marrow, lymph. lung, skin Skin, lung Lung. neurologic. peritoneum Thymus. ovary Lung. bone marrow, ovary Peritoneum Lung, bone marrow hlamniary, bone marrow, bladder. liver. testis, neurologic Lymphatic. lung Skin. lung. lymphatic Lung, bone marrow, kidney Liver, pancreas Lung, lymphatic, bone marrow, uterus. mammary Lung, stomach Skin Skin Cenix. uterus. prostate Vagina. cervix, uterus. ovary. mammary Kidney Lymphatic Lymphatic Kidney Bone marrow - - Bladder Bone marrow Bone marrow Bone marrow Bone marrow - - Bladder, bone marrow Bone marrow Skin Bone marrow Skin Skin Skin Liver Vagina, cervix Endometrium Lymphatic Lymphatic ~ ~ ~ ~~ * Classified by the Agency for Research on Cancer as known or possible human carcinogens. This list IS not all inclusive. t Cancer not observed. Chemotherapy combinations can significantly raise the risk of secondary tumors, especially nonlymphocytic leukemias. The combination of lomustine, cyclophosphamide, and vincristine led to a leukemia incidence of 14% over 4 years after treatment.'79 Nitrogen mustard, vincristine, prednisone, and Procarbazine for the treatment of Hodgkin's disease yield leukemia rates up to 17%.'80;\lore than six cycles increased the relative risk from 9 to 14. Radiation further increases the risk of leukemia.181 . .,I ' r7 R*,E. FERENCES 1. b r i a CC Chemical and ph\sical carcinogenesis Advances and perspectives Cancer ' ka 1991.51 50238-50448 2 HUI J. Cautions wrist the unmcderateuse of snuff London R Baldwin and J Jacobson, 1761 3. POtt P. Chirurgjcal obsenauons relative to the cancer of the scrotum London L Haw-. W Clark. and R Collm 1775 '. Rehn L Blasengeschwulste bel fuchsin-arbeiten Arch Klin Chir 1895.50 588-600 . &lore P. Borgomano C Leucemiealignee au cours de I'intoxication benzenique Sur )'niglnemuquede cen;unes leucemis agumet leur relations avec les anemies graves J Med Lyon 1928.9 227 ' bll R,Hill AB Smoking and w c m o m a of the lung Preliminary report Br Med J 'QQ2 739-748 7. Wynder EL, Graham EA.Tobacco smoking as a possible etiologic factor in bronchogenic carcinoma. JAMA 1950.143329-336. 8. Yamagiwa K. lchikawa K. Experimental study of the pathogenesis of carcinoma. J Cancer Res 1918;3:1-21. 9. Brugge J. Curran T. Harlow E. McConnick F. 0.igins of human cancer. Cold Spring Harbor, NY:Cold Spring Harbor Laboratory Piess, 1991. 10. Cooper C, Grover P. e&. Chemical carcinagenesis and mutagenesis I and II. Berlin: Springer-Verlag. 1990. 11. Foulds L. The experimental study of tumor progression: A review. Cancer Res 1054;14: 317-339. 12. Rowley JD. Molecular cytogenetics: Rosetta stone fcr understanding cancer-The 29th G. H. A. Clowes memorial award lecture. Cancer Res 1990.50:3816-3825. 13. Schimke RT.The search for early genetic events in tumorigenesis: An amplification paradigm. Cancer Cells 19902:149-351. 14. Pet0 R. Roe FJ. Lee PN. Levy L. Clack J. Cancer and ageing in mice and men. Br J Cancer 1975;32411-426. 15. Fearon ER. Vogelstein B. A genetic model for colorectal tumorigenesis. Cell 1990;61: 759-767. 16. Butterworth BE. SlagaTJ. Nongenotoxic mechanismsin carcinogenesis: Banbury Report. vol25. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press. 1987. 17. Wilson VL, SmiiR RA. Longoria J. Liotta MA, Harper CM, Harris CC. Chemical carcinogen-induced decreases in genomic 5-methyldeoxycytidlnecontent of normal human bronchial epithelial cells. Proc Natl Acad Sci USA 1987;84:3298-3301. 18. Lutz WK.Dose-response relationship and low dose extrapolation in chemical carci- nogenesis.Carcinogenesis 1990;11:1243-1247. 19. Swenberg JA. Richardson FC, Boucheron JA, et al. High- to lowdose extrapolation: Criticaldeterminants involved in the dose response of carcinogenicsubslulces.Envimn Health Perspect 1987;76:57-63. 20. Loeb LA. Mutator phenotype may be required for multistage carcinogenesis. Cancer Res 1991:51:3075-3079. 2 10 P r i r i c i p l c , s of'Curcinoytw~sisC: hemical 21. Lutz WK. I'.ndogcnous genotoxic agcnts and pros rsses as a basis of spontancuus carcinogenesis. Muvdt Res 1990;238:287-295. 42. Breimer LH. Molecular mechanisms of oxygen rarlisal carcinogenesis and mutagcncsis: The role of DNA base damage. Mol Carcinogcn :9903:188-197. 23. Weinstein IB. Mitogenesis is only one factor in carcinogenesis. Science 1991;251. 387-388. 24. Ames BN, Gold LS. Too many rodent carcinogens: Mitogenesis increases mutagenesis. Science 1990;249970-971. 25. Infante PF. Prevention versus chemophobia: A defense of rodent carcinogenicity tests [comments]. Lancet 1991:337:538-540. 26. Hay A. Carcinogenesis: Testing times for the tests [news[. Nature 1991.350:555-556. 27. Gonzalez FJ, Crespi CL. Gelboin HV. DXA-expressed human cytochrome P450s: A new age of molecular toxicology and human risk assessment. Mutat Res 1991;247: 113-127. 28. Nebert DW. Nelson ER. Coon MJ. et al. The P450 superfamily: Update on new sequences. gene mapping and recommended nomenclature. DNA Crll Biol 1991;lO:l20. 29. Gonzalez FJ. Molecular genetics of the P-450 superfamily. Pharmacol Ther 199045: 1-38. 30. Autrup H. Harris CC. Metatolism of chemical carcinogens by human tissues. In: Harris CC. A u t ~ pH. eds. Human carcinogenesis. Yew York: Academic Press, 1983169194. 31. Harris CC. Human tissues and cells in caxinogenesis research. Cancer Res 1987;47. 1-10, 32. Harris CC. lnterindividual variatior among humms in carcinogen metabolism. DNA adduct formation and DNA repaj:. Carcinogenesis 1989;10:1563-1566. 33. Hall M, Grover PL. Polycyclk aromatic hydrocxbons: Metabolism, activation. and tumour initiation. In: Cooper CS. Grover PL. eds. Chemical carcinogenesis and mutagenesis 1. Berlin: Sprinser-Verlag. 1990327-372. 34. Barbacid M. R4S genes. Annu Rev Biochem 1987:56:779-827. 35. Bouck N. Tumor angiogenesis: The role of oncogenes and tumor suppressor genes. Cancer Cells 1990;9:179-185. 36 Liotta LA. Steeg P S . Stetler-Stevenson WG: Cancer metastasis and angiogenesis: An imbalance of VJsitive and negative regulation. Cell I991:64:327-336. 37. Bishop JM: Molecular themes in oncogenesis. Crll 1991:64.235-248. 38 Brown K. Buchmann A. Balmain A. Carcinogen-induced mutations in the mouse c-Ha-RAS gene provide evidence of multiple p i t h \ .ns for tumor progression. Proc Natl Acad Sci USA 1990237538-542. 39. Newcomb EW. Diamond LE. Sloan SR.Corornin.~11.Guerrerro 1. Pellicer A. Radiation and chemical activation of R4S oncogenes in ditfr.rr~~miiouse strrins. Environ Health Perspect 1989;81:33-37 40. hlarshall CJ, Vousden KH. Phillips DH. .Acti\itioii oic-Ha-RIS-I proto-oncogene by in Yitro modification with a chemical cxcino~r.iik. ~ r u i d ) p ) r e ndeiol-epoxide. Nature 1984:310:586-589. 41. Sukumar S.An experimental analysis of c m c e r h i e uf RIS oncogenes in multistep carcinogenesis, Cancer Cells 199C:2:190-204 42. Bailleul B, Surani MA. White S . et ai. Shin h:perLrritosis and papilloma formation in transgenic mice expressing a R4S oncopene from .I supmb3sdl keratin promoter. Cell 1990;62:697-708. 43. Leder A. Kuo A. Cadiff RD. Smn E. Leder P \-HJ-S-+S :rmsgene abrogates the initiation step in mouse skin tumorigenesis: Effects o i Fi:zrki esters and retinoic acid. Proc Natl Acad Sci USA 199087:9178-918'2. 44. Knudson AG Jr Mutation and cancer: S r . ~ t i s t ~i:z.di\ of retinoblastoma Proc Natl Acad Sci USA 1971;68:820-823. 19903 I 18-121. 48 Herskowitz 1. Functional inactivation of gen 1987:329:219-222. 1-3. mt negnti\e mutations. Nature 1x1 canccr Culd Spring Harbor. Photubiol [BJ 1989.3:143-I55 i98a;i 1-18, 53 Skandalis A, Glickman B\V Endogenous ge.;? >\<:'?is for the study of mutational 54 Yale. JL. Chcn RH. hlaher YM. SlcCormicL !J ?,:;d.eic~ationof niut~tionsinduccd Iibrobl~sts.C~rc~nogrnes1i9s Y1:12.71-:~ monitoring iii hiimaris hy PS 1988.?04 48 1-492. 56. I uechlrs 1-1. (;st .::bull1 11110 .I l i i i i ~ \~iIc~Iin~.t' !is.11 qci11~i1ic' f-x\ \..id Sc.i l ' S 4 198-i.81 6271- 6275 57 Donigcr J. Novario V. DiPaolo JA. C a r c i n x c n ~with divcrse mutagenic activities initiate neoplastic guinea pig cells that acquirc the same N-RAS point mutation. J Biol Cheni 1987;262:3813-3819 58 Brookes P. Chemical carcinogens and PytS gene activation. Mol Carcinogen 19892: 305-307 59 Topal MD. DNA repair. oncogcncs and carcinogenesis. Carcinogenesis 19889691- 696. 60 Wood ML.Dizdaroglu M. Gajewski E. Essiqmann JM. Mechanistic studies of ionizing radiation and oxidative mutagenesis: Genetic effects of a single 8-hydroxyguanine (7-hydro-8-oxyguanine) residue insencd at a tiniqw site in a viral genune. Biochemistry I990;349:7024-7032. 61 Lindahl T. Nyberg B. Rate of depurination of native deoxyribonucleicacid. Biochemistry 1972;11.3610-3618. 62 Soussi T. Caron de Fromentel C. hlay P. Structural aspects of the p53 protein in relation to gene evolution. Oncogene 19905945-952. 63 Hollstein M, Sidransky D. Vogelstein B. Harris CC. P53 mutations in human cancers. Science 1991;253:49-53. 64 Hsu IC, Metcalf RA. Sun T.Welsh 1. Wang NJ. Hams CC. P53 gene mutational hotspot in human hepatocellular carcinomas from Qidong, China. Nature 1991;350 427-428. 65 Bressac B. Kew M, Wands J. Ozturk 11. Selective G to T mutations of p53 gene in hepatocellular carcinoma from southern Africa. Nature 1991;350:429-431. 66 !dcMahon G. Davis EF, Huber U.Kim Y. Wogan GN. Characterization of e-Ki-RAS and N-RAS oncogenes in aflatoxin B,-induced rat liver tumors. Prof Sat1Acad Sci USA 1990;87~1104-1108. 67 Hams CC. Hepatocellular carcinogenesis Recent advances and speculations. Cancer Cells 1990;2:146-148. 68 Committee on the Biological Effects of Ionizing Radiation National Research Council (BEIR IV): Health risks of radon. Washingon. DC: National Academy Press, 1988:l. 69 Saracci R. The interactions of tobacco smoking and other agents in cancer etiology. Epidemiol Rev 1987;9.175-193. 70 Jackson JH. Schraufstatter IU. Hyslop P.A. et al. Role of oxidants in DSA damage: Hydroxyl radical mediates the svnergisiic DS.4 damaging effects of asbestosand cigarette smoke. J Clin Invest 1987;80:1090-109j. 71 \lossman BT. Marsh JP. Evidence supportmi: d role for active o.wgenspecies in arbewos- induced toxicity and lung disease. Entiron Health Perspect 1989:81.91-94. 72 Kamp D\V. Dunne XI. Weitzman SA. D u ~ a5151. The interaction of asbestos and neu- trophils injures cultured human pulmonarr epithelial cells: Role of hydroqen peroxide. I Lab Clin Mcd 1989:114.604-612 73 llossman BT. Eastman .A. Landesrnan J I I Bresnick E. Effects of crocidolite and chry- sotile asbestos on celluldr uptake and rnerabolism of benzotaipyrene m hamster inched epithelial cells. Environ Health Perspect 1983.51.331-335. 74 Shields PC. Harris CC. Environmental c a s e s of cancer kled Chn Sonh .Am 19907.1: 263-277 -r I3 LaLe LB. Omcnn GS Cosl-effecri\enessoishort-terttnests for carcinogenicity Nature 1986;324:29-31. 76 Ashby J. hlorrod RS Detection of human crrcinogens. Sature 1991.352-185-186. 77 Ames BN. McCann J. Yamasaki E. l l e t h d s for detectin5 cnrcinoqens and mutagens irith the Salmo~rullo/rnammalian-micro.;ome mutagenicity test. \!ut31 Res 197531: 347-364. 78 Yamasaki E. Ames BN. Concentration of mutagens from unne by absorption nith the nonpolar resin XAD-2: Cigarette smokers hate mutagenic urine. Proc Sat1 .%cadSci CS.A 1977.74.3555-3559. 79 Santella RM. In vitro testing for carcinoeens and mutagens. In Brandt-Rauf PW. ed. Occuprtional medicine-state of t h e m re, i e ~ sPhiladelphia Hank\ 6 Belfus 1987 39-46 80 Late LB Enneter F k Roscnkranz HS O-nenn GS lnformaiion \slue of the rodent bioassay Nature 1988 336 631-633 81 4mes BN Magarv R Gold LS Rankinqpossibie carctnqenic hazards h e n c e 1987 2% 271-280 82 Lijmrkv \\ In tito testing lor carcinoceri~it~In Cooper CS Gro\er PL d s Chemical carcinogcncsir and niutagcncsts I Berlin Spnnger ierlag 1990 179-209 83 Tomatis L Bdrtsch H The contribution 01 expenmental studies to nd. s e a s m e n t of carcinogcnic agents in humans Exp Piihol 1990 40 251-266 8.1 Hill AB The cnrironmcnt and diseare -\.~iciationor causation Proc Rota1 Soc Med 196558295-300 85 Tennant RW hlargolin BH Shelby \ID et dl Prediction uf chemical carcinogenicitv in rodents from in iitro genetic to~icit\SAS Sclence 1987 236 933-941 86 Ashb\ J Tennant R W Definitire relattonships among chemicrl .iructure Carcino- genicitv rnd mutdgcnicit\ for 301 rheniirals tested b\ the L S \TP \lutai Res 1991 257229-306 87 Purchase II. Range uf c~peririicntrel \Idt.nce in rs~cssingpotential humin carcino- qcni~itbArch Toxicol Suppl 1980 3 283-2Y3 88 Ruc\ell hl Grubcr XI Ri5k rssc9rnicnt ,n en\irunmentdl polic\ makinc Science 1987 236286-290 89 Culliton RJ US go\crniiient orders neu look dioxin \aturc 1991 352 753 90 Shields PC Harrir CC Xlolecular epidcmiolop and the genetics of en\ironmental canccr J,\MA I99I 266 681-687 Y 1 \\oe.iii CN hl.irkers of c*xposuretu crrcinogens Emiron Health Penpect 1989.81 9-17 9 2 Xlen~chiniP Abhundandolo A Somdtic gene mutations in humans In Gledh~llBL Nauru F eds Trends in biological dosrrnern \ew turk \ \ ~ l c \ - L ~ s1s991 267-279 03 4lhertini RJ Robi\on SH Human population monitonng In LI AP Hefhch RH. eds. Genetic to\ico~ogv A treatisc Cdlducll \J Tcltord Press 1991 375-420 ~ s.Blological dosimetn uitn L\togenetic eildpuint\ In Glrdhill Bl. \Iauro k ends in biological dosimetn \eu York \\'ile)-l~iw 1991 351-362 K Quantitaii\e exaluatior of DUA binding data for risk estimation and for a t i o n of direct and indirpc: carcinogen5 J C.incer Re\ Clin Oncol 198h 1 I!? 121; -Ho@mannD. Hechi SS 4 d \ a n c ~ i;n tohacco carcinugenevr In C w p u i s Cro\cr p~ & Chemical carcinogenrs and m u t ~ g c i i r \Ii ~Hcrlin Spriiigrr \erla< IWO 'gj-102. wingJE. Phenow Kl Health effiils of involuntary snioking N Engl J hled 1988.319 1452-1 460. b8. Wu-Wi]liamsAH. Samer J\l E n r ironmental tobacco smoke Expusur~~-responsre- lationships in epidemiologc studic~ Risk Anal 1990.1039-48 99: D,,,,~ BP. Veda1 S. San RH. et d DSA adducts in bronclii.4 biopsies Int J Cancer 1ggl;48:485-492 Phillips DH. Hetrer .4. Slanin C\ Garner RC. King hlhl. Correlation of DSA adduct levels in human lung \tith c1gare::e smoking Nature 1988.336.790-792. lol,Genste 0.Camus ASl. Castegrdu 51. et a!. Comparison of pulmonan DSA adduct levels. measured b) "P~postlabel1injand aryl hydrocarbon hydroxylase activit! in lung pvenchymaof smokers and el-smokers Carcinogenesis 3991;12:3301-3305 lm,l m t t i A. RossiGA. Bagnasco 51 De Flora S. Bcnzolalpyrene dioleposide-DS.4adducts &mIvmacrophages of smohen Carcinogenesis 1991.12:1281-1285 103,Evenon RB. Randerath E. Santell R\I. Cefalu RC. Avitts TA. Randerdrh I; Detection ofsmoking-related co\alent D S i ddducts in human placenta Science 1986:231 54- 57. 104. Phillips DH. Schoket B Hewer i Bale! E. Kostic S. \'incze 1. Influence of cigarette smokingon the levels of Dh.4 ada-xts in human bronchial epithelium and v hite blood c&. Int J Cancer 1990.46569-575 1%. Chacko M. Gupta RC. E\aluation o i DNA damage in the oral mucosa of tobacco users &non-usen by nP-postlabeling ass+ Carcinogenesis 1988;9:2309-2313. 106. m e U a SG. Kagan SS. E a p n 1: et al Mass spectrometric analysis uf iobacco-specific nitrosamine hemoglobin addum in snuff dippers. smokers, and nonsmokers Cancer Res 1990.505438-5445. 107. Wilson VL. Basu AX. Essicmanr. J \ l Smith RA. Harris CC. 06-alkvldeos!guanosine detection by "P-postlabeling anl nucleotide chromatographic analysis. Cancer Res 1988;48:2I 56-2161 108. Shields PG. Povey AC. \\ilson VL \Veston A . Harris CC Combined high performance l i d chromatograph! /"P.pos:i&ling assay of Ni-niethyldeo~yyanosine.Cancer Res 3990506580-6984 109. P e l w M. Castegnaro Sl. \lala\ edit C et al. JIPostlabelling analysis of urinar)-mutagens fivm smokers of black tobacco implicates 2-aniino-l-methyl-G-phen~limidazoll.jblpyridine (PhlP) a s a major DUdamagingagent. Carcinogenesis 1991;12.713-717 110. Hanis CC. Vahakangas I; Yewnm \U.et al Detection of henzo[a]pyrene diolepoxideDNA adducts in peripheral b i o d I\ mphocytes and antibodies to the adducts in semm from coke oven \rorhers Proc Sal .4cad Sci USA 1985;82:6672-6676. 111. Van Schwten FJ. Van Leemen FE. Hillebrand hU. et al. Determination of benzolalpyrene diol epoxide-Dh i adducts in white blood cell DNA from coke-oven workers: The impact of smoking JSCI 1990;82:927-933. 112. Peren FP. Hemmtnki K. Young TL. Brenner D. Kelly G. Santella Rhl. Detection of polycyclicaromatic h\drocarbon-l)S-\ adducts in white blood cells of foundn workers. Cancer Res 1988;48:2286-22'1: 113. Bryant MS. Vineis P. Skipper PL. Tannenbaum SR. Hemoglobin adducts of aromatic amines:Associations ~ i t shmohng status and type of tobacco. Proc Nail Acad Sci USA 1988;859788-979I 114.He~~~miKn.ltGi rzybowskaE. C h o r q AI. et al. DNA adducts in humans environmentally 10 aromatic compounds in an mdustrial area of Poland. Carcinogenesis 1990.11: 1229-1231. 115. Phiuips DH. Hemminki K. Alhonen A, Hewer A, Grover PL. Monitoringoccupational exponve to carcinogens: Detection by 3*P-postlabelling of aromatic DNA adducts in white blood cells from iron found? workers. Mutat Res 1988;204:531-541. Il8. F O h PG. Miglietta Lhl. Quan .Ul.Quart E,Kabat GC. Hecht SS.Evaluation of "P- -ling analysis of DS.4 from exfoliated oral mucosa cells as a means of monitoring exposure of the oral cavity to genotoxic agents. Carcinogenesis 198910:1429-1434. 'I7. Rothman N. Poirier MC. h s e r \lE. et al. Formation of polycyclic ammatic hydrocarbonDNA d u c t s in peripheral white blood cells during consumption of charcoal-broiled 9 beef. Cveinogenesis 1990;I1:1241-1243. ,'I8. Mmhester DK. W'esion A. Choi JS. et al. Detection of benzolajpyrene diol epoxide- DNA d u c t s in human placenu. Proc Nail Acad Sci USA 1988;85:9213-9247. 'Is GmPman JD, Cain LG. Kensler M'.Aflatoxin exposure in human populations: Mea- surements and relationship to cancer. CRC Crit Rev Toxicol 1988:19113-145. CP.Jiang YZ.Sabbioni G. Chapot B, Montesano R. Evaluation of methods for W t i t a t i o n of aflatoxin-albumin adducts and their application to human exposure -ent. Cancer Res 199050245-251. Huh NH. Satoh MS, Shina J. hiewskv MF. Kuroki T. lmmunoanalvtical detection of ~ 04-echylthymme in liverDNA oi mdi\;duals with or without malignkt tumors Cancer 1989.49 93-97 'awh Weston A hlanchexer DK, et al Alkyl and aryl carcinogen adducts detected 'p inh u n peripheral lung Cdcmogenesis 1989,lO2149-2153 Hemon DC.Shank RC Meth\Lated purines in human liver DNA after probable dimethyhtrosamine pasontng Cancer Res 1980.4031 16-31 17 Foiles PG. Miglietta Lhl Akerkar SA Everson RB. Hecht SS Detection of OB'"%ldeoxyguanosine in human placental DNA Cancer Res 1988,484184-4188 "mbenhauer D.Wild CP Mom- R et al 0' methyldeoxyyanosine in oesophageal '), DNA mong individuals at high n s k of oesophageal cancer In1 J Cancer 1985.36 661-665 127 128 129 130. 131.Conney AH Induction of microsomal enzymes by foreign chcmicals and carcinogenesis by polycyclic aromatic hydrocarbons. l'hr C . ti A Clowes Memorial Lecture. Cancer Res 1982;42:4875-4917 132. Manchcster DK.Jacob) EH. Sensitivity of human placental monooxygena:e activity to maternal smoking. Clin Pharmacol Ther 1981;30687-692. 133. Nuwak D. Schmidt-Preuss U, Jorres R. Liehke F. Rudiger H W Formation of DNA adducts and water-soluble metabolites of benzolajpyrene in human monocvtes is genetically controlled. Int J Cancer 1988;41:169-173. 134 Kouri RE.hlcKinney CE. Slomiany DJ. Snodgras DR. Wray NP. McLemoreTL Positive correlation between high an1 hydrocarbon hvdroylase activity and primary lung cancer as analyzed in cryopresened lymphocytes Cancer Res 1982;42:5030-5037. 135. Farmer PB. Bailey E. Gorf Shl. et al. Monitoring human exposure to ethvlene oxide by the deiermination of haemoglobin adducts using gas chromatography-mass spectroniet ?.Carcinogenesis I986.7:637-640. 136. Hawke U.Farrell GC. Increased binding of benzolalpyrene metabolites IO lymphocytes from patients with lung cancer. Cancer Lett 1986:30:289-297. 137. Peterson DD. hlcKinne! CE. lkeya K. et al Human CYPIAI gene: Co-segregation of the enzyme inducible phenotype and a restriction fragment length polymorphism. Am J Hum Genet 1991.48:720-725. 138 McLemore TL. Adelherg S. Liu MC. McMahon NA. Expression ofcytochrome P4501.41 gene in lung cancer patients. Evidence for cigarette smoke-induced expression in normal lung and altered gene regulation in primary pulmonary carcinomas. JNC1 1990:82 1333-1339 139 Kawajiri K. Nakachi I;. lmai K. Yoshii A. Shincda N. Watanahe J . Identification of geneticall! high risk indibiduals to lung cancer by DNA polymorphisms of the cytochrome P4501A1 gene FEBS Lett 1990:263.131-133 140. Crespi CL.Penman R\V. Gelboin HV. Gonzalez FJ A tobacco smoke-derived nitrosamine. 4-(meth! Initrosamino)-l-(3-pyridyl)-l-hutanonies.activated by multiple human cvtochrome P450s including the polyniorphic human cytochrome P4502D6.Carcinogenesis 1991.12:1197-1201. 141. Idle JR. Mahgoub A . Sloan TP. Smith RL. Mbanefo CO. Bababunmi EA. Some observations on the oxidation phenotype status of Nigerian patients presenting with cancer. Cancer Lett 1981.1 1.331-338. 142. Kaisary A. Smith P. Jaczq E. et a1 Genetic predisposition to bladder cancer: Ability to hydroxvlate debrisoquine and mephenytoin as risk factors. Cancer Res 1987;47:54885493. 143. Caporaso N. H a y s RB. Dosemeci M , et al. Lung cancer risk. occupational exposure, and the debrisoquine metabolic phenotype. Cancer Res 1989;49:3675-3679. 144. Beland FA, Kadlubar FF. Metabolic activation and DNA adducts of aromatic amines and nitroaromatic h!drocarhons In: Cooper CS, Grover PL. eds. Chemical carcinogenesis and mutagenesis I. Berlin: Springer-Verlag. 1930:268-325. 145. Lang NP. Chu DZ, Hunter CF. Kendall DC. Flammang TJ. Kadlubar FF. Role of aromatic amine acet)ltransferase in human colorectal cancer. Arch Surg 1986;121: 1259-1261. 146. llett KF, David BM, Detchon P, Castleden H'M, Kwa R. Acetylation phenotype in coloreetal carcinoma. Cancer Res 1987;47:;466-1469. 147.Ladero JM, Gonzalez JF, Benitez J. et al. Acetylator polymorphism in human colorectal carcinoma. Cancer Res 1991;51:2098--2100 148. Jerina DM. Bend JA. Glutathione S-transferases In: lollow D.Kocsis JJ, Synde- R. Vaino H, eds. Biological reactiv? metabolites. New Yorl:: Plenum Press, 1977:207236. 149. Robertson IGC. Guthenherg C. Mannervik U,Jernstrom B. The glutathione conjugation of benzo(a)pyrenediol-epoxide by human glutathione transferases. In:Cooke M, Dennis AJ,eds. Polynuclear aromatic hydrocarbons: A decade of progress. Columbus, OH: Battelle Press. 1988:799-808. 150. Mannervik B. The isoenzymes of glutathione transferase. Adv Enzymol 1985;57:357- 417. 151. Seidegard J, Per0 RW. The hereditary transmission of high glutathione transferase activity towards trans-stilbene oxide in human mononuclear leukocytes. Hum Genet 1985;6966-68 152. Seidegard J. Pero RW. Miller DG, Beattie Ei. A glutathione transferase in human leukocytes as'$ marker for the susceptibility to lung cancer. Carcinogenesis 1986;7: 751 -753 153. Seidegard J, Per0 R\V, Markowitz MM. Roush 6 .Miller DG. Beattie Ei.Isoenzyme(s) of glutathione transferase (class Mu) as a marker for the susceptibility to lung cancer: A follow up study. Carcinogenesis 1990;11:33-36. 154. Malkin D. Li FP. Strong LC, et al. Germ line p53 mutations in a familial syndrome of hrcast cancer, sarcomas. and other neoplasms. Science 1990;250:1233-1238. 155. Srivastava S. Zou ZQ. Pirollo K. Blattner WA, Chang EH. Gcrmline transmission of a mutated p53 in a cancer-prone hniily with Li-Fraumeni syndrome. Nature 1990;348: 747-749. 2 12 Principles of Carciwgenesis: Chemical 156. Sanders BM. Jay M. Draper GJ. Roberts EM. No i ocular cancer in relatives of rctinoblastoma patients. Br J Cancer 198960:358-335. 157. KrontirisTG.DiManino NA. Colb M. Parkinson DR. Unique allelic restriction fragments of the human Ha-RAS locus in Icukocyte and tumcwr DNAs of cancer patients. Nature 1985;313:369-374. 158. Sugimura H. Caporaso NE. Hoover RN, et al. Association of rare alleles of the Harvey RAS protooncogene locus with lung cancer. Cancer Res 1990;50:1857-1862. 159. Vineis P. Caporaso N. The analysis of restriction fragment length polymorphism in human cancer: A review from an epidemiologicalperspective. Int J Cancer 1991;47: 26-30. 160. Kawashima K, Shikama H, lmoto K. et al. Close correlationbetween restrictionfragment length polymorphism of the L-MYC gene and metastasis ofhuman lung cancer to the lymph nodes and other organs. Proc Natl Acad Sci USA 1988;85:2353-2356. 161. lshizaki K, Kato M. Ikenaga M. Hunda K. Ozawa K. Toguchida J. Correlation of LMYC genotypes to metastasis of gastric cancer and breast cancer. JNCl 1990;82 238-239. 162. Tamai S . Sugimura H, Caporaso NE, et al. Restriction fragment length polymorphism analysis of the L-MYC gene locus in a case-control study of lung cancer. lnt J Cancer 1990;46:41 1-415. 163. Setlow RB. Variations in DNA repair among humans. In: Harris CC. Autrup H, eds. Human carcinogenesis. New York: Academic Press, 1983231-254. 164. Hoeijmakers JHJ.Bootsma D. Molecular genetics of eukaryotic DNA excision repair. Cancer Cells 1990;2:?11-320. 165. Lindahl T. Wood RD. Karran P. Molecgiar deficiencies in human cancer-prone syndromes associatedwith hypersensitivityto DNA damaging agents. In: Brugge J. Curren T. Harlow E. McCormick F. eds: Origins of human cancer. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press. 1991:163-170. 166. Per0 RW. Bryngelsson C. Blyrgelsson T, Norden A. A genetic componentof the variance of N-acetoxy-Z-acetylarninJ8uorene-inducedDNA damage in mononuclear leukocytes determined by a twin study. Hum Genet 1983;65:181-184, 167. Pero RW.Johnson DE, Markowitz M, et al. DNA repair synthesis in individualswith and without a family history of cancer. Carcinogenesis 1989;10693-697. 168. Scudiero DA. Me:er SA. Clatterbuck BE, Mattem MR. Ziolkowski CH. Day RS.Sen- ty of humaii cell strains having different abilities to repair 0'-methylguanine in DNA to inactivation by alkylating agents including chloroethylnitrosoureas.Cancer Res 1984;44:2467-2474. 169. Grafstrom RC. Pegg AE. Trump BF. HarrisCC. O'-alkylguanine-DNA alkyltransferase activity in normal human tissues and cells. Cancer Res 1984;44:2855-2857. 170. Yarosh DB. The role of O'methylguanine-DNA methyltransferase in cell survival, mutagenesis and carcinogenesis. Mutat Res 1985;145:I -16. 171, Grafstrum RC. Fornace AI Jr. Autrup H, Lechner JF, Harris CC. Formaldehyde damage to DNA and inhibition of DNA repair in human bronchial cells. Science 1983;220: 216-218. 172. Sagher D. Karrison T. SchwartzJL. Larson R. Meier P. Strauss B. Low W-alkylguanine DNA alkyltransferase activity in the peripheral blood lymphocytes of patients with therapy-related acute nonlymphocytic leukemia. Cancer Res 1988483084-3089. 173. Rudiger HW. Schwartz U.Serrand E. et al Reduced 0 6 - m e t h y l ~ e r e p i r i fnibroblast cultures from patients with lung cancer. Cancer Res 1989;495623-5626. 174. Ludlum DB. Therapeutic agents as potential carcinogens. In: Cooper CS, Grover PL, eds. Chemical carcinogenesis and mutagenesis 1. Berlin: Springer-Verlag.1990153- 175. 175. Casciato DA, Scott JL. Acute leukemia following prolonged cytotoxic agent therapy. Medicine (Baltimore) 1979;58:32-47. 176. Marselos M.Vainio H. Carcinogenic properties of pharmaceutical agents evaluated in the IARC monographs programme. Carcinogenesis 1991;12:1751-1766. 177. International Agency for Research on Cancer. hfonographs on the evalwuion of car- cinogenic risks to humans: Overall evaluations of carcinogenicity-an updating of lARC monographs volumes 1 to 42. Lyon: International Agency for Research on Cancer. 1987. 178. Vyas RC, Adhvaryu SG. Shah VC. Effects of CCNU therapy on human chromosomes. Mutat Res 1988;206:163-166. 179. Pedersen-BjergaardJ. Osterlind K. Hansen \I. Philip P. Pedemn AG. Hansen HH. Acute nonlymphocytic leukemia, preleukemia. and solid tumors following intensive chemotherapy of small cell carcinoma of the lung. Blood 1985;66:1393-1397. 180. Kaldor JM. Day NE. Clarke EA. et al. Leukemia followingHodgkin's disease (comments]. N Engl J Med 1990;3227-13. 181. Val- P. Santoro A, Fossati-BellaniF. Banfi A. Bonadonna G. Second acute leukemia and other malipmcies following treatment for H o d w s disease.J Clin Oncol 19864: 830-837. 182. Parkin DM. Laara E. Muir CS. Estimates of the worldwide frequency of sixteen major cancers in 1980. Int J Cancer 1988;41:184-197. i83. Tomatis L. Aitio A, Wilboum J. Shuker L. Human carcinogens so far identified. Jpn J Cancer Res 198980795-807.