Figure 7-1 Chemical structures of selected carcinogens.
Causes of Cancer
Epidemiology and Causal Criteria
The landmark findings of Kennaway, Avery, Hershey, Chase, and others in the early part of the 1900s guided a period of rapid advancement in the laboratory concerning the molecular basis of cancer. In contrast, the discovery of specific human carcinogens has been largely guided by epidemiologic studies of cancer incidence. The study of worldwide cancer incidence patterns, including analysis of cancer risk among migrant populations, has confirmed the critical role of environment in determining cancer risk. Studies of exposure cohorts and observational studies of cancer incidence have been especially crucial in the identification of the biologic, physical, and chemical agents capable of causing cancer. Similarly, epidemiologic studies have revealed numerous lifestyle choices and socioeconomic factors associated with increased risk of cancer.
Cancer risk is known to vary extensively worldwide. 4 For instance, liver cancer risk varies 20- to 40-fold internationally; the incidence is highest in eastern Asia and lowest in northern Europe and Central America. Prostate cancer rates are high in the United States, Canada, and Scandinavia, especially in comparison with the rates in China and other Asian countries. Similarly, breast cancer risk has historically been higher in the United States and European countries than in Asia, Africa, and South America. These observations suggest that (1) genetic differences among ethnic groups alter cancer risk and/or (2) differences in environmental exposures among geographic locations affect the risk of developing cancer.
Capitalizing on known ethnic variation in cancer rates, analysis of cancer risk in migrant populations has been undertaken and has yielded important information concerning the relative contribution of environment versus genetics in cancer etiology. In these studies, the rate of cancer in migrant cohorts is compared with the rate of cancer among people of the same ethnicity living in the country of origin and to the cancer rate of people in the destination population. For example, breast cancer incidence among Asian immigrants to the United States has been compared with that of women still living in their country or region of origin. 5 The breast cancer risk of Asian American women born in the East has been shown to rise with increasing number of years lived in the West. Ultimately, the risk of breast cancer among Asian American women approaches that of U.S.-born White women and is significantly higher than that of Asian women still living in the country of origin. Numerous studies of this kind demonstrate that even while in the first generation following relocation, immigrant populations assume a pattern of cancer risk in common with native populations rather than with populations in their country of origin. These studies imply that environmental factors play a significant role in determining cancer risk. Similarly, studies of cancer risk in twins have suggested the importance of environmental factors in determining overall cancer risk.
Recent population-based evidence further underscores the overall importance of environmental factors in determining cancer risk. Cancers that were once associated with affluence and/or the Western lifestyle are on the rise in less developed countries. Rates of colon, breast, and lung cancers in developing countries have increased as their economies have transitioned. 4,6 Multiple factors likely contribute to this trend, including non-genetically controlled influences such as tobacco use, diet, and physical activity.
In addition to population-based evidence, case-control and cohort studies have been used to identify specific environmental agents and factors that are now considered to be human carcinogens. To assess the likelihood that a particular environmental exposure is causally linked to cancer, epidemiologic data are interpreted in the context of mechanistic data and other considerations. The strength of evidence for a causal role in cancer development is evaluated using criteria developed as a modification of Bradford-Hill’s criteria (1965) for assessment of evidence of causation 7 :
1. Strength of Association: Large-magnitude effects on cancer risk are less likely than small-magnitude effects to be due to chance.
2. Temporal Relationship: To be causal, the environmental exposure must have happened in advance of the appearance of cancer.
3. Biologic Plausibility: Relationships that can be supported by laboratory evidence or a plausible mechanistic hypothesis are more likely to be causal relationships.
4. Dose-Response Relationship: Studies that demonstrate a gradient in disease outcome whenever a gradient in exposure has occurred provide stronger support for a causal relationship than those studies that do not demonstrate a positive correlation between dose and response.
5. Consistency: The most probable causal relationships are consistently demonstrated in multiple studies of the exposure-disease relationship.
Using these criteria, numerous cancer-causing agents and/or risk factors have been identified for further characterization.
Known Cancer Risk Factors
In a landmark paper published in 1981, Doll and Peto summarized available epidemiologic data to estimate the percentage of U.S. cancer deaths attributable to a variety of environmental and lifestyle influences. Their analyses suggested that as many as 60% of all cancer deaths could be attributed to two environmental factors: diet and tobacco use. 8 More than 30 years later, these estimates appear to remain valid; diet and tobacco use continue to be primary determinants of cancer mortality. Additional factors cited by multiple investigators and regulatory agencies as contributing to cancer risk include occupation, radiation, alcohol, pollution, infections, medications, and reproductive and socioeconomic factors.
Smoking
Tobacco use remains the single most important and avoidable factor in determining cancer risk. 9,10 Smoking is estimated to contribute to at least 30% of all cancer deaths. Lung, bladder, esophageal, pancreatic, uterine, oral, and nasal cavity cancers, among others, have all been associated with tobacco use. Approximately 90% of all lung cancer deaths can be attributed to smoking. Lung cancer risk is greatest for persons who begin smoking at an early age and continue smoking for many years, and the risk of tobacco smoke–induced lung cancer is directly proportional to the dose inhaled. Tobacco smoke is a complex mixture of chemicals, 55 of which are known or suspected human carcinogens (Table 7-1 ). On absorption in the lungs, these agents may act locally or at distal sites to (1) induce DNA damage and (2) alter cellular growth and proliferation. A synergistic effect has been noted in the case of combined tobacco use and heavy alcohol use. Despite antitobacco sentiment, approximately one fifth of U.S. citizens are still smokers, and smoking rates in countries such as China remain high; therefore, smoking-induced cancers are likely to continue to be prevalent worldwide.
Table 7-1
Carcinogens in Tobacco Smoke
| Carcinogen Class | No. of Compounds | Example Compound |
| Polycyclic aromatic hydrocarbons | 10 | Benzo[a]pyrene 5-Methylchrysene Dibenz[a,h]anthracene |
| Aza-arenes | 3 | Dibenz[a,h]acridine |
| N-nitrosamines | 7 | 4-(Methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) N-Nitrosodiethylamine |
| Aromatic amines | 3 | 4-Aminobiphenyl |
| Heterocyclic amines | 8 | 2-Amino-3-methylimidazo[4,5–f]quinoline |
| Aldehydes | 2 | Formaldehyde |
| Miscellaneous organic compounds | 15 | 1,3-Butadiene Ethyl carbamate |
| Inorganic compounds | 7 | Nickel Chromium Cadmium Arsenic |
| Total | 55 |
Adapted from Hecht SS. Tobacco smoke carcinogens and lung cancer. J Natl Cancer Inst 1999;91:1194.
Diet
The effects of diet on cancer risk have been attributed both to dietary chemical constituents and to overall energy consumption. As many as 14% to 30% of cancer deaths have been attributed to overweight and obesity. Overweight and obesity, as defined by the ratio of weight to height known as body mass index (BMI), are prevalent at epidemic proportions in the United States and other developed countries. Overweight and obesity have been associated with elevated risk of cancers of the colon, breast, endometrium, kidney, liver, pancreas, gallbladder, ovary, cervix, rectum, and esophagus as well as risk of non-Hodgkin’s lymphoma and multiple myeloma. In addition, animal studies have consistently demonstrated that restricting calorie intake can significantly reduce cancer risk, whereas inducing obesity can significantly elevate cancer risk. Despite these findings, a complete understanding of the mechanistic basis for the effect of dietary energy balance status on cancer formation is not conclusively known. 11 Elevated steroid hormone production in adipose tissue has been proposed as the basis for obesity-induced endometrial and breast cancers; adipose-derived leptin, adiponectin, and proinflammatory molecules may affect cancer development more broadly. Recent studies have suggested that alterations in circulating insulin-like growth factor 1 (IGF-1) levels may account for some of the effects of altered dietary energy balance status on cancer risk. 12
In addition to excess calorie intake, certain dietary constituents may affect cancer risk. 13 In the United States, cancer risk due to food additives is presumed to be quite low because the U.S. Food and Drug Administration (FDA) strictly regulates food additive use. In 1958, an amendment to the Food, Drugs, and Cosmetic Act of 1958, referred to as the Delaney Clause, was approved and stated that “the Secretary (of the FDA) shall not approve for use in food any chemical additive found to induce cancer in man, or, after tests, found to induce cancer in animals.” Presumably, therefore, cancer risk due to food additive consumption is quite low. Nonetheless, inadvertent food contaminants such as the plasticizer bisphenol A remain a source of concern. Bisphenol A is a weak endocrine-disrupting agent that has been associated with a variety of health effects including increased cancer risk. Fungal toxins such as aflatoxins are food contaminants resulting from mold growth on foodstuffs. Several of these toxins have been shown to be extremely potent mutagens and in some cases potent carcinogens (e.g., aflatoxin B1 [AFB1]). Red meat consumption has been associated with elevated colorectal cancer risk, possibly due in part to the carcinogenic nitrosamine and heterocyclic amine content of preserved or heat-treated meats.
Although examples of carcinogenic dietary constituents can be identified, a possibly greater dietary determinant of cancer risk is consumption of anticarcinogenic fruits and vegetables. Consumption of fruits and vegetables has consistently been linked to reduced cancer risk for a variety of cancer types. Fruits and vegetables contain numerous antioxidant compounds, which may guard against oxidative DNA damage or other forms of carcinogenic assault. In fact, tea phenols such as epigallocatechin-3-gallate (EGCG), the turmeric component curcumin, grape-derived resveratrol, and lycopene from tomatoes are all proposed cancer preventive agents. 14 On the other hand, excess consumption of herbal health supplements is an emerging dietary concern due to their widespread use in the absence of proper validation or safety assessment. As an example, renal failure was noted in women who consumed weight-reducing Chinese herbal pills. The pills were inadvertently substituted with a nephrotoxic herb, Aristolochia fangchi, containing aristolochic acids. Aristolochic acids are mutagenic and carcinogenic, and a high rate of urothelial carcinoma was noted in the population of women who consumed these pills. 15
Occupation
Many carcinogens have been identified at the cost of human exposure and cancer incidence that occurred as a result of industrialization. Human epidemiologic studies highlight the potency of chemical and physical carcinogens and how lack of understanding leads to lack of preparation and protection. 16–18 In the 1800s, high incidence of bladder cancer among workers in the aniline dye industry was recognized. Later, evidence was reported demonstrating that 2-napthylamine and benzidine were two carcinogenic agents responsible for this unusual cancer incidence. Also during the early 1900s, nearly 5000 workers were hired to apply luminous radium-containing paint to watch and instrument dials. Because of their occupational radiation exposure and a lack of precautionary practices, a large excess of bone cancers was noted among this cohort. Thousands of workers were exposed to vinyl chloride before its ability to induce angiosarcoma of the liver was recognized. Since the 1970s, strict workplace regulations and protective measures in the United States have largely prevented such dramatic incidents. The Occupational Safety and Health Administration (OSHA) was signed into existence in 1970 by the U.S. government with the goal of ensuring worker safety and health by improving the workplace environment. OSHA sets the legal limit for worker exposure to hazardous compounds in the United States. These limits are referred to as permissible exposure limits (PELs). PELs have been issued for approximately 500 chemicals, a portion of which are known or suspected carcinogens. Also created in 1970, the Environmental Protection Agency (EPA) is charged with protecting human health and the environment. In addition to other roles, the EPA regulates the release of industrial pollution, including carcinogens. Before these institutions were in place, employment in a wide variety of settings was linked to elevated risk of numerous cancers (Table 7-2 ).
Table 7-2
Environmental Carcinogens Associated with Occupation
| Occupation | Carcinogen Exposure | Associated Cancer Type |
| Iron and steel founding | PAH, chromium, nickel, formaldehyde | Lung |
| Copper mining and smelting | Arsenic | Skin, bronchus, liver |
| Underground mining | Radon (ionizing radiation ) | Lung |
| Aluminum production | PAH | Lung |
| Coke production | PAH | Lung, kidney |
| Painting | Chromium, solvents | Lung |
| Furniture and cabinet making | Wood dust | Nasal sinus |
| Boot and shoe manufacture | Leather dust, benzene | Nasal sinus, leukemia |
| Rubber industry | Aromatic amines, solvents | Bladder, leukemia |
| Nickel refining | Nickel | Nasal sinus, bronchus |
| Vinyl chloride manufacture | Vinyl chloride | Liver |
| Dye and textile production | Benzidine-based dyes | Bladder |
PAH, Polycyclic aromatic hydrocarbons.
Despite regulatory measures, occupational exposure to carcinogens continues. In the U.S. President’s Cancer Panel Report of 2008-2009, 19 members highlighted 14 types of environmental contaminants from industrial, manufacturing, and agricultural sources (polyhalogenated biphenyls, asbestos, chromium, perchloroethylene/trichloroethylene, particulate matter, mercury, formaldehyde, endocrine-disrupting chemicals, atrazine, DDT, nitrogen fertilizers, phosphate fertilizers, and veterinary pharmaceuticals) due to their cancer-causing potential and high probability of human exposure. The group estimated that millions of workers continue to be exposed to high levels of these and other agents each year. The families of exposed workers also experience higher than average exposure due to home contamination and may be at elevated cancer risk. As examples, chromium used in leather tanning, manufacture of dyes and pigments, wood preservation, and chrome plating is an established risk factor for lung cancer. Perchloroethylene, heavily used in dry-cleaning businesses, is classified as a probable carcinogen by the International Agency for Research on Cancer (IARC), and formaldehyde (a group 1 human carcinogen) is a synthetic starting material in manufacturing and a widely used disinfectant and preservative.
Causes of Cancer by Organ Site
Although an extensive list of known human carcinogens has been collected, the cause of many common cancers is still unknown. As shown in Table 7-3 , gastric, liver, and cervical cancers are each clearly linked with biologic carcinogens: Helicobacter pylori, hepatitis B virus (HBV), and human papillomavirus (HPV), respectively. The vast majority of lung cancer cases can be linked to tobacco use, and mesothelioma incidence is strongly correlated with exposure to asbestos. In contrast, the causes of most brain, pancreas, and prostate cancers remain largely unknown. For many other cancer types such as bone cancers, relatively rare exposures have been causally linked to incidence, yet the associated attributable risk is quite low. The remainder of cases continues to be largely unexplained. In general, linking particular cancers to specific exposure events can be problematic, and further work is necessary to uncover the primary causes of a significant number of cancers. Limiting factors include the inability to accurately estimate exposure dose and duration and a lack of understanding of combinatorial effects in multi-exposure events and finally lack of adequate biomarkers of exposure.
Classes and Types of Carcinogens
Carcinogen Evaluation and Classification
The U.S. National Toxicology Program (NTP), the World Health Organization’s International Agency for Research on Cancer (IARC), the U.S. EPA, and other agencies characterize and report the carcinogenicity of environmental agents and other factors (including drugs). Each entity independently evaluates the available evidence to rate the cancer-causing potential of a chemical, chemical mixture, occupational exposure, physical agent, biologic agent, or lifestyle factor. The most frequently referenced database is the IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. IARC defines carcinogens as agents “capable of increasing the incidence of malignant neoplasms, reducing their latency, or increasing their severity or multiplicity.” Agents are selected for evaluation on the basis of two factors: (1) evidence of potential carcinogenicity and (2) known exposure of humans. During the scientific review and evaluation of potential carcinogens, a working group is formed and charged with summarizing available data concerning anticipated exposure levels, human epidemiologic data, and studies of cancer-producing capacity in animals. Although the goal of the IARC Monographs has been to identify carcinogens regardless of an explanatory mechanism, information on mechanisms can also be used as supporting data. All agents evaluated by IARC are classified into one of five categories as shown in Table 7-4 . As of the most recent report, 108 agents, groups of agents, or exposure scenarios are listed as “Carcinogenic to Humans” (a partial listing is shown in Table 7-5 ). An additional 64 are listed as “Probably Carcinogenic to Humans.” These agents are extremely diverse in structure, potency, and mechanism.
Table 7-3
Exposures Associated with Human Cancers, as Identified by the IARC (Partial Listing)
| Cancer Site | Carcinogenic Agents with Sufficient Evidence in Humans | Agents with Limited Evidence in Humans |
| Oral cavity | Alcohol, betel quid, HPV, tobacco smoking, smokeless tobacco | Solar radiation |
| Stomach | Helicobacter pylori, rubber production industry, tobacco smoking, x-rays, gamma radiation | Asbestos, Epstein-Barr virus, lead, nitrate, nitrite, pickled vegetables, salted fish |
| Colon and rectum | Alcohol, tobacco smoking, radiation | Asbestos, Schistosoma japonicum |
| Liver and bile duct | Aflatoxins, alcohol, Clonorchis sinensis, estrogen-progestin contraceptives, HBV, HCV, Opisthorchis viverrini, plutonium, thorium-232, vinyl chloride | Androgenic steroids, arsenic, betel quid, HIV, polychlorinated biphenyls, Schistosoma japonicum, trichloroethylene, x-rays, gamma radiation |
| Pancreas | Tobacco smoking, smokeless tobacco | Alcohol, thorium-232, x-rays, gamma radiation, radioiodines |
| Lung | Tobacco smoking, aluminum production, arsenic, asbestos, beryllium, bis (chloromethyl) ether, chloromethyl methyl ether, cadmium, chromium, coal combustion and coal tar pitch, coke production, hematite mining, iron and steel founding, MOPP, nickel, painting, plutonium, radon, rubber production, silica dust, soot, sulfur mustard, x-rays, gamma radiation | Acid mists, manufacture of glass, indoor emissions from household combustion, carbon electrode manufacture, chlorinated toluenes and benzoyl chloride, cobalt metal with tungsten carbide, creosotes, engine exhaust, insecticides, dioxin, printing processes, welding fumes |
| Skin—melanoma | Solar radiation, UV-emitting tanning devices | |
| Other skin cancers | Arsenic, azathiopurine, coal tar pitch, coal tar distillation, cyclosporine, methoxsalen plus UVA, mineral oils, shale oils, solar radiation, soot, x-rays, gamma radiation | Creosotes, HIV, HPV, nitrogen mustard, petroleum refining, UV-emitting tanning devices |
| Mesothelioma | Asbestos, erionite, painting | |
| Breast | Alcohol, diethylstilbestrol, estrogen-progesterone contraceptive and menopausal therapy, x-rays, gamma radiation | Estrogen menopausal therapy, ethylene oxide, shift work resulting in circadian disruption, tobacco smoking |
| Uterine cervix | Diethylstilbestrol (exposure in utero), estrogen-progestogen contraception, HIV, HPV, tobacco smoking | Tetrachloroethylene |
| Ovary | Asbestos, estrogen menopausal therapy, tobacco smoking | Talc-based body powder, x-rays, gamma radiation |
| Prostate | Androgenic steroids, arsenic, cadmium, rubber production industry, thorium-232, x-rays, gamma radiation, diethylstilbestrol (exposure in utero) | |
| Kidney | Tobacco smoking, x-rays, gamma radiation | Arsenic, cadmium, printing processes |
| Urinary Bladder | Aluminum production, 4-aminobiphenyl, arsenic, auramine production, benzidine, chlornaphazine, cyclophosphamide, magenta production, 2-naphthylamine, painting, rubber production, Schistosoma haematobium, tobacco smoking, toluidine, x-rays, gamma radiation | Coal tar pitch, coffee, dry cleaning, engine exhaust, printing processes, occupational exposures in hair dressing and barbering, soot, textile manufacturing |
| Brain | X radiation, gamma radiation | |
| Leukemia and/or lymphoma | Azathiopurine, benzene, busulfan, 1,3-butadiene, chlorambucil, cyclophosphamide, cyclosporine, Epstein-Barr virus, etoposide with cisplatin and bleomycin, fission products, formaldehyde, Helicobacter pylori, HCV, HIV, human T-cell lymphotropic virus type 1, Kaposi’s sarcoma herpesvirus, melphalan, MOPP, phosphorus-32, rubber production, semustine, thiotepa, thorium-232, tobacco smoking, treosulfan, X radiation, gamma radiation | Bischloroethyl nitrosourea, chloramphenicol, ethylene oxide, etoposide, HBV, magnetic fields, mitoxantrone, nitrogen mustard, painting, petroleum refining, polychlorophenols, radioiodines, radon-222, styrene, teniposide, tetrachloroethylene, trichloroethylene, dioxin, tobacco smoking (childhood leukemia in smokers’ children) |
HBV, Hepatitis B virus; HCV, hepatitis C virus; HIV, human immunodeficiency virus; HPV, human papillomavirus; IARC, International Agency for Research on Cancer; MOPP, mustargen-oncovin-procarbazine-prednisone chemotherapy; UVA, ultraviolet A light.
Adapted from Cogliano et al. Preventable exposures associated with human cancers. J Natl Cancer Inst 2011;103:1835.
Table 7-4
IARC Classification of Suspected Carcinogenic Agents
Group 1: Carcinogenic to humans: Sufficient evidence of carcinogenicity in humans exists or sufficient evidence of carcinogenicity in animals is supported by strong evidence of a relevant mechanism of carcinogenicity in humans.
Group 2A: Probably carcinogenic to humans: Limited evidence of carcinogenicity in humans exists but sufficient evidence of carcinogenicity in animals has been demonstrated. Alternatively, inadequate evidence in humans with sufficient evidence in animals may be supported by strong evidence that a similar mechanism of carcinogenicity would occur in humans.
Group 2B: Possibly carcinogenic to humans: Limited evidence of carcinogenicity in humans exists but inadequate evidence in experimental animals. Alternatively, this classification can be used for agents for which there are inadequate data in humans but sufficient evidence in animals or strong mechanistic data.
Group 3: Unclassifiable as to carcinogenicity in humans: Inadequate evidence in humans and animals exists. Alternatively, sufficient evidence of carcinogenicity may exist in animals but strong mechanistic data predict a lack of carcinogenicity in humans.
Group 4: Probably not carcinogenic to humans: Evidence suggesting a lack of carcinogenicity in humans and animals exists.
IARC, International Agency for Research on Cancer.
Types of Carcinogens
Carcinogens can be grouped into one of three categories according to their composition: (1) physical carcinogens, (2) biologic carcinogens, and (3) chemical carcinogens. The term physical carcinogen encompasses multiple types of radiation (e.g., ultraviolet [UV] and ionizing radiation). Biologic carcinogens refer to viral and bacterial infections that have been associated with cancer development (e.g., human papillomavirus [HPV] and hepatitis B virus [HBV]). Most carcinogens can be categorized as chemical carcinogens. As examples, heavy metals, organic combustion products (e.g., B[a]P), hormones, and fibers (e.g., asbestos) are considered to be chemical carcinogens. Note that in the discussion that follows, only selected carcinogens that are known to be carcinogenic in humans are described (see Table 7-5). For a more comprehensive listing of carcinogenic agents, including those listed in other IARC categories, refer to the WHO IARC monograph database (http://monographs.iarc.fr/ENG/Monographs/PDFs/index.php) and additional references. 20,21
Physical Carcinogens
Examples of physical carcinogens include UV and ionizing radiation. Radiation refers to flow of energy-bearing particles; ionizing radiation refers to radiation that is of sufficiently high energy to remove an electron from an atom or molecule with which it collides. Exposure to ionizing radiation of various forms has been shown to cause multiple types of cancers. In addition, solar radiation is of sufficient energy to elicit photochemical damage to the skin, ultimately leading to cancer formation.
Table 7-5
Selected IARC Known Human Carcinogens
| 4-Aminobiphenyl | Hepatitis B virus |
| Arsenic | Hepatitis C virus |
| Asbestos | Human immunodeficiency virus type 1 |
| Azathioprine | Human papillomavirus |
| Benzene | Human T-cell lymphotropic virus |
| Benzidine | Melphalan |
| Benzo[a]pyrene | 8-Methoxypsoralen |
| Beryllium | Mustard gas |
| N,N-Bis(2-chloroethyl)-2-naphthylamine | 2-Naphthylamine |
| Bis(chloromethyl)ether | Nickel compounds |
| Chloromethyl methyl ether | N′-Nitrosonornicotine (NNN) |
| 1,4-Butanediol dimethanesulfonate | Phosphorus-32 |
| Cadmium | Plutonium-239 |
| Chlorambucil | Radioiodines |
| 1-(2-Chloroethyl)-3-(4-methylcyclohexyl)-1-nitrosourea | Radium-224 |
| Chromium[VI] | Radium-226 |
| Cyclosporine | Radium-228 |
| Cyclophosphamide | Radon-222 |
| Diethylstilbestrol | Silica |
| Epstein-Barr virus | Solar radiation |
| Erionite | Talc-containing asbestiform fibers |
| Estrogen-progestogen menopausal therapy | Tamoxifen |
| Estrogen-progestogen oral contraceptives | 2,3,7,8-Tetrachlorodibenzo-para-dioxin |
| Estrogen therapy | Thiotepa |
| Ethylene oxide | Treosulfan |
| Etoposide | Vinyl chloride |
| Formaldehyde | X- and gamma (γ)-radiation |
| Gallium arsenide | Aflatoxins |
| Helicobacter pylori | Soots
Tobacco Wood dust |
IARC, International Agency for Research on Cancer.
The incidence of skin cancers such as melanoma, basal-cell carcinoma, and squamous-cell carcinoma has risen dramatically in recent years. 22 The risk of developing skin cancer is highest in equatorial regions and correlates with the number of blistering sunburns encountered during childhood. Correlative studies such as these, in addition to mechanistic studies at the cellular and organismal levels, indicate that most skin cancers arise because of exposure to solar radiation. In particular, UV radiation in the 100- to 400-nm range appears to be causative. The health effects of UV radiation vary according to wavelength. Consequently, UV radiation is examined in three regions of wavelength: UVA, 315 to 400 nm; UVB, 280 to 315 nm; UVC, 100 to 280 nm. In contrast to UVC radiation, UVB and UVA can bypass the earth’s atmosphere, including stratospheric ozone; therefore, UVA and UVB are believed to contribute to a much higher attributable risk of cutaneous carcinogenesis than UVC. Moderate UVB exposure results in an erythema response, and UVB is well absorbed by cellular molecules such as DNA, melanin, amino acids, carotene, and urocanic acids. 23,24 UVB is more potent in inducing skin tumors in hairless mice than UVA. However, exposure to UV light of any wavelength results in DNA damage and mutation in in vitro models, and UVA also induces tumors in hairless mice. For this reason, excess exposure to any wavelength of UV light is considered unsafe, and tanning beds have been placed on the IARC’s list of human carcinogens.
For UV radiation to produce an adverse reaction in skin, photon energy must be absorbed by the target biomolecules such as DNA. Although melanin produced by resident melanocytes is a critical UV radiation absorption filter, unfiltered photons may generate oxidative stress and/or damage DNA. UV irradiation of DNA results in the formation of pyrimidine dimers and other photodamage such as DNA strand breaks and pyrimidine-pyrimidone photoproducts. 25 When these lesions are not repaired, DNA mutations can result. The hallmark UVB radiation-induced mutations are C→T or CC→TT transitions. Target genes for solar radiation–induced mutations include but are not limited to TP53 (squamous-cell carcinomas [SCCs], basal-cell carcinomas [BCCs], melanoma), CDKN2A (melanoma), BRAF (melanoma), NEDD9 (melanoma), and PTCH (BCCs, possibly SCCs). UV irradiation of skin keratinocytes also alters numerous cell signaling pathways such as growth arrest and DNA damage-response (i.e., p53, GADD45, mismatch repair genes), apoptotic (i.e., bcl-2, fas), and mitogenic (i.e., ras, ERK) signaling pathways. 26
In addition to solar radiation, ionizing radiation in the form of x-rays, nuclear fallout, and therapeutic irradiation as well as energy deposition from radon gas also contribute to the incidence of human cancers. Epidemiologic studies of radiation workers and atom bomb survivors of Hiroshima and Nagasaki as well as the use of animal models have led to the characterization of ionizing radiation as a “universal carcinogen.” 27 Ionizing radiation can induce tumors in most tissues and in most species examined because of its unique ability to penetrate tissues and induce DNA damage via energy deposition. 28
Radon-222 is a radioactive gas that is produced by radioactive decay of uranium-238, which is found ubiquitously in soil, rock, and groundwater. Concern over accumulation of radon in indoor air, especially in underground spaces, has led to study of the health effects of inhaled radon. Radon decay results in the release of alpha particles (two protons and two neutrons), which do not deeply penetrate tissues but possess the capacity to damage DNA in areas of contact. Inhalation of radon has been associated with lung cancer incidence due to exposure of the bronchial epithelium to decay products. 29 Uranium miners have been shown to succumb to lung cancer at a much higher rate than the general population because of their exposure to radon in underground air supplies. At the reduced exposure level detected in homes, radon carcinogenic potential is low, although not insignificant. WHO officials consider radon to be “the second most important cause of lung cancer second to tobacco in many countries” (http://www.who.int/phe/radiation/backgrounder_radon/en/index.html).
Biologic Carcinogens
Biologic carcinogens also play an important role in human carcinogenesis. Approximately 20% of human cancers are associated with infectious agents including bacteria, parasites, and viruses. These are discussed in more detail in Chapter 6 and are not discussed further in this chapter.
Chemical Carcinogens
Chemical carcinogens can be classified into one of four groups according to their chemical nature: organic carcinogens, inorganic carcinogens, fibers, and hormones. The first experimental confirmation of the existence of organic chemical carcinogens came in 1915, when Yamagiwa and Ichikawa demonstrated that multiple applications of coal tar could induce skin tumors on the ears of rabbits. 30 It was later shown that the active carcinogenic agent was composed entirely of carbon and hydrogen. Since that time, numerous carbon-based carcinogens have been identified in studies using experimental animals and in epidemiologic studies of human populations. These organic compounds range from industrially produced and utilized solvents, to naturally occurring but chemically complex combustion products and mycotoxins, to simple alkyl halides such as vinyl chloride (see Figure 7-1).
Organic Carcinogens
Benzene
Benzene is a widely used solvent and is present in gasoline, automobile emissions, and cigarette smoke. Historically, high-level exposure to benzene was commonplace, and, in general, benzene exposure has been the cause of great concern due to its carcinogenic properties. Exposure to benzene occurs in industrial settings such as in rubber production, chemical plants, oil refineries, and shoe manufacturing. Because benzene is a volatile aromatic solvent, inhalation exposures predominate. 31
The carcinogenic properties of benzene have long been recognized; an increased risk of leukemia has been shown in workers exposed to high levels of benzene. Benzene exposure is associated with myelodysplastic syndromes. In addition, the strongest associations of benzene and cancer risk are found with risk of acute myeloid leukemia and non-Hodgkin’s lymphoma. Benzene is a recognized clastogen and induces oxidative stress upon metabolic activation. Along with mutagenic effects, benzene is believed to alter cell-signaling pathways that control hematopoiesis in hematopoietic stem cells. 32 Workplace exposure restrictions have reduced human exposure to high levels of benzene. Current research is aimed at assessing risk associated with chronic low-level exposure scenarios.
Polycyclic Aromatic Hydrocarbons
Polycyclic aromatic hydrocarbons (PAHs) are a diverse group of intensively studied organic compounds including benzo[a]pyrene. Many PAHs can be metabolically activated to become highly reactive, electrophilic mutagens. PAHs are converted to “bay region” diol epoxides as depicted in Figure 7-2 . These diol epoxides covalently bind to DNA, forming a DNA adduct, and their overall reactivity is predictive of their carcinogenic potency. 2,33 For example, benzo[a]pyrene diol epoxide reacts extensively with the exocyclic amino group of guanine to produce mutagenic DNA adducts (Figure 7-3 , and see section entitled Initiation and Mutational Theory of Carcinogenesis). In addition, certain PAH metabolites may act synergistically with bay region diol epoxide metabolites to promote tumor formation in a manner unrelated to DNA adduct formation. 34 PAHs are formed during combustion of organic matter such as coal, mineral oil, and oil shale. Therefore, PAH exposure occurs in the form of automobile exhaust, soot, coal tar, cigarette smoke, and charred food products. Many PAHs have been found to be carcinogenic in animal studies, and PAH exposure is associated in humans with lung, skin, and urinary cancers, among others. The carcinogenic potential of PAHs is highly variable. Examples of potent to moderately carcinogenic PAHs include 3-methylcholanthrene, B[a]P, dibenzo[a,h]anthracene, 5-methylchrysene, and dibenz[a,j]anthracene, whereas benzo[e]pyrene, dibenz[a,c]anthracene, chrysene, benzo[c]phenanthrene and fluoranthene are relatively weak or inactive carcinogens. Because humans are exposed to mixtures of PAH that are produced during combustion, estimates of carcinogenic potential associated with diverse exposure scenarios are highly variable.
Figure 7-2 Selected polycyclic aromatic hydrocarbon (PAH) bay region dihydrodiol epoxides.
Aflatoxin B1
One of the most potent liver carcinogens is the fungal metabolite aflatoxin B1 (AFB1). AFB1 and other aflatoxins are produced by Aspergillus mold species, such as Aspergillus flavus and Aspergillus parasiticus. Exposure to aflatoxins occurs via consumption of contaminated nuts and grain, such as peanuts and corn, on which Aspergillus species grow. Humid conditions and poor storage contribute to the growth of these molds. In numerous epidemiologic studies, the incidence of hepatocellular carcinoma (HCC) has been correlated with aflatoxin intake. AFB1 is highly mutagenic in in vitro assays. AFB1 is converted to an epoxide metabolite responsible for its mutagenic and carcinogenic action. The DNA base targeted by activated AFB1-epoxide is G (N7 position; see Figure 7-3), and the mutations induced are predominantly GC→TA transversions. Significantly, the TP53 gene is mutated (GC→TA point mutation in codon 249) in a high proportion of human HCCs that arise in areas where aflatoxin exposure is high. 35,36 Evidence suggests that TP53 mutation at codon 249 may occur as a result of combined exposure to HBV and AFB1, and studies have shown elevated risk of HCC in individuals exposed to both HBV and aflatoxin over individuals exposed to either agent alone.
Figure 7-3 Sites of adduct formation associated with carcinogenicity of selected agents.
Benzidine
Benzidine is a member of a large class of carcinogens referred to as aromatic amines. The carcinogenic nature of benzidine was discovered in the context of bladder cancer induction in workers in the dye industry. 37 In the past, benzidine-based azo dyes were synthesized in vast quantities in the United States and abroad. In the 1970s, their use was significantly curtailed because of health concerns. However, numerous workers were exposed to these carcinogens before regulation. On activation, benzidine and certain benzidine-based dyes can covalently react with DNA, and benzidine has been shown to induce chromosomal damage in vivo. 38 Benzidine is a bladder carcinogen in multiple species, including humans, dogs, mice, rats, and hamsters, although species differences in activation of the parent compound have made the study of benzidine-induced bladder cancer challenging. 39
Nitrosamines and Heterocyclic Amines
Shortly after the identification of benzo[a]pyrene, N-nitrosodimethylamine was shown to induce liver tumors in rats. These results were provocative at the time because of the stark differences in physical properties between the PAHs and the water-soluble N-nitroso compounds. Since the initial discovery of N-nitrosodimethylamine, a wide variety of N-nitroso compounds have been shown to be powerful carcinogens in multiple experimental models and suspected carcinogens in lung and gastrointestinal cancers in humans. 40 Following metabolic activation, N-nitrosamines can react with DNA to initiate carcinogenesis. Exogenous and endogenous sources of N-nitroso compounds have been described. N-nitrosamines are present in smoked meats and in meats containing the antimicrobial and color-enhancing agent nitrite. In both cases, nitrogen oxides are formed, which react with the amines present in meat. Alternatively, the formation of N-nitroso compounds can occur endogenously because of low pH conditions in the gastric system or as result of the presence of intestinal bacteria that catalyze N-nitroso compound formation.
Heterocyclic amines are also formed in muscle meats on high-temperature processing. Most heterocyclic amines tested are mutagenic in in vitro assays, and several induce gastrointestinal tumors in rodents. 41,42 The two heterocyclic amines found most abundantly in cooked meat and best absorbed into the circulation are 2-amino-1-methyl-6-phenylimidazo-(4,5-b)-pyridine (PhIP) and 2-amino-3,8-dimethylimidazo-(4,5-f)-quinoxaline (MeIQx). At high temperatures, these heterocyclic amines are formed via reactions among creatinine, creatine, sugars, and amino acids.
N-Nitrosamine exposure is also associated with tobacco use 43 : 4-(Methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK), 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL), and N-nitrosonornicotine (NNN) are carcinogenic tobacco-alkaloid–derived N-nitrosamines present in unburned and burned tobacco products. PAHs and NNK are the most abundant pulmonary carcinogens in tobacco smoke. In contrast to PAHs, which induce SCCs, NNK induces adenocarcinoma of the lung in animal models. Furthermore, adenocarcinoma of the lung has become the most common lung cancer type in the United States. This fact may reflect changes in cigarette manufacturing in the past 30 to 40 years that have resulted in rising levels of NNK and falling levels of B[a]P. In addition, in smokeless tobacco products such as snuff, N-nitrosamines are prominent agents involved in the induction of oral cancer. These N-nitrosamines require metabolic activation for carcinogenic activity and form DNA adducts similar to other organic carcinogens discussed earlier.
Inorganic Carcinogens
Beryllium
In 1946, Hardy and Tabershaw reported “delayed chemical pneumonias” in workers exposed to beryllium. In that same year, Gardner and Heslington reported experimentally induced osteosarcomas in beryllium-injected rabbits. Subsequent studies in the 1950s demonstrated that inhalation exposure of rodents resulted in induction of lung tumors. Since that time, beryllium has been recognized as a human carcinogen capable of inducing lung cancer in exposed workers. Occupational exposures to beryllium include inhalation of beryllium-containing dusts during processing of ores, machining of beryllium metal and alloys, and manufacturing of aerospace materials, ceramics, sports equipment, and electronics. Beryllium is weakly mutagenic in bacterial and mammalian mutagenesis test systems; however, it shows strong transformation capacity in Balb/3T3 and Syrian hamster secondary embryo cells. 44 In addition to genotoxic effects, beryllium has been shown to alter the expression of numerous cancer-related genes (i.e., c-fos, c-jun, c-ras, MAP kinases, DNA repair genes,). Despite these reports, the carcinogenicity of beryllium has recently been called into question. 45,46 Current occupational exposure levels are much lower than historical values, and previous reports of beryllium’s effects may have been confounded by inadequate smoking history information.
Cadmium
Cadmium is a heavy metal present in soil, air, and water and is listed as a priority pollutant by the U.S. EPA. Occupational exposures to cadmium occur during the manufacture of nickel-cadmium batteries, pigments, and plastic stabilizers as well as electroplating processes, metal smelting, and electronic waste recycling. 47 In addition, cigarette smoke contains cadmium. Release of industrial cadmium waste into the environment is of particular concern because of its long biologic half-life. On absorption, cadmium can accumulate in the body because it is poorly excreted and effectively stored in liver and kidney as a result of binding to metallothionein. Furthermore, once absorbed, no effective detoxification pathways for cadmium exist. The half-life of cadmium in humans is estimated at 15 to 20 years.
Cadmium exposure has been linked to human lung cancer and may affect the risk of prostate and kidney cancers. Although the carcinogenicity of cadmium has been confirmed in rodent models, the precise mechanism is unknown. 47 Cadmium binds only weakly to DNA and is only weakly mutagenic in bacterial and mammalian assays, and high concentrations are required to induce oxidative stress. Cadmium may act via non-genotoxic mechanisms to activate proto-oncogenes and disrupt normal cellular processes. For example, cadmium has been shown to alter E-cadherin–mediated cell adhesion, inhibit DNA repair, and alter expression of numerous genes in vitro including c-fos, c-myc, metallothionein, and genes encoding heat shock proteins.
Arsenic
Arsenic is widely distributed in the environment, being found in the earth’s crust in both inorganic [arsenite-As(III) and arsenate-As(V)] and methylated forms [monomethylated arsenic (MMA) and dimethylated arsenic (DMA)]. As(III), as well as MMA(III) and DMA(III), have been associated with skin, lung, urinary bladder, kidney, and liver cancers. 48 Human exposure to arsenic occurs via contaminated drinking water, diet, or contact with wood preserved with arsenicals; during mining of tin, gold, and uranium; and during application of arsenical pesticides. Signs of chronic exposure to arsenic in drinking water include altered skin pigmentation and hyperkeratosis of the palms of the hand and soles of the feet, which may ultimately lead to skin lesions and skin cancer.
Much attention has been given to assessing the health impact of arsenic contamination in drinking water sources. The current WHO guidelines for arsenic exposure recommend no more than 10 μg/L arsenic in water intended for human consumption. Since the 1980s, millions of people in China, India, Bangladesh, the United States, Chile, and Argentina have been exposed to arsenic in the drinking water far in excess of this limit. Already, numerous epidemiologic studies in Taiwan, the United States, Chile, and Argentina have demonstrated excess cancer risk in areas with known high exposure to arsenic in drinking water. 49 Unfortunately, identifying a safe level of arsenic in drinking water has been difficult because most epidemiologic studies show adverse effects at high doses; data concerning health risk at low exposures are unavailable. After intense debate, the limit in the United States was lowered to 10 μg/L in 2006.
For years, the study of arsenic and cancer was hindered by a lack of experimental evidence of carcinogenicity in animals. Only recently have studies of methylated arsenic and early life exposures provided adequate validation in vivo. As(III) and As(V) are transported into cells, As(III) more readily than As(V). On absorption, As(V) is reduced to As(III); As(III) can then be methylated. Historically, methylation of As(III) was considered to be a detoxification reaction, but recent evidence contradicts this dogma. 50 MMA(III) and DMA(III) are at least as cytotoxic, mutagenic, and clastogenic as As(III). Nonetheless, when methylated, arsenic is readily excreted in urine. DMA can be detected in urine shortly after exposure; also, because of the wide distribution of arsenic, exposure can be assessed via hair and fingernail deposits months or years after exposure. Therefore, methylated arsenic in hair and fingernail samples serves as a useful biomarker of exposure.
Numerous mechanisms of action have been proposed for arsenic carcinogenicity. 48 Arsenic exposure is known to generate reactive oxygen species. Like many transition metals, arsenic can participate in Fenton reactions that produce oxidative stress. Furthermore, arsenic may activate superoxide-generating NAD(P)H oxidase. In this way, arsenic is thought to induce DNA and protein damage that may initiate carcinogenesis. Arsenic has also been shown to elevate the total level of tyrosine phosphorylation in cells. Specifically, arsenic may alter phosphorylation-dependent epidermal growth factor receptor (EGFR) and mitogen-activated protein kinase (MAPK) signaling. In addition, arsenic has been shown to alter NFκ signaling, apoptosis rates, cell cycle regulation, DNA repair, and genome stability. More recent evidence suggests that epigenetic dysregulation may underlie these adverse effects of arsenic exposure. 51
Chromium
Chromium in the hexavalent state [Cr(VI)] is a human carcinogen. The carcinogenic properties of chromium have been identified via epidemiologic studies of exposed workers in industries such as chrome plating, welding, leather tanning, and stainless steel production. Exposure to chromium generally occurs via inhalation and primarily affects risk of lung cancer. Because of environmental contamination, consumption of chromium in drinking water is also possible; however, the health consequences of the low-level exposure are unclear. 52
The oxidation state of chromium determines not only its bioavailability but also its cellular reactivity. 53 Cr(VI) readily enters cells via anion channels, whereas Cr(III) only slowly crosses the cell membrane. On entry to the cell, Cr(VI) is likely reduced, as Cr(VI) does not readily react with DNA in in vitro analyses. Chromium in lower oxidation states [Cr(III), Cr(IV), and Cr(V)] is more reactive; Cr(III) is believed to be the ultimate DNA reactive form. The reduced forms of chromium can also induce oxidative stress. In addition to or as a result of oxidative stress, chromium alters cell signaling pathways. Signaling molecules affected include NFκB, AP-1, p53, and HIF-1.
Fibers
Asbestos
The term asbestos refers to a group of naturally occurring silicate mineral fibers. There are numerous types of asbestos fibers that are classified according to their morphologic characteristics, including whether the fibers are curly (serpentine) or straight (amphibole). The shape and length-to-width ratio are important determinants of whether a particular asbestos fiber type will be carcinogenic. 54 This is likely because the size of the fiber determines the ability of the fiber to reach the deep lung tissues and penetrate the lung. Long (>4 μm) and thin (<0.5 μm diameter) fibers are the most carcinogenic. Extensive exposure to asbestos has occurred because the flame-resistant and durable characteristics of asbestos have led to its use as an insulating agent in schools, factories, homes, and ships, as construction material, and as a raw material for automobile brake and clutch parts. A large cohort of workers was exposed to high levels of asbestos when ship building peaked during World War II.
The toxic effects of asbestos exposure have been known for many years. 54 For example, more than 40 years ago, crocidolite asbestos exposure of South African miners was linked to mesothelioma incidence. Mesothelioma is a rare cancer of the membranous lining of the abdomen and chest. Numerous animal studies and in vitro experiments support the conclusion that asbestos can induce tumors. In fact, few cancer cause-effect relationships are as striking as asbestos and mesothelioma; most cases of mesothelioma can be related to asbestos exposure. In addition to mesothelioma, asbestos exposure has been associated with lung and laryngeal cancers. The initial accumulation of evidence of asbestos carcinogenicity was obscured by the differences in carcinogenicity of asbestos fibers of varying shape and by the long latency for the development of tumors following exposure. Since identification of asbestos as a cancer-causing agent, asbestos usage in the United States has greatly declined because of the introduction of a replacement material (fiberglass) and OSHA regulation of asbestos exposure. Nonetheless, mesothelioma rates have not declined in the United States in the past 15 years.
Numerous biologic hypotheses concerning the mechanism(s) by which asbestos induces tumors have been proposed. 55 Because long, thin fibers are the most carcinogenic, asbestos fibers may penetrate the lung and irritate the lining of the chest wall. The chronic inflammation and scarring would then contribute to tumor formation. Alternatively, the fibers may pierce spindle fibers during mitosis and thereby induce chromosome damage. Finally, asbestos fibers may induce oxidative stress and/or alter EGFR and MAPK cell signaling. Significantly, epidemiologic studies show that cigarette smoking acts synergistically with asbestos exposure to induce lung tumors. In addition to asbestos, exposure to other fibers such as plant-derived silica fibers (biogenic silica) has been shown to be carcinogenic. 56
Hormones
The etiology of numerous cancers is believed to be influenced by hormonal or dietary factors, and hormones under certain conditions are considered to be known human carcinogens (see Table 7-5). As previously mentioned, overweight and obesity are associated with elevated cancer risk. This effect may be mediated by endocrine dysregulation such as altered adiponectin, leptin, insulin, and IGF-1 levels. In addition, prostate, ovarian, breast, testicular, and endometrial cancers are hormonally driven. 57,58 A role for hormones in cancer etiology was established when castration and ovariectomy studies revealed that hormone-dependent cancers could be prevented by removing the primary hormone-synthesis organs. As an example of the action of hormones in cancer formation, estrogen activates hormone-responsive receptors. Stimulation of these receptors, such as the estrogen receptors, can increase the cellular proliferation rate to promote tumorigenesis. Endogenously synthesized hormones and administered hormones have been shown to influence cancer formation. Hormone replacement therapy and estrogen-only birth control therapy have been associated with increased risk of hormone-dependent cancers. An even more dramatic example of synthetic hormone-induced cancer is that of women who were exposed to estrogenic diethylstilbestrol (DES) in utero. DES was taken by pregnant women to prevent abortion; however, a large percentage of their female offspring developed clear-cell carcinomas of the vagina and cervix after the onset of puberty.
Mechanisms of Chemical Carcinogenesis
Multistage Nature of Carcinogenesis and the Multistage Model of Mouse Skin Carcinogenesis
In the early days of carcinogenesis research, it was noted that wounding of the skin of mice previously treated with mutagenic coal tar led to skin tumor formation. To explain these findings, a multistage model of carcinogenesis was proposed. 59 The model holds that tumors arise in cells that have first undergone a mutating event initiated by an electrophilic metabolite such as that formed from benzo[a]pyrene found in coal tar. Subsequently, cell proliferative stimuli promote the initiated cell population to expand, resulting in premalignant clonal outgrowths. Finally, additional genetic alterations accumulate in these lesions, leading to the development of a neoplasm that becomes invasive and ultimately metastatic. Over the years, this model has been refined to encompass the fundamental role of stem cells as the targets of initiation, the importance of stemlike characteristics in the plasticity of developing tumors, and the critical role of tumor microenvironment in carcinogenesis. 60 In addition, the mechanistic importance of DNA methylation changes and histone modifications in the initiation, promotion, and progression phases has been acknowledged.
Numerous animal models have been developed to study the multistep manner in which various epithelial and other tumors develop and progress. In one of the best-characterized models, the mouse two-stage skin carcinogenesis model, a subcarcinogenic dose of a mutating agent is delivered. 61 This is followed by multiple exposures to growth-promoting stimuli and the appearance of tumors on the dorsal skin (Figure 7-4 ). This model has provided an excellent paradigm in which to examine the carcinogenic potential of environmental agents and has been used to reveal the mechanistic bases of multistage carcinogenesis by environmental agents.
Initiation and Mutational Theory of Carcinogenesis
During the first stage of multistage carcinogenesis, DNA mutations result as a consequence of electrophilic carcinogen exposure, oxidative damage to DNA, DNA strand breaks, or other DNA insults. Mutations are believed to occur in multipotent stem cells and are inherited by daughter cells. Theodor Boveri first proposed the concept that cancer arises as a result of damage to genetic material at the turn of the 20th century. In the 1950s and 1960s, James and Elizabeth Miller, after observing that a wide variety of structurally diverse chemicals could induce cancer in animal models, suggested that chemical carcinogens required metabolic activation to electrophilic intermediates. These electrophilic intermediates could then covalently bind to proteins, RNA and DNA. The term electrophile theory of chemical carcinogenesis was coined to describe their concept. The Millers’ work was supported by data reported in 1964 by Brooks and Lawly, which demonstrated that the degree of covalent binding of carcinogenic PAHs to DNA correlated with carcinogenic potential. 62 Subsequently, many chemical carcinogens have been shown to bind and alter DNA integrity, thereby inducing mutations. Carcinogens that alter DNA to induce cancer in this manner are referred to as genotoxic carcinogens. The majority of carcinogens identified to date are genotoxic carcinogens.
Figure 7-4 Multistage model of mouse skin carcinogenesis. BaP, Benzo[a]pyrene; DMBA, 7,12-dimethylbenz[a]anthracene; LOH, loss of heterozygosity; MNNG, N-methyl-N-nitro-N-nitrosoguanidine; TPA, 12-O-tetradecanoyl phorbol-13-acetate.
DNA mutations occurring in proto-oncogenes and tumor suppressor genes are particularly critical to the initiation of carcinogenesis. These normal cellular genes are targeted during carcinogenesis and play critical roles in tumor formation. Proto-oncogene mutations are dominant, in that activation of a single copy of a proto-oncogene to an oncogene may be significant for carcinogenesis. Proto-oncogenes are discussed in greater detail in Chapter 2; however, a list of proto-oncogenes and the cancers with which they are associated is provided in Table 7-6 . In contrast to proto-oncogenes, the normal cellular function of tumor suppressor genes is to negatively regulate cell growth. According to Knudson’s two-hit theory, tumor suppressor genes require that both copies of the gene be lost or inactivated because tumor suppressor mutations are recessive in nature. For instance, inheritance of one mutated copy of p53 is not significant until the second copy is lost (“second hit”), resulting in loss of heterozygosity (LOH). Examples of tumor suppressor genes and associated cancers are also provided in Table 7-6.
In the multistage mouse skin carcinogenesis model, initiation occurs via application of a genotoxic carcinogen (e.g., N-methyl-N-nitro-N-nitrosoguanidine [MNNG], 7,12-dimethylbenz[a]anthracene [DMBA], or B[a]P; see Figure 7-4). A subcarcinogenic dose of the initiating agent is applied to the shaven dorsal skin of the mouse. The critical mutations for tumor development are believed to occur in epidermal multipotent stem cells, a major population of which resides in the bulge region of the hair follicle. Although no phenotypically aberrant cells are apparent in the “initiated” skin, small populations of epidermal cells can be identified as early as 1 week after treatment with DMBA that contain signature mutations. The initiation stage is irreversible and cumulative. That is, the dose required for initiation can be divided and applied in portions over time or applied in a single dose with essentially the same result. In addition, commencement of the promotion phase can be delayed, because the DNA mutations induced by the initiating agent are permanent.
Table 7-6
Selected Proto-oncogenes and Tumor Suppressor Genes and Some Associated Cancers
Adapted from Perkins AS, Stern DF: Molecular biology of cancer. Oncogenes. In: DeVita VT, Hellman S, Rosenbert SA, eds. Cancer: Principles & Practice of Oncology. 5th ed. Philadelphia, Pa: Lippincott-Raven Publishers; 1997.
In the mouse skin model, the most frequently mutated proto-oncogene following initiation with PAH is Hras1 (reviewed in Ref. 61). Mutations at G38 of codon 13, in the case of B[a]P, and at A182 of codon 61, in the case of DMBA, lead to constitutive activation of the gene product of Hras1. These mutations can be detected in the resulting tumors, reflecting the clonal origin of the papillomas. In addition, the specificity of these mutations is directly related to the major sites of DNA adduct formation arising from the carcinogenic diol-epoxides of these two PAH carcinogens. In a variety of rodent models of multistage cancer (rat azoxymethane-induced colonic lesions, mouse diethylnitrosamine-induced liver foci, mouse urethane-induced lung adenomas), mutations in ras oncogenes frequently occur, highlighting the importance of ras, and oncogenes in general, in the development of cancer. Activating mutations in Hras1 are believed to confer, at least in part, resistance to terminal differentiation of keratinocytes during tumor promoter treatment, thus conferring a selective growth advantage to these cells. Findings in animal models have established the irreversible and cumulative nature of tumor initiation and underscore the specificity of critical DNA mutations in proto-oncogenes or tumor suppressors induced by genotoxic carcinogens.
During carcinogenesis, numerous types of DNA lesions occur following exposure to carcinogenic agents. For example, in the case of electrophilic carcinogen attack, specific points within the DNA nucleotides are targeted for adduction (see Figure 7-3). As noted previously, B[a]P targets primarily the N 2 exocyclic amino group of guanine, whereas other PAHs may target adenine in addition to guanine (e.g., DMBA). As shown in Figure 7-3, alkylating agents target numerous sites within DNA bases. However, certain sites (e.g., 06 methylguanine and 04 methylthymine for methylating agents) may be the most important for carcinogenesis by this class of carcinogen. During replication, mispairing due to DNA adducts or other DNA lesions may become fixed, and the ultimate effect depends on the location of the mutation. Mutations affect coding sequences, intronic signals, untranslated regions, or promoter elements; consequently, protein function or expression levels may be altered. Following DNA double-strand breaks, incorrect rejoining of DNA has been shown to cause rearrangement of DNA coding and promoter regions. In addition to these qualitative changes, quantitative changes in gene copy number (gene amplification or gene deletion) may also affect key cancer-associated genes.
Promotion
During the tumor promotion phase of carcinogenesis, growth stimuli and other factors promote clonal expansion of initiated cells. This stage is characterized by altered gene expression and proliferation of initiated cells, some of which maintain stemlike characteristics. Most tumor promoters are thought to exert their effects through cellular receptors or cell growth, differentiation, and/or apoptotic signaling pathways. Inflammatory mediators or other stromal factors may mediate these effects by providing a permissive environment for tumor growth. Promoting agents do not directly affect DNA but act primarily via non-genotoxic, reversible mechanisms.
Growth Factor Receptor Signaling Pathway Engagement
In the mouse two-stage skin carcinogenesis model, the promotion stage is elicited by multiple applications of promoting agents delivered over the course of weeks or months (see Figure 7-4). In this model, promotion must occur following initiation with a mutating agent. As opposed to initiation, the promotion stage is initially reversible, does not elicit DNA mutation, is prolonged in nature, and appears to be nonadditive. Typical skin tumor-promoting agents include the phorbol ester 12-O-tetradecanoyl phorbol-13-acetate (TPA); the phosphatase inhibitor okadaic acid; the organic peroxide benzoyl peroxide; and the anthrone derivative chrysarobin (see Figure 7-5 for the diversity of structures of chemical tumor promoters). In addition, UV light, repeated abrasion, full-thickness wounding, and certain silica fibers possess the ability to function as tumor promoters. The endpoint of promotion in the mouse two-stage skin carcinogenesis model is the development of premalignant, clonal outgrowths referred to as squamous papillomas. These lesions (hyperplastic epidermis folded over a core of stroma) are still well differentiated and do not possess the ability to invade surrounding tissue. Once the cells of the papilloma acquire additional mutations that allow autonomous growth, the promotion stage is no longer reversible.
Figure 7-5 Chemical structures of selected tumor promoters.
Promoting agents act primarily by altering gene expression. 63 The initial interaction of promoting agent with the cell depends on the nature of the promoter. In the mouse skin model, the receptor for TPA (the most frequently used promoting agent) has been identified as protein kinase C (PKC). Stimulation of PKC results in a cascade of events that allow for expansion of the initiated cell population. PKC-mediated signaling events include induction of ornithine decarboxylase (ODC) activity, activation of the MAPK pathway, and upregulation of ligands for the EGFR. EGFR activation leads to further activation of multiple signaling pathways involved in proliferation and survival. For instance, Akt and Stat3 signaling are believed to affect cell-cycle parameters via altered cyclin Dl expression. 64,65 Although the initial mechanism for other skin tumor promoters (e.g., okadaic acid, benzoyl peroxide, chrysarobin) is different from that of TPA, all tumors promoter ultimately elicit key biologic and molecular changes. These changes include induction of ODC, induction of growth factors and cytokines, production of eicosanoids, and increased DNA synthesis. Growth factors and cytokines known to be altered by tumor-promoting stimuli include TGF-α, TGF-β, IL-1, IL-6, and TNF-α, among others.
During tumor promotion, immune inflammatory cells play diverse and complex roles. Increasingly, cytokines are acknowledged as active players in the growth-promoting environment of developing neoplasia. These cells, along with fibroblasts and other cells present in the stroma, release signaling molecules such as EGF, VEGF, and FGF2 to facilitate proliferation and contribute to proinvasive capacity by affecting expression of matrix metalloproteinases and other proteases. In contrast, other immune surveillance cells can play tumor-antagonizing roles.
Examples of Tumor Promoters
Tumor-promoting agents have been identified for a number of rodent tissues other than mouse skin, indicating the generality of this phenomenon to other organs and species (reviewed in Ref. 66). Some examples of tumor promoters that act on organs other than skin include 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) (liver), butylated hydroxyanisole (BHA) (lung), sodium saccharin (urinary bladder), and bile acids (colon) (see Figure 7-5 again for chemical structures). Certain components of tobacco smoke can also act as tumor promoters.
Progression
The final stage of carcinogenesis is referred to as progression. The original tumor mass increases in size, additional mutations accumulate, and invasion and metastasis occur. Genome stability is compromised, and a mutator phenotype may develop to permit further accumulation of mutations. Crucial to progression of solid tumors is the ability of cancer cells to invade the surrounding stroma, enter the bloodstream, and extravasate to colonize distal sites. For cells to break away from the primary tumor, downregulation of cell adhesion, often by repression of E-cadherin expression, must occur. Subsequently, the cells must acquire mobility and ability to invade the surrounding stroma and blood vessel basement membranes. Invasion requires the action of degradation enzymes such as the matrix metalloproteinases (MMPs). These changes are collectively referred to as the epithelial-to-mesenchymal transition (EMT). Once established, the metastatic colony must develop adequate nutrition and oxygen supply via stimulation of angiogenesis.
The process whereby normal cells accrue an increasingly aberrant and invasive phenotype as a consequence of selection has been likened to evolutionary theory at the population level. 67 However, in addition to an accumulation of genetic mutations, tumor promotion and tumor progression can also involve heritable but non-genotoxic lesions. These changes, referred to as epigenetic reprogramming, involve altered DNA methylation patterns and changes to the histone code. In the case of DNA methylation, the promoter regions of tumor suppressor genes may become hypermethylated, thereby silencing gene expression. Alternatively, the expression of proto-oncogenes may be upregulated via hypomethylation. Recent proposals suggest that environmental agents elicit epigenetic changes and alter the risk of tumor promotion and/or progression. 68,69 Interestingly, epigenetic reprogramming may precede the neoplastic process in some cases, and modifications to the histone code may affect DNA damage response or DNA repair during initiation, promotion, or progression. 69
During progression in the two-stage mouse skin carcinogenesis model, premalignant papillomas convert to SCCs (see Figure 7-4). 61 This conversion process is accompanied by additional genetic alterations, including the development of aneuploidy. Characteristic changes in gene expression such as elevation in gamma-glutamyltranspeptidase, α6β4 integrin, and keratin-13 expression as well as loss of E-cadherin expression are also common. The histologic appearance of SCCs can be distinguished from papillomas by downward growth and loss of ordered differentiation of epidermal keratinocytes.
Multistage Environmental Carcinogenesis in Humans
The applicability of multistage carcinogenesis concepts to human cancer is supported by a number of observations. First, human environmental carcinogen exposure outside of occupational settings usually occurs in low doses repeatedly delivered over the course of months or years. Each individual dose alone is likely insufficient to produce cancer. In addition, it is unlikely that a single dose of a sole agent is the cause of most human cancers. Second, there is considerable evidence from human epidemiologic and experimental animal studies that certain human carcinogens such as tobacco smoke and UV light exhibit a strong tumor-promoting activity. Furthermore, many components of the human diet appear to influence cancer in humans through a tumor-promotion type of effect. Finally, histochemical and molecular examination of tumors at various stages indicates that human cancers develop via multiple steps. It has been postulated that human cancers require as many as four to six sequential genetic events for their development. 70
Numerous human cancers, particularly those of epithelial origin, appear to develop in a multistage progression. For instance, regions of dysplasia and carcinoma in situ appear to precede invasive carcinoma when melanoma, head and neck squamous-cell carcinoma, and cervical cancer lesions are examined. Supporting the multistage nature of cancer development, genetic alterations have been shown to accumulate during tumorigenesis in these lesions. For example, during colorectal carcinogenesis, mutations in the adenomatous polyposis coli (APC) gene appear to initiate tumorigenesis. 71 A portion of the resulting dysplastic foci further accumulate mutations in the K-ras oncogene and other oncogene and tumor suppressor genes and progress from adenomas to invasive carcinomas. A similar pattern of accumulation of molecular abnormalities has been noted for squamous-cell lung carcinoma. As the severity of the histopathologic appearance of these lesions increases, the frequency of loss of heterozygosity events also increases. 72
Examination of tumor DNA has also validated a role for environmental carcinogens in the etiology of human cancer. When tumor suppressor and oncogene gene sequences are examined, characteristic mutation spectrums can be identified and associated with specific carcinogen exposure. The mutation spectrum of the p53 tumor suppressor gene has been intensively studied. 73 A database of more than 10,000 reports of p53 mutations in human cancers has been collected. Depending on the cancer type, mutations are frequently reported at amino acids 130-142, 151-164, 171-181, 193-200, 213-223, 234-258, and 270-286, which are part of the DNA-binding domain of p53. Sixty-one percent of lung cancer samples have mutations at codon 157 in addition to mutations in codons 248 and 273. A large percentage of these mutations are the result of G→T transversions. In vitro analyses indicate that exposure of normal human bronchiolar epithelial cells to benzo[a]pyrene diol epoxide results in DNA adducts in p53 in the same mutation hot spots as in lung cancer: codons 157, 248, and 273. 74 These results provide strong evidence for a link between chemical carcinogen exposure (B[a]P of cigarette smoke) and human lung cancer. In addition, aflatoxin B1 exposure correlates strongly with liver cancer and p53 mutation at codon 249, whereas sunlight exposure, which is known to induce CC→TT transition mutations, correlates with CC→TT tandem mutations at hot spots for skin cancer in p53.
Endogenous Defense Systems against Chemical Carcinogenesis
Various genetically influenced defense systems determine the ultimate outcome following exposure to an environmental carcinogen. Chemical carcinogens are often metabolized to less toxic derivatives and cleared from the body. Damage to cellular DNA may be repaired, or the damaged cell may be eliminated via apoptosis. Furthermore, the cell possesses a number of endogenous defense mechanisms against carcinogen-induced oxidative stress. Interindividual differences in the efficacy of these defense mechanisms are known to exist and influence host susceptibility to environmentally induced cancer.
Carcinogen Metabolism
Subsequent to absorption through the gastrointestinal tract, xenobiotics travel via the portal vein to the liver where first-pass metabolism occurs. Hepatic tissues are highly concentrated with metabolic enzymes specialized in chemical conversion referred to as biotransformation. In other cases, such as inhalation or dermal exposure, biotransformation enzymes at the site of exposure can begin immediately to convert the parent compound into metabolites. Biotransformation enzymes are theorized to have evolved primarily as natural defenses against environmental chemical exposure.
Phase I and Phase II Biotransformation Reactions
The reactions catalyzed by biotransformation enzymes have been categorized into groups referred to as phase I and phase II reactions because of their often sequential roles in the conversion of xenobiotics. Phase I reactions include oxidation, reduction, and hydrolysis reactions and, generally, expose functional groups that enable phase II biotransformation to proceed. Phase II biotransformation reactions catalyze glucuronidation, sulfation, acetylation, methylation, and glutathione conjugation reactions, among others. 75 Numerous enzymes that catalyze these reactions have been identified and classified according to gene family (Table 7-7 ).
Phase I metabolites, in general, display minimally increased hydrophilicity. In contrast, phase II biotransformation reactions catalyze the addition of cofactor molecules to the parent compound, resulting in a significant increase in hydrophilicity. In certain instances, phase II conjugation reactions may also target the parent compound for export via specialized efflux pumps. Therefore, in general, phase II biotransformation reactions ultimately result in metabolites that are less toxic and more readily excreted. In contrast, phase I biotransformation of carcinogens often results in reactive metabolites capable of covalent modification of cellular macromolecules. It is important to note, however, that these are generalizations. Examples of phase I–mediated detoxification have been noted, and phase II–mediated chemical activation has been documented.
According to the Millers’ electrophilic theory of carcinogenesis, all mutagenic compounds must be inherently chemically reactive or converted via biotransformation to a reactive form. Carcinogens that do not require metabolic activation are referred to as direct carcinogens; indirect carcinogens require metabolic activation. The conversion of the parent compound to a reactive state converts a procarcinogen to an ultimate carcinogen. Ultimate carcinogens, like direct carcinogens, are electrophilic and attack nucleophilic groups in DNA to initiate carcinogenesis, as discussed in the section titled Initiation and Mutational Theory of Carcinogenesis. Although categorizing biotransformation reactions according to the phase I versus phase II nature of the metabolism is useful, the endpoint of carcinogen exposure is often determined by a combination of oxidation, reduction, and conjugation reactions.
PAH Biotransformation
PAHs are widely studied substrates for cytochrome P450 (CYP450)-mediated biotransformation. CYP450s, a class of enzymes present in the endoplasmic reticulum of most cells, have been implicated in numerous carcinogen activation reactions. In humans, the CYP450 family consists of more than 50 genes. which are grouped on the basis of sequence similarity into families (1, 2, 3, . . .), subfamilies (A, B, C, . . .), and individual CYP450s (1, 2, 3, . . .) (e.g., CYP450 1A1, 1A2, 1B1, etc.). 64 CYP450s catalyze oxidation, reduction, oxygenation, dealkylation, desulfuration, dehalogenation, and hydroxylation reactions. CYP450-mediated reactions can detoxify direct carcinogens or activate indirect carcinogens.
Once absorbed, certain PAHs can be biotransformed into electrophilic mutagens via the sequential action of phase I enzymes (Figure 7-6 23). First, PAH double-bond oxidation is catalyzed by CYP450 enzymes (e.g., CYP1A1). For example, in the case of B[a]P, CYP450-mediated oxidation forms the epoxide intermediate, benzo[a]pyrene-(7R,8S)-epoxide. Next, microsomal epoxide hydrolase (mEH) catalyzes hydrolysis of arene oxide to a trans dihydrodiol. Finally, a CYP450-catalyzed oxidation reaction forms the ultimate carcinogen (i.e., benzo[a]pyrene-7,8 diol-9,10-epoxide), a diol-epoxide metabolite. In human lung tissue, both B[a]P epoxidation steps are catalyzed primarily by CYP1A1. Studies using CYP1B1-deficient mice highlight the importance of this p450 enzyme in PAH activation. These mice are resistant to DMBA-induced carcinogenesis, due to a lack of conversion of DMBA from procarcinogen to ultimate carcinogen. 76
PAHs can be detoxified by glutathione S-transferases (GSTs). GST-mediated glutathione conjugation of PAH epoxides can deactivate the ultimate carcinogen, prevent activation to reactive diol epoxides, or accelerate clearance of PAHs following exposure. In this way, GSTs act as an endogenous defense system against PAH-induced carcinogenesis.
Aflatoxin Biotransformation
Metabolism plays a critical role in determining the carcinogenicity of the mycotoxin AFB1. AFB1 must first be activated to the ultimate carcinogen, exo-8,9-AFB1-epoxide (see Figure 7-6). This reaction is predominantly catalyzed by CYP450 3A4 in humans. 35 Alternatively, CYP450s can metabolize AFB1 to inactive products such as AFM1, AFQ1, or AFB1 endo-8,9-epoxide (AFBO).
Glutathione conjugation catalyzed by GSTs plays a critical role in protecting against mutagenic and carcinogenic effects of AFB1 metabolites. Generally, GSTs facilitate xenobiotic clearance by catalyzing glutathione conjugation of a variety of electrophilic substrates. 77 In humans, cytosolic GSTs are categorized according to gene sequence similarity into at least six classes: Alpha (A), Mu (M), Omega (O), Pi (P), Theta (T), and Zeta (Z). Individual GST family members demonstrated unique, though overlapping, substrate specificity.
Subsequent to activation, GST-mediated glutathione (GSH) conjugation can detoxify the AFB1 epoxide, and this reaction is a major factor underlying the substantial species variation in sensitivity to AFB1-induced carcinogenesis. For example, rats are highly sensitive to AFB1-induced hepatocarcinogenesis, whereas mice are comparatively resistant. In line with this observation, mice express mGSTA3-3, which demonstrates high activity toward AFBO, whereas rat GST-mediated deactivation of AFB is significantly less in comparison. Mutational studies of recombinant mGSTA3-3 indicate that the high activity of this protein toward AFBO is due to multiple, critical amino acid residues in the substrate binding site that are not present in homologous rat GSTA3-3. 78
Figure 7-6 Biotransformation either activates or deactivates the ultimate carcinogen (A) Sequential action of cytochrome P450 (CYP450) and microsomal epoxide hydrolase (mEH) activates B[a]P. (B) CYP450 activates while glutathione S-transferase (GST)-mediated conjugation of glutathione (GSH) deactivates aflatoxin B1 (AFB1). (C) Vinyl chloride is activated to its epoxide metabolite by CYP450. (D) The 2-amino-1-methyl-6-phenylimidazo-(4,5-b)-pyridine (PhIP) is first metabolized by CYP450, then activated by N-acetyltransferase. (E) GST mediates activation of ethyldibromide.
Vinyl Chloride Biotransformation
Vinyl chloride is the starting material for the production of polyvinyl chloride, used in the fabrication of products such as PVC pipe. The mutagenicity of this liver carcinogen is dependent on metabolic activation by CYP450, and detoxification is mediated by mEH. 18 As shown in Figure 7-6, vinyl chloride is a relatively simple compound. In the presence of oxygen and NADPH, CYP450 2E1 catalyzes formation of a highly unstable epoxide moiety across the central carbon double bond. This epoxide, chloroethylene oxide, is the ultimate carcinogen capable of covalently binding DNA. Chloroethylene oxide can be detoxified via the action of mEH as noted previously or by GST-mediated glutathione conjugation.
Benzidine Biotransformation
Benzidine, an aromatic amine bladder carcinogen, must also undergo metabolic activation to initiate carcinogenesis. 38 CYP450 enzymes catalyze the activation of benzidine via N-oxidation. Subsequent to N-oxidation, N-acetyltransferase (NAT)-catalyzed O-acetylation forms electrophilic N-acetoxy derivatives capable of attacking DNA. In contrast, N-acetylation is also believed to compete with N-oxidation and, therefore, is considered a detoxification reaction when it occurs before the formation of the N-OH metabolites. N-glucuronidation of oxidized benzidine catalyzed by UDP-glucuronosyltransferase (UGT) is a second detoxification mechanism, because N-glucuronidation facilitates excretion. Therefore, in the case of benzidine biotransformation, phase II reactions activate and detoxify the carcinogen.
Heterocyclic Amine Biotransformation
Heterocyclic amines, found in cooked meat and fish, are initially activated to genotoxic metabolites via CYP450-mediated oxidation to the N-hydroxyl derivative (see Figure 7-6). 42 In particular, this reaction is catalyzed in the liver predominantly by CYP450 1A2 (CYP1A2). The hydroxylated heterocyclic amine metabolites are then further activated by acetyltransferases and sulfotransferases to the ultimate carcinogen, a highly reactive electrophile. GSTs and UDP-glucuronosyl transferases are thought to deactivate the ultimate carcinogen and permit elimination. Therefore, during the biotransformation of heterocyclic amines, phase II enzymes activate and detoxify the carcinogen.
Ethylene Dibromide (1,2-Dibromoethane) Biotransformation
An additional example of phase II–mediated carcinogen activation is that of the halogenated aliphatic ethyldibromide. 79 Ethylene dibromide is a potent mutagen used as an industrial solvent, gasoline lead scavenger, and fumigant. Following glutathione conjugation of the parent compound, S-2-bromoethyl glutathione spontaneously forms an episulfonium ion (see Figure 7-6). This sterically strained molecule is the reactive ultimate carcinogen and primarily attacks the N 7 position of guanine. Again, although GSTs commonly detoxify xenobiotics, glutathione conjugation of ethylene dibromide leads to carcinogen activation.
DNA Repair
A second form of endogenous defense against environmental carcinogenesis is DNA repair. Various forms of carcinogen-induced DNA damage, such as DNA adducts, DNA crosslinks, and double- and single-strand breaks, have been reported following exposure to various carcinogenic insults. To maintain genomic integrity, DNA repair genes have evolved. 80 More than 125 DNA repair enzymes and DNA damage response genes have been identified. The importance of these genes is highlighted by inherited syndromes (e.g., xeroderma pigmentosum [XP], Fanconi anemia, Bloom syndrome, and ataxia-telangiectasia) wherein DNA repair defects render the individual more susceptible to cancer development. These DNA repair proteins can be generally categorized according to the repair pathways in which they function or according to their ability to signal for or regulate DNA repair (see Chapter 4). The predominant human DNA repair pathways include base excision, nucleotide excision, base mismatch, and DNA strand break repair. Simpler, direct repair pathways have also been reported. In cases where DNA damage is excessive, crosstalk may activate apoptotic pathways. In this case, programmed cell death can be viewed as a protective host response.
Defense against Oxidative Stress
Numerous carcinogens, including UV light and heavy metals, are thought to act by inducing oxidative stress and oxidative DNA damage. Endogenous defense systems have evolved to detoxify reactive oxygen species such as hydroxyl radical and superoxide anion. These defense systems range from the free radical scavengers glutathione and vitamin E, to glutathione synthesis enzymes, to antioxidant enzymes such as glutathione transferases, peroxidases, superoxide dismutase and catalase, to DNA repair enzymes that are specialized for repair of oxidative DNA damage.
Host Susceptibility to Environmental Carcinogenesis
As described in the preceding sections, biotransformation of carcinogens, DNA repair, and other cellular events play roles in defending cells against carcinogenic insults. Sequence variation in the genes involved in these cellular processes has been described and is expected, in some cases, to alter cancer risk. In addition, susceptibility to certain carcinogens may vary according to other clinical factors such as developmental stage, disease status, or co-exposure to additional environmental agents. The combined and interacting effects of carcinogen exposure, genetic background, and clinical susceptibility determine the individual’s ensuing risk of developing cancer.
Genetic Susceptibility
For years researchers have sought to define the relationship between toxicant exposure outcome and the mediating effects of genetic polymorphisms. Genetic polymorphisms are defined as genetic variations occurring with 1% or greater prevalence in a human population. Polymorphisms can occur in the form of large deletions, small deletions, small insertions, and individual base changes, especially single-nucleotide polymorphisms (SNPs). These polymorphisms can occur in exons, introns, and promoter regions and produce a wide variety of effects ranging from changes in the protein function, to expression-level alterations, to changes in protein stability. The physiologic relevance of genetic variation may be a function of the severity of the alteration and the level of carcinogen exposure. Thus, genetic variation may be most important in the context of low-level exposure, because high-level carcinogen exposure may overwhelm any differences in effects that may result from genetic variation.
Biotransformation Enzyme Polymorphisms and Cancer Risk
Although genetic variation can influence cell signaling, cellular differentiation, apoptosis rate, cellular proliferation rate, or other biochemical processes during chemical carcinogenesis, historically more emphasis has been placed on identification of genetic determinants of carcinogen metabolism and DNA repair. SNPs occur in GST gene exons, introns, and promoter regions, and two-gene deletion polymorphisms have been described. 81 The GSTM1 and GSTT1 genes are deleted in about 50% and about 20% to 60% of the population, respectively. Polymorphisms have also been described for CYP450, NAT, and mEH genes. A number of these alterations have been shown experimentally to affect either the expression level or catalytic activities of their corresponding proteins. 82
One of the most studied biotransformation enzyme/cancer risk relationships is the relationship between GSTM1 deletion polymorphism and lung cancer risk. GSTM1 detoxifies PAHs such as those in cigarette smoke, and meta-analysis of epidemiologic data suggests that GSTM1 deficiency is a moderate risk factor for lung cancer. 71 Similarly, GSTM1 deletion may increase the risk of colon and bladder cancers. 72,73 However, some studies of GSTM1 genotype and cancer phenotype have reported inconsistent outcomes; therefore the relative contribution of this polymorphism requires further investigation. Studies of GSTM1 deletion polymorphism in the context of other carcinogen-response gene polymorphisms may be critical.
GWAS and Cancer Risk
More recently, genome-wide association studies (GWAS) have largely supplanted single-gene studies as the preferred method for probing the relationship between genetic background, carcinogen exposure, and disease risk. 83 GWAS allow high-throughput scans of the entire genome in cancer case and control groups without prior hypotheses concerning which genes or gene regions may contribute to cancer susceptibility. Numerous groups have attempted to identify the primary genetic determinants of cancer risk using this technique. Their work has revealed that multiple, common alleles each makes a small contribution to the heritability of cancer susceptibility, with no single gene or gene pathway dictating cancer risk.
Clinical Susceptibility
The relatively modest findings derived from GWAS and earlier epidemiologic studies illustrate the complex nature of predicting cancer risk phenotype based on genotype alone. Additional factors such as carcinogen dose and tissue-specific expression of biotransformation genes must be taken into account in the context of low-penetrance phenotypes. In addition, it is well known that developmental stage of the individual, concurrent diseases, co-exposures to other dietary and environmental chemicals, and the frequency, dose, and timing of carcinogen exposure affect the clinical susceptibility to cancer development. Co-exposures to other agents can alter expression or activity of phase I and phase II enzymes and confound the genotype-phenotype relationship. In addition, in case-control and cohort studies, accurate estimates of carcinogen dose are virtually impossible because of recall bias, lack of adequate exposure monitoring technology, effects of dosing frequency or timing, and other confounding factors that influence clinical susceptibility.
Researchers have begun to realize the importance of developmental stage in determining the physiologic response to carcinogen exposure. Accumulating evidence suggests that exposures during the critical periods of organogenesis and tissue differentiation may permanently affect disease risk later in life. 84 Although gene mutation in stem cells or other critical cell populations may contribute, epigenetic reprogramming is believed to be the primary mechanism whereby early life exposures affect health outcomes in childhood or adulthood. During development, epigenetic programs are in constant flux and open to environmental cues. Exposure to agents that affect DNA methylation or histone modification may permanently “reprogram” the response to carcinogens within the affected tissue. The strongest evidence for epigenetic reprogramming of cancer susceptibility comes from studies of reproductive tract and breast cancers. As an example, in utero exposure to the xenoestrogen DES is associated with a significantly increased risk of vaginal clear-cell carcinoma. Studies in animal models show that DES exposure alters DNA methylation patterns. Therefore, the stage at which an individual is exposed to environmental agents could be a critical determinant of susceptibility. Even a brief exposure to epigenetic reprogramming agents during development could potentially alter host susceptibility to carcinogenesis. Despite these inherent challenges in assessing co-exposures, temporal effects, and multiple genetic factors, the ultimate goal is to achieve cancer risk modeling that takes into account both inheritance of polymorphisms in genes encoding carcinogen defense pathways and other clinical factors that affect susceptibility.
Cancer Prevention
Because a significant fraction of cancer risk appears to be attributable to environmental factors, cancer prevention should be an attainable goal. Multiple approaches to cancer prevention have been proposed and include chemoprevention and, more simply, exposure reduction. As new products and pollutants are introduced into the environment, vigilance in hazard identification should largely prevent population-wide health crises such as those that led to the discovery of many occupational carcinogens in the 1970s and earlier. In addition, careful analysis of current dietary and other environmental exposures will increase the understanding of existing hazards. Finally, understanding of the underlying molecular mechanisms associated with the carcinogenic process will allow for the design of effective chemoprevention strategies.
Hazard Identification
Assays
An important aspect of cancer prevention is hazard identification. To effectively prevent human exposure to carcinogens, the carcinogen must be recognized as such. Hazard identification occurs via multiple avenues under the direction of numerous institutions. Academic institutes, corporations, and government agencies all contribute to the identification of carcinogenic agents. Initial screening is often conducted using short-term, in vitro techniques. Several widely used assays have been developed and measure the mutagenicity of suspected carcinogens.
Ames Assay
The Ames assay of mutagenicity utilizes Salmonella typhimurium bacterial strains with unique growth requirements to detect mutagenicity of test compounds. 85 In these assays, histidine-synthesis–deficient Salmonella strains are initially grown in the presence of exogenous histidine and are subsequently exposed to test compounds. Mutations in histidine-synthesis genes revert the bacterial strain to a histidine-independent status, which can be detected by growth in minimal-histidine media. Only those bacteria that have acquired specific mutations in histidine-synthesis genes are able to form colonies. Because bacterial strains cannot activate procarcinogens via CYP450 biotransformation, inclusion of mammalian metabolic enzymes is an important feature of this “reversion” assay.
HPRT Assay
The hypoxanthine-guanine phosphoribosyltransferase (HPRT) assay uses cultured human somatic cells to detect mutagenic agents. The normal function of HPRT in cells is to recycle nucleotide bases from degraded DNA. To detect mutations in the HPRT gene, cells are first exposed to the test compound and then exposed to a toxic nucleotide analogue, 6-thioguanine (6TG). When HPRT is nonmutated and functioning, 6TG is incorporated into DNA, triggering cell death. However, when HPRT is inactivated by mutations, no 6TG is incorporated, and the cells remain viable. Therefore, the number of surviving cells after a defined period of cell growth following 6TG exposure reflects the mutagenicity of the test agent.
Additional In Vitro Carcinogen Identification Assays
In addition to the HPRT and Ames assays, several other direct and indirect in vitro assays for the detection of genetic damage have been developed. Assays of changes at the chromosomal level in human cells include (1) the chromosome aberration assay, wherein metaphase chromosomes are examined for abnormalities; (2) the sister chromatid exchange (SCE) assay, wherein exchanges of identical pieces of chromosomes in duplicated sister chromatids are examined in metaphase cells; (3) the micronucleus assay, wherein the number of chromosome fragments referred to as micronuclei are counted; and (4) the comet (or single-cell gel electrophoresis) assay, wherein DNA strand breaks in individual cells are visualized using fluorescence microscopy.
In Vivo Assays
Two-year bioassays in rodents are currently used extensively for carcinogen identification. Whole-animal assays are conducted to determine the carcinogenic potential of an agent when delivered over the lifespan, in a more physiologically relevant model. Of the approximately 200 agents classified as human carcinogens, almost all have been shown to cause cancer in rats or mice, highlighting the utility of animal studies in the identification of carcinogens. Rodents are administered the test compound via the exposure route most relevant to human scenarios at two doses: the maximum tolerated dose (MTD) and one half the MTD. The compound is administered for a majority of the lifespan of the animal, and tumor incidence at all sites is recorded. Generally, the rat is recommended for the first 2-year carcinogenicity study. These data are then supplemented with additional short- or medium-term in vivo studies or with a 2-year carcinogenicity study in another rodent species such as the mouse. Short- and medium-term testing may include the use of transgenic or “knockout” mouse models wherein an oncogene is overexpressed or a tumor suppressor gene allele is missing, although the validity of using these genetically altered models is still under debate.
The costly nature of in vivo screens (more than 800 rodents and histopathological analysis of more than 40 tissues) limits their utility. 86 Only approximately 1500 chemicals have been adequately examined. Recent effort has been directed toward developing predictive models for more thorough or higher throughput screening of new drugs and other agents. For example, the vast majority of known carcinogens is mutagenic or genotoxic; therefore, several groups have proposed an expanded battery of DNA-based tests, including tests of DNA adduct formation, DNA strand breaks, and DNA repair. In addition, toxicogenomic strategies for the identification of gene expression profiles that are predictive of genotoxic and non-genotoxic carcinogenicity have been proposed.
Non-genotoxic Carcinogens
The identification and analysis of non-genotoxic carcinogens are less straightforward than are those for genotoxic agents. These agents are identified in the context of the 2-year rodent bioassay. Agents that are identified as carcinogenic in these in vivo assays but do not directly interact with DNA are classified as non-genotoxic carcinogens. 87 Non-genotoxic carcinogens characteristically induce tumors in only one or a few species and only after a threshold dose is achieved. Non-genotoxic carcinogens are not detected in in vitro assays of mutagenicity such as the Ames or HPRT assays. Many of these non-genotoxic carcinogens possess properties similar to tumor promoters, suggesting, together with a lack of genotoxicity, that they work mechanistically differently from classical, genotoxic carcinogens. Considerable debate is ongoing concerning the best way to identify and regulate such compounds (see subsequent sections).
Risk Assessment and Regulation of Carcinogen Exposure
As carcinogen exposure scenarios are identified, an assessment of predicted exposure dose and the expected degree of health hazard is conducted. This informs an estimation of overall cancer risk associated with the observed exposure. Estimating cancer risk helps investigators determine when and if behavior modifications should be enforced. This process of predicting cancer risk in a given exposure scenario is referred to as risk assessment, whereas the response to predicted risk is referred to as risk management. The EPA is responsible for risk assessment in areas of known or suspected exposure of the population to carcinogens and makes recommendations for risk management to minimize health consequences due to environmental contamination.
Risk assessment concerning mutagenic carcinogen exposure assumes that no threshold dose exists. That is, no safe exposure level can be identified because any exposure dose could, in theory, induce a mutation in a critical target gene, thereby elevating cancer risk. Extrapolation of a safe level of human exposure to non-genotoxic carcinogens is more complex and requires multiple assumptions. For instance, it is assumed that an agent found to be a non-genotoxic carcinogen in rodents would be toxic to humans and that the no observable adverse effect level (NOAEL) in rodents could be applied to humans. Such decisions are greatly enhanced by mechanistic information so that judgments can be made concerning potential threat to human health. In 2005, the EPA released new risk assessment guidelines that acknowledge the different mode of action for non-genotoxic carcinogens and provide for this deviation during risk-management decision making.
Prevention Strategies
The goal of risk assessment and risk management is to prevent cancer by anticipating and circumventing carcinogen exposure. However, the etiology of certain cancers is still unknown. In many cases, risk assessment is impossible or risk management measures are unavailable. Furthermore, some carcinogen exposures are unavoidable, or avoidance is not practically feasible. For instance, therapeutic radiation and certain chemotherapy drugs are known carcinogens; however, the risk-to-benefit ratio still favors voluntary exposure, despite health risk. In these instances, prevention tactics are needed to counteract the carcinogenic process, especially in the absence of effective treatment options. Several approaches to prevention have been taken in recent years with varying degrees of promise. 88
Vaccination
Vaccination is among the most promising of approaches for biologic carcinogens such as HBV, HPV, and Helicobacter pylori. 89 The development of vaccines to block initial infection with carcinogenic bacteria or virus would presumably prevent or reduce associated cancers. As an example, HPV vaccines have been developed to limit the spread of the virus and reduce the incidence of cervical cancer. In addition to this traditional use of vaccination, the use of vaccines against oncoantigens has also been proposed to prevent cancer via stimulating immune mechanisms to attack small cancerous lesions. Oncoantigens, which are tumor-associated molecules, are used to stimulate persistent immune memory mechanisms. When the antigen is later detected via immune surveillance, an effective adaptive immune response is mounted. In theory, the immune system is primed to detect and destroy any cancer cells expressing the oncoantigen. The success of vaccines in the prevention of tumors in animal models has been documented; however, the utility of such vaccines to prevent human tumors must still be validated.
Chemoprevention
Chemoprevention strategies for reducing the incidence of cancer have also been proposed. For instance, chemicals that upregulate biotransformation enzymes (in particular, phase II enzymes) have been investigated as chemopreventive agents. 75,90 Because most phase II biotransformation reactions reduce chemical reactivity of the parent compound, the rationale for inducing phase II enzymes or their cofactors is to reduce the mutagenicity of initiating agents. For instance, oltipraz administration has been shown to attenuate AFB1 toxicity in rats. Oltipraz elevates GST activity, likely via activation of antioxidant response elements within GST promoter regions. Oltipraz may also inhibit the activation of aflatoxin by CYP450. The challenge associated with enzyme induction as a chemopreventive approach is that not all phase II biotransformation reactions are detoxification reactions. Because humans are exposed to a wide variety of carcinogens, induction of biotransformation enzymes may be simultaneously beneficial and detrimental. Therefore, the decision to use of this type of chemopreventive agent must weigh multiple factors such as carcinogen target organ, agent distribution, and exposure scenario.
In addition to agents that induce the expression of detoxification enzymes, agents that combat or prevent oxidative stress are potential chemopreventive agents. Oxidative stress is believed to contribute to the formation of multiple cancer types; consequently, treatment with antioxidant agents may block carcinogenesis. In this regard, selenium, α-tocopherol, EGCG, and lycopene are potent antioxidants under study for chemopreventive properties. Similarly, inflammation is believed to contribute to formation of numerous cancer types. Agents such as cyclooxygenase-2 (COX-2) inhibitors (i.e., celecoxib) as well as other nonsteroidal antiinflammatory drugs (NSAIDs) have been proposed as chemopreventive agents to combat procarcinogenic inflammation. In addition, hormonal agents have been proposed for the chemoprevention of cancers of reproductive organs such as breast and prostate cancers. 91 For instance, selective estrogen receptor modulators (SERMs) have been proposed to prevent breast cancer by blocking the action of procarcinogenic estrogen. Tamoxifen, an antiestrogenic agent, was first approved for the treatment of advanced breast cancer but also reduced the risk of contralateral breast cancer occurrence. Tamoxifen has since been approved as a chemopreventive agent in high-risk patients.
Summary and Conclusions
Cancer is known to develop over many years and is determined by the interaction of host genetic factors as well as environmental exposures. Environmental factors appear to play a major role in determining cancer risk. Of the known cancer risk factors, smoking and diet account for a significant proportion of cancer deaths. There are many types of environmental carcinogens including biologic agents (e.g., viruses), chemicals (e.g., PAH), and physical agents (e.g., solar radiation). The linkage between environmental exposure and cancer in humans is strong in some cases (e.g., asbestos and mesothelioma of the lung), whereas in other cases the environmental etiologic factors are less well understood (e.g., breast and prostate cancers). Epidemiology studies, together with studies in model systems, especially animal model systems, provide the evidence used to determine the relative risk of specific environmental exposures. Categorizing cancer risk from environmental agents is an ongoing process conducted by the NTP and the IARC. Study of genetic polymorphisms in various genes involved in the carcinogenic process is leading to a better understanding of the overall risk associated with environmental exposures and identification of high-risk populations to target prevention strategies. Research on the underlying mechanisms associated with environmental carcinogenesis provides the basis for early detection and identification of target molecules for chemoprevention and/or intervention strategies. Carcinogens are known to target oncogenes and tumor suppressor genes through DNA damage and/or to alter cellular signaling pathways in bringing about the changes associated with cancer formation in specific tissues. Ultimately, environmental carcinogenesis occurs via the stepwise accumulation of genetic alterations leading to invasive and metastatic lesions. Finally, although many regulatory mechanisms exist to protect the public, diligence is required to guard them from future unintended carcinogen exposures. It will remain prudent to closely monitor the environment for potential human carcinogens.
References
1. Nature and nurture—lessons from chemical carcinogenesis . Nat Rev Cancer . 2005 ; 5 : 113 – 125 .
2. Historical overview of chemical carcinogenesis . In: Penning T.M. , ed. Chemical Carcinogenesis . New York, NY : Springer ; 2011 : 1 – 26 .
3. The identification of a carcinogenic compound in coal-tar . Br Med J . 1955 ; 2 ( 4942 ) : 749 – 752 .
4. Global patterns of cancer incidence and mortality rates and trends . Cancer Epidemiol Biomarkers Prev . 2010 ; 19 ( 8 ) : 1893 – 1907 .
5. Migration patterns and breast cancer risk in Asian-American women . J Natl Cancer Inst . 1993 ; 85 ( 22 ) : 1819 – 1827 .
6. Worldwide variations in colorectal cancer . CA Cancer J Clin . 2009 ; 59 ( 6 ) : 366 – 378 .
7. Cancer epidemiology . In: Pollock R.E. , ed. UICC Manual of Clinical Oncology . New York, NY : John Wiley & Sons, Inc ; 2004 : 91 – 120 .
8. The causes of cancer: quantitative estimates of avoidable risks of cancer in the United States today . J Natl Cancer Inst . 1981 ; 66 ( 6 ) : 1191 – 1308 .
9. Progress and challenges in selected areas of tobacco carcinogenesis . Chem Res Toxicol . 2008 ; 21 ( 1 ) : 160 – 171 .
10. Tobacco smoke carcinogens and lung cancer . In: Penning T.M. , ed. Chemical Carcinogenesis . New York, NY : Springer ; 2011 : 53 – 74 .
11. Molecular mechanisms of cancer development in obesity . Nat Rev Cancer . 2011 ; 11 ( 12 ) : 886 – 895 .
12. Obesity, energy balance, and cancer: new opportunities for prevention . Cancer Prev Res (Phila) . 2012 ; 5 ( 11 ) : 1260 – 1272 .
13. Environmental causes of human cancers . Eur J Cancer . 2001 ; 37 ( Suppl 8 ) : S67 – S87 .
14. Cancer chemoprevention through dietary antioxidants: progress and promise . Antioxid Redox Signal . 2008 ; 10 ( 3 ) : 475 – 510 .
15. Urothelial carcinoma associated with the use of a Chinese herb (Aristolochia fangchi) . N Engl J Med . 2000 ; 342 ( 23 ) : 1686 – 1692 .
16. Ionizing radiation as a carcinogen . In: Sipes G. , McQueen C.A. , Gandolfi A.J. , eds. Comprehensive Toxicology . New York, NY : Elsevier Science, Ltd ; 1997 : 281 – 302 .
17. Carcinogenic aromatic amines and amides . In: Sipes G. , McQueen C.A. , Gandolfi A.J. , eds. Comprehensive Toxicology . New York, NY : Elsevier Science, Ltd ; 1997 : 141 – 170 .
18. Vinyl chloride and the liver . J Hepatol . 2009 ; 51 ( 6 ) : 1074 – 1081 .
19. . Environmental Factors in Cancer . Indianapolis, Ind. Bethesda, Md : Dept. of Health & Human Services National Institutes of Health, National Cancer Institute ; October 21, 2008 .
20. Chemical carcinogenesis . In: Klaassen C.D. , ed. Casarett & Doull’s Toxicology, The Basic Science of Poisons . New York, NY : McGraw-Hill ; 1996 : 201 – 267 .
21. Comprehensive toxicology . In: Bowden G.T. , Fischer S.M. , eds. Chemical Carcinogens and Anticarcinogens . Vol 12 . New York, NY : Elsevier Science Inc ; 1997 .
22. Ultraviolet radiation and skin cancer . Int J Dermatol . 2010 ; 49 ( 9 ) : 978 – 986 .
23. Cell cycle kinetics following UVA irradiation in comparison to UVB and UVC irradiation . Photochem Photobiol . 1996 ; 63 ( 4 ) : 492 – 497 .
24. The optics of human skin . J Invest Dermatol . 1981 ; 77 ( 1 ) : 13 – 19 .
25. Early events in UV carcinogenesis—DNA damage, target cells and mutant p53 foci . Photochem Photobiol . 2008 ; 84 ( 2 ) : 382 – 387 .
26. From genes to drugs: targeted strategies for melanoma . Nat Rev Cancer . 2012 ; 12 ( 5 ) : 349 – 361 .
27. The cancer epidemiology of radiation . Oncogene . 2004 ; 23 ( 38 ) : 6404 – 6428 .
28. Radiation carcinogenesis . Carcinogenesis . 2000 ; 21 ( 3 ) : 397 – 404 .
29. What are the risks of low-level exposure to alpha radiation from radon? Proc Natl Acad Sci U S A . 1997 ; 94 ( 12 ) : 5996 – 5997 .
30. Polycyclic aromatic hydrocarbon carcinogenesis . In: Harvey R.G. , ed. Polycyclic Hydrocarbons and Carcinogenesis. ACS Symposium Series 283 . Philadelphia, Pa : American Chemical Society ; 1985 : 1 – 17 .
31. Benzene in the environment: an assessment of the potential risks to the health of the population . Occup Environ Med . 2001 ; 58 ( 1 ) : 2 – 13 .
32. Current understanding of the mechanism of benzene-induced leukemia in humans: implications for risk assessment . Carcinogenesis . 2012 ; 33 ( 2 ) : 240 – 252 .
33. Covalent polycyclic aromatic hydrocarbon-DNA adducts: carcinogenicity, structure and function . In: Penning T.M. , ed. Chemical Carcinogenesis . New York, NY : Springer ; 2011 : 181 – 208 .
34. Synergistic mechanisms in carcinogenesis by polycyclic aromatic hydrocarbons and by tobacco smoke: a bio-historical perspective with updates . Carcinogenesis . 2001 ; 22 ( 12 ) : 1903 – 1930 .
35. Aflatoxin: a 50-year odyssey of mechanistic and translational toxicology . Toxicol Sci . 2011 ; 120 ( Suppl 1 ) : S28 – S48 .
36. Aflatoxin and hepatocellular carcinoma . In: Penning T.M. , ed. Chemical Carcinogenesis . New York, NY : Springer ; 2011 : 113 – 134 .
37. Carcinogenicity of azo colorants: influence of solubility and bioavailability . Toxicol Lett . 2004 ; 151 ( 1 ) : 203 – 210 .
38. Metabolism of N-acetylbenzidine and initiation of bladder cancer . Mutat Res . 2002 ; 506-507 : 29 – 40 .
39. Benzidine mechanistic data and risk assessment: species- and organ-specific metabolic activation . Pharmacol Ther . 1996 ; 71 ( 1-2 ) : 107 – 126 .
40. Risk of colorectal and other gastro-intestinal cancers after exposure to nitrate, nitrite and N-nitroso compounds: a follow-up study . Int J Cancer . 1999 ; 80 ( 6 ) : 852 – 856 .
41. Formation and human risk of carcinogenic heterocyclic amines formed from natural precursors in meat . Nutr Rev . 2005 ; 63 ( 5 ) : 158 – 165 .
42. Meat-related mutagens/carcinogens in the etiology of colorectal cancer . Environ Mol Mutagen . 2004 ; 44 ( 1 ) : 44 – 55 .
43. Biochemistry, biology, and carcinogenicity of tobacco-specific N-nitrosamines . Chem Res Toxicol . 1998 ; 11 ( 6 ) : 559 – 603 .
44. Beryllium: genotoxicity and carcinogenicity . Mutat Res . 2003 ; 533 ( 1-2 ) : 99 – 105 .
45. Occupational exposure to beryllium and cancer risk: a review of the epidemiologic evidence . Crit Rev Toxicol . 2012 ; 42 ( 2 ) : 107 – 118 .
46. Beryllium and lung cancer: a weight of evidence evaluation of the toxicological and epidemiological literature . Crit Rev Toxicol . 2009 ; 39 ( Suppl 1 ) : 1 – 32 .
47. Mechanisms of cadmium carcinogenesis . Toxicol Appl Pharmacol . 2009 ; 238 ( 3 ) : 272 – 279 .
48. Arsenic in the aetiology of cancer . Mutat Res . 2006 ; 612 ( 3 ) : 215 – 246 .
49. Some drinking-water disinfectants and contaminants, including arsenic . IARC Monogr Eval Carcinog Risks Hum . 2004 ; 84 : 1 – 477 .
50. Molecular mechanisms of arsenic carcinogenesis . Mol Cell Biochem . 2004 ; 255 ( 1-2 ) : 57 – 66 .
51. An emerging role for epigenetic dysregulation in arsenic toxicity and carcinogenesis . Environ Health Perspect . 2011 ; 119 ( 1 ) : 11 – 19 .
52. Chromium in drinking water: sources, metabolism, and cancer risks . Chem Res Toxicol . 2011 ; 24 ( 10 ) : 1617 – 1629 .
53. Molecular mechanisms of Cr(VI)-induced carcinogenesis . Mol Cell Biochem . 2002 ; 234-235 ( 1-2 ) : 293 – 300 .
54. Malignant mesothelioma . Lancet . 2005 ; 366 ( 9483 ) : 397 – 408 .
55. Molecular pathways: targeting mechanisms of asbestos and erionite carcinogenesis in mesothelioma . Clin Cancer Res . 2012 ; 18 ( 3 ) : 598 – 604 .
56. Case-control study of lung cancer among sugar cane farmers in India . Occup Environ Med . 1999 ; 56 ( 8 ) : 548 – 552 .
57. Prevention of hormone-related cancers: breast cancer . J Clin Oncol . 2005 ; 23 ( 2 ) : 357 – 367 .
58. Prevention of hormone-related cancers: prostate cancer . J Clin Oncol . 2005 ; 23 ( 2 ) : 368 – 377 .
59. Multistage carcinogenesis . In: Penning T.M. , ed. Chemical Carcinogenesis . New York, NY : Springer ; 2011 : 27 – 52 .
60. Hallmarks of cancer: the next generation . Cell . 2011 ; 144 ( 5 ) : 646 – 674 .
61. Multi-stage chemical carcinogenesis in mouse skin: fundamentals and applications . Nat Protoc . 2009 ; 4 ( 9 ) : 1350 – 1362 .
62. Evidence for the binding of polynuclear aromatic hydrocarbons to the nucleic acids of mouse skin: relation between carcinogenic power of hydrocarbons and their binding to deoxyribonucleic acid . Nature . 1964 ; 202 : 781 – 784 .
63. Growth factor signaling pathways as targets for prevention of epithelial carcinogenesis . Mol Carcinog . 2011 ; 50 ( 4 ) : 264 – 279 .
64. Disruption of Stat3 reveals a critical role in both the initiation and the promotion stages of epithelial carcinogenesis . J Clin Invest . 2004 ; 114 ( 5 ) : 720 – 728 .
65. Role of PI3K/Akt signaling in insulin-like growth factor-1 (IGF-1) skin tumor promotion . Mol Carcinog . 2005 ; 44 ( 2 ) : 137 – 145 .
66. Mechanisms of tumor promotion . In: Slaga T.J. , ed. Tumor Promotion in Internal Organs . Vol. 1 . Boca Raton, Fla : CRC Press, Inc ; 1983 .
67. Models of carcinogenesis: an overview . Carcinogenesis . 2010 ; 31 ( 10 ) : 1703 – 1709 .
68. The role of epigenetics in environmental and occupational carcinogenesis . Chem Biol Interact . 2010 ; 188 ( 2 ) : 340 – 349 .
69. Epigenetics and cancer: towards an evaluation of the impact of environmental and dietary factors . Mutagenesis . 2007 ; 22 ( 2 ) : 91 – 103 .
70. Modelling the molecular circuitry of cancer . Nat Rev Cancer . 2002 ; 2 ( 5 ) : 331 – 341 .
71. Lessons from hereditary colorectal cancer . Cell . 1996 ; 87 ( 2 ) : 159 – 170 .
72. Sequential molecular abnormalities are involved in the multistage development of squamous cell lung carcinoma . Oncogene . 1999 ; 18 ( 3 ) : 643 – 650 .
73. TP53 Mutations in human cancers: origins, consequences and clinical use . In: Levine A. , Lane D. , eds. Cold Spring Harbor Perspectives in Biology . Cold Spring Harbor, NY : Cold Spring Harbor Laboratory Press ; 2012 : 1 – 17 .
74. Preferential formation of benzo[a]pyrene adducts at lung cancer mutational hotspots in P53 . Science . 1996 ; 274 ( 5286 ) : 430 – 432 .
75. Detoxification of chemical carcinogens and chemoprevention . In: Penning T.M. , ed. Chemical Carcinogenesis . New York, NY : Springer ; 2011 : 159 – 180 .
76. Role of cytochrome p4501 family members in the metabolic activation of polycyclic aromatic hydrocarbons in mouse epidermis . Chem Res Toxicol . 2004 ; 17 ( 12 ) : 1667 – 1674 .
77. Concise review of the glutathione S-transferases and their significance to toxicology . Toxicol Sci . 1999 ; 49 ( 2 ) : 156 – 164 .
78. Identification of amino acid residues essential for high aflatoxin B1-8,9-epoxide conjugation activity in alpha class glutathione S-transferases through site-directed mutagenesis . Toxicol Appl Pharmacol . 1998 ; 152 ( 1 ) : 166 – 174 .
79. Evidence for formation of an S-[2-(N7-guanyl)ethyl]glutathione adduct in glutathione-mediated binding of the carcinogen 1,2-dibromoethane to DNA . Proc Natl Acad Sci U S A . 1983 ; 80 ( 17 ) : 5266 – 5270 .
80. DNA polymerases and cancer . Nat Rev Cancer . 2011 ; 11 ( 2 ) : 96 – 110 .
81. Glutathione S-transferase polymorphisms: cancer incidence and therapy . Oncogene . 2006 ; 25 ( 11 ) : 1639 – 1648 .
82. Role of xenobiotic metabolic enzymes in cancer epidemiology . Methods Mol Biol . 2009 ; 472 : 243 – 264 .
83. Current status of genome-wide association studies in cancer . Hum Genet . 2011 ; 130 ( 1 ) : 59 – 78 .
84. Developmental reprogramming of cancer susceptibility . Nat Rev Cancer . 2012 ; 12 ( 7 ) : 479 – 486 .
85. The Ames Salmonella/microsome mutagenicity assay . Mutat Res . 2000 ; 455 ( 1-2 ) : 29 – 60 .
86. Characterizing and predicting carcinogenicity and mode of action using conventional and toxicogenomics methods . Mutat Res . 2010 ; 705 ( 3 ) : 184 – 200 .
87. Mechanisms of nongenotoxic carcinogenesis and assessment of the human hazard . Regul Toxicol Pharmacol . 2000 ; 32 ( 2 ) : 135 – 143 .
88. Cancer prevention: from 1727 to milestones of the past 100 years . Cancer Res . 2009 ; 69 ( 13 ) : 5269 – 5284 .
89. Vaccines for tumour prevention . Nat Rev Cancer . 2006 ; 6 ( 3 ) : 204 – 216 .
90. Modulation of xenobiotic metabolising enzymes by anticarcinogens—focus on glutathione S-transferases and their role as targets of dietary chemoprevention in colorectal carcinogenesis . Mutat Res . 2005 ; 591 ( 1-2 ) : 74 – 92 .
91. An update on chemoprevention strategies in prostate cancer for 2006 . Curr Opin Urol . 2006 ; 16 ( 3 ) : 132 – 137 .
