In 1713 Bernardino Ramazzini, of whom we have already spoken, published a text entitled De Morbis Artificum, or Diseases of Workers, which was translated in 1964 by Wright. It is in this text that Ramazzini describes the high incidence of mammary cancer in nuns, attributing it to their celibate life. However, the remainder of the book does not consider neoplastic disease to any extent in the more than 50 occupations described. Scrotal cancer in one-time chimney sweeps, described by Pott, was a later example of occupational cancer. During the nineteenth century, several reports of the association of specific cancers with mining, smelting, dyeing and lubrication processes and industries were reported. Unfortunately, in many instances, little was done for many years to protect the worker, a neglect of which we are still not entirely free today (Table 11.9).
The association of occupational exposure to asbestos and the subsequent development of bronchogenic carcinoma and mesothelioma have been well established (cf. Selikoff, 1978). However, in virtually all of the studies undertaken in this field, the highest incidence of bron- chogenic carcinoma was found in those individuals exposed to asbestos who also had a history of cigarette smoking. In fact, risk ratios for the development of bronchogenic carcinoma of a
Table 11.9 Carcinogenic Risks Associated with Occupations
large cohort of asbestos insulation workers in the United States and Canada as well as asbestos miners in Finland showed a dramatic increase in risk with smoking as compared with either smoking or asbestos exposure alone (Thiringer and Järvholm, 1980). This relationship can be seen in Table 11.10.
Hammond and Selikoff (1973) have suggested that this differential indicates the probable necessity for multiple environmental agents for the induction of bronchogenic carcinoma in the individuals exposed to asbestos. Carcinomas occurring in smokers exposed to asbestos exhibit a greater incidence of adenocarcinoma than those seen in smokers alone (Mollo et al., 1990; Johansson et al., 1992). In contrast, there appears to be little or no association between cigarette smoking and the occurrence of mesothelioma (Muscat and Wynder, 1991). Most pleural and peritoneal mesotheliomas in men can be attributed to exposure to asbestos, although such an association is less clear in women (Spirtas et al., 1994). The latency period from the first expo- sure to the diagnosis of mesothelioma may range from 20 to 45 years (Roggli, 1995; cf. Pisani et al., 1988). Asbestos fibers differ in structure. The predominant types are noted in Table 11.11, together with the evidence of their carcinogenicity in inducing mesotheliomas in the human (Pisani et al., 1988). Virtually all fiber types are carcinogenic (Huncharek, 1994) with the possi- ble exception of anthophyllite (Meurman et al., 1974), as noted in the table. Further evidence in support of the direct carcinogenic effect of asbestos is the fact that mesotheliomas can be in- duced in rodents by appropriate exposure to this material (Wagner and Berry, 1969). Although other fibrous materials such as fiberglass have been shown to be carcinogenic in the rodent (Stanton et al., 1977), there is essentially no evidence to date that fiberglass causes human cancer.
The mechanism of asbestos carcinogenesis is unclear. However, asbestos fibers can cause cytogenetic effects in cell culture as well as “transformation” (Chapter 14) of Syrian hamster embryo cells in culture (Oshimura et al., 1984; Dopp et al., 1995). Hei et al. (1992) have also demonstrated that asbestos fibers may induce large deletions in hamster-derived cells in culture; this could be related to its clastogenic effects. In addition, asbestos fibers are able to generate reactive oxygen species in vitro, suggesting a molecular mechanism for the cytogenetic and transformation effects noted (Moyer et al., 1994). Although the mechanism of asbestos induc- tion of cancer is as yet unknown, the demonstration that the type and size of fiber are important for the carcinogenicity of this material indicates that its carcinogenic action may be similar to that of “plastic film” carcinogenesis (Chapter 3). Several inorganic compounds and their primary elements have also been shown to be carcinogenic in the human (Table 11.9). There is only lim- ited evidence that cadmium and its compounds are carcinogenic for the human, the best evi- dence being pulmonary carcinogenesis in epidemiological studies, where an actual dose- response relationship may exist (cf. Waalkes et al., 1992). The evidence is significantly less for a causative factor in human prostate cancer (Waalkes and Rehm, 1994). On the other hand, chro- mium compounds have been causally related to cancers of the respiratory system, especially the lung, as well as the stomach, bone, and the urogenital tract (cf. Cohen et al., 1993; Costa, 1997).
As noted earlier (Chapter 3), there is substantial evidence for the interaction of chromium com- pounds with DNA, induction of protein-DNA crosslinks, and direct mutagenic effects on DNA (cf. Cohen et al., 1993). Chromium compounds are clastogenic in humans and animals both in vivo and in vitro (cf. Vainio and Sorsa, 1981). Exposure to nickel-containing dust in refineries and other working environments is causally related to cancers of the respiratory tract, especially the nose and lungs, after exposure periods extending from 5 to 25 years (cf. Magos, 1991). In most instances such neoplastic development was associated with other evidence of toxicities, such as chronic pulmonary irritation, rhinitis, and sinusitis.
A situation similar to the problem of asbestos exposure may be seen in underground hema- tite miners in several areas of Europe as well as in this country. Since iron oxide has not been found to be carcinogenic in laboratory animals—although hematite dust enhances the carcino- genic activity of hydrocarbons in the lung—the exact agent or agents producing the increased incidence of lung cancer in this occupational group have not yet been identified. Similarly, workers in the furniture industry and related fields exposed to high levels of wood dust, espe- cially from hardwood (Leclerc et al., 1994), are more prone to develop adenocarcinoma of the nasal cavity and sinuses. A similar finding has been demonstrated for workers in the shoe indus- try in England; again, the offending agents are unknown. In previous years, during the manufac- ture of isopropyl alcohol, oils were produced whose exact structure was not known but whose presence was obviously related to the process. Workers in this industry had a significant increase in cancer of the larynx and nasal cavity after chronic inhalation of these oils. Today the manufac- turing process has been changed so that there is no risk to the worker.
One of the most interesting inorganic elements whose compounds have been shown to be carcinogenic in the human is arsenic. As indicated in Chapter 3, the evidence for the carcino- genic effects of arsenic is far better in the human than in experimental animals. Only in the last two decades has it been shown that arsenic compounds are carcinogenic in rodents, primarily if not exclusively in the Syrian golden hamster, without the addition of other agents (Yamamoto et al., 1987). More recently, a number of studies have demonstrated that organic arsenicals, espe- cially methylated arsenicals, may act as promoting or progressor agents in several different ex- perimental systems and species (Yamanaka et al., 1996; Hayashi et al., 1998; Yamamoto et al.,1997; Germolec et al., 1997). Furthermore, in cells isolated from the Syrian hamster embryo and explanted to tissue culture, arsenical compounds in the medium induced cell transformation (Chapter 14) and cytogenetic effects (Lee et al., 1985). One of the earliest extensive reviews of arsenical cancer in the human was that of Neubauer (1947), in which he described clinical char- acteristics of cancer of the skin associated with ingestion of medicinals as well as occupational exposures. Since that time and especially during the last decade, there have been extensive epidemiological investigations on various exposures to arsenic compounds by a variety of envi- ronmental routes. A related environmental exposure to arsenic is in drinking water. One of the most extensively studied cohorts of individuals exposed to high levels of arsenic in drinking wa- ter was described in Taiwan (Chen et al., 1992). The concentration of arsenic in the well water was associated significantly with an increased risk for cancers of the liver, nasal cavity, lung, skin, bladder, and kidney in both males and females. While some have perhaps questioned the applicability of these findings to other situations, at least two other studies (Hopenhayn-Rich et al., 1996; Tsuda et al., 1995) have obtained similar evidence for the effect of arsenic in drinking water on the development of bladder and lung cancer. These findings have considerable practical importance in the United States, where a very large proportion of drinking water contains low levels of inorganic arsenic. However, a recent study by Buchet and Lison (1998) from Belgium has indicated that a low to moderate level of environmental exposure to inorganic arsenic (20 to 50 µg/L drinking water) does not seem to affect the causes of mortality in the exposed popula- tion. Generally, from these epidemiological studies, one may propose that inhaled arsenic in- duces primarily respiratory cancer, whereas ingested arsenic is associated with increased risk of cancer at multiple sites, including the skin and a variety of organs (Byrd et al., 1996). Further- more, from the epidemiological investigations as well as experimental studies demonstrating the effective clastogenicity of arsenic and its compounds while producing an absence of point muta- tions (Stöhrer, 1991), as well as the effects on multistage carcinogenesis in animals, arsenicals may be more closely related to progressor agents (Chapter 9), exerting their effects at the late, progressor stage of the carcinogenic process (Brown and Chu, 1983).
The use of specific organic chemicals in industrial processes dates back to the middle of the last century; the carcinogenic effects of several of these chemicals have been completely or partially defined in the human species. The earliest of these were some aromatic amines used in the dye industry. In 1895, Rehn noted bladder neoplasms in three men working in a dye factory that used aniline as the basic ingredient for the preparation of fuchsin dyes (Rehn, 1895). Since that time, bladder neoplasms among aromatic amine workers have been referred to as “aniline tumors,” although it is now known that the bladder carcinogen in Rehn’s report was not aniline but 4-aminobiphenyl (Connolly and White, 1969). A related and also extensively produced chemical was β-naphthylamine, which was purified by distillation and sublimation in closed rooms during the last century. As noted in Figure 11.7, these distillation conditions resulted in the development of bladder cancer in virtually 100% of the individuals who worked in those rooms. While the danger to the human of the aromatic amines listed in Table 11.9 is now well known and considerable safety precautions are employed where such chemicals are used, there has still been some recent evidence that other aromatic amines, such as o-toluidine, may be a causal agents for bladder cancer in some industries (Ward et al., 1996).
A far more pervasive and common environmental chemical that has been designated as carcinogenic for the human is benzene. Although the clear carcinogenic effect of benzene in laboratory test animals has been demonstrated conclusively only within the last two decades, an association between exposure to benzene and leukemia has been known since before 1930 (De- lore and Borgomano, 1928). In chronic bioassay studies, benzene induced neoplasms in a num- ber of sites in both rats and mice (Huff et al., 1989). In the human, the toxicity for the hematopoietic system has been well documented, the most common effect being aplastic anemia (Smith, 1996). Although workers in the chemical and rubber industries may be exposed to greater concentrations of benzene for longer periods of time than the general population, expo-
Figure 11.7 Increasing risk of bladder carcinoma following exposure to various aromatic amine carcin- ogens. Note the 100% incidence in those distilling 2-naphthylamine. (From Connolly and White, 1969, with permission of the authors and publisher.)
sure to benzene is pervasive (Figure 11.8). As noted in Figure 11.8, a typical smoker takes in roughly 2 mg of benzene per day, primarily from mainstream cigarette smoke. Nonsmokers in- hale about one-tenth of this amount, predominantly from ambient air but also from automobile vapors (Wallace, 1996). Outdoor air contributes about 40% of the total benzene dose for non- smokers, while the remainder involves primarily gasoline and automobile vapor emissions asso- ciated with automobile driving and maintenance as well as environmental tobacco smoke (ETS) exposures. As expected from its myelotoxicity, there is strong evidence that benzene induces leukemia, predominantly if not exclusively acute myelogenous leukemia, as noted from the combination of six studies reported by Lamm et al. (1989) (Figure 11.9). Although there is some suggestion that benzene may induce multiple myeloma (Goldstein, 1990), other studies have not borne out this suggestion (Wong, 1995).
Medinsky et al. (1994) have pointed out that acute myelogenous leukemia appears to in- volve repeated exposure to cytotoxic concentrations of benzene. Other studies (Wong, 1995; Schnatter et al., 1996) have demonstrated apparent thresholds of benzene exposure below which there is no evidence of neoplastic development. Smith (1996) has pointed out that benzene is not a “classic” chemical carcinogen in that it does not form a highly electrophilic metabolite on me- tabolism, and there is relatively little binding to DNA of radiolabeled benzene or its metabolites, although DNA adducts can be demonstrated by the 32P-postlabeling technique (Bodell et al.,1996). However, both benzene and its metabolites induce chromosomal alterations in both hu- mans and animals (cf. Snyder and Kalf, 1994). Chronic exposure to benzene induces breaks and translocations in specific chromosomes in vivo, which are significant at exposures of greater than 31 ppm (Figure 11.10) (Smith et al., 1998). Thus, while there has been much speculation as to the mechanism(s) of benzene carcinogenesis, it would appear that the known characteristics
Figure 11.8 Sources of benzene exposure in smokers (A) and nonsmokers (B). A typical smoker takes in roughly 2 mg of benzene per day, while a nonsmoker typically inhales about 0.2 mg of benzene per day, assuming in the latter incidence an average exposure of 15 µg/m3 and an alveolar respiration rate of 14 m3/day. (Adapted from Wallace, 1996, with permission of the author and publisher.)
of this agent allow its classification as a progressor agent in human neoplastic development (Chapter 9).
Although the dioxins and polyhalogenated biphenyls induce neoplasms in experimental animals, it is important to note here the present controversy that suggests an effect of dioxins on the induction of soft tissue sarcomas. Hardell et al. (1995) have presented evidence that chronic exposure to phenoxy acids or chlorophenols is associated with significant increases in soft tissue sarcomas in Swedish workers, but others (cf. Bond and Rossbacher, 1993) were not able to con- firm this in a similar study in Finland. Since these compounds are usually contaminated by diox- ins and related toxic compounds, the argument has been made that such studies indicate the potential carcinogenicity of dioxins for humans. 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)
Figure 11.9 The percent observed/percent expected by type of leukemia for six studies as described by Lamm et al. (1989). ALL, acute lymphatic leukemia; CLL, chronic lymphatic leukemia; AML, acute myelogenous leukemia; CML, chronic myelogenous leukemia. (Reproduced with permission of the authors and publishers.)
and related halogenated compounds are pervasive environmental chemical contaminants which are of no practical or useful importance. TCDD and many of its congeners are minor by-prod- ucts of a variety of synthetic processes utilized in the preparation of antiseptics and herbicides. They are also produced by incomplete combustion of a variety of halogen-containing organic chemicals and thus occur in stack exhausts and cigarette smoke. TCDD is extremely toxic to a variety of species (cf. Pohjanvirta and Tuomisto, 1994) and is carcinogenic on long-term expo- sure as well as being one of the most potent promoting agents known for experimental liver and skin cancer (Chapter 7; Lucier et al., 1993). In humans, the carcinogenic potential of TCDD has been confused by exposures that are almost always in association with multiple other chemicals and confounding effects such as smoking. However, the IARC has classified TCDD and phe- noxy herbicides as carcinogenic to humans on the basis of limited evidence from epidemiologi- cal studies and sufficient evidence from animal studies (McGregor et al., 1998). An analysis of four Swedish studies was one of the bases for this decision, although other earlier investigations (Kang et al., 1987; Riihimäki et al., 1982) did not support such an association. Somewhat less clear is the association of non-Hodgkin lymphoma with herbicide and TCDD exposure, al- though the increasing incidence of this disease in a number of countries could be based, at least in part, on such chronic exposure (Zahm and Blair, 1992; Hardell et al., 1994). Still, because of the relative lack of genotoxic effects of TCDD in general and its extremely potent promoting action in experimental animals, its carcinogenic effects may be more closely related to this latter characteristic.
Unlike TCDD, the 2-carbon halogenated compounds trichloroethylene and vinyl chloride are produced and used extensively worldwide for a variety of occupational procedures. About a quarter of a million tons of trichloroethylene are produced in the world each year and used as a solvent in numerous industrial processes. The IARC (1995) has concluded that there is limited evidence in humans for the carcinogenicity of this compound, but a committee of the National Research Council in the United States (James et al., 1996) felt that the evidence both in humans and animals was less convincing, although they did choose to calculate a cancer risk based on estimates by the U.S. Environmental Protection Agency. A much less controversial subject is the association of malignant neoplasms of the liver with exposure to vinyl chloride, the monomer used in the production of a variety of plastics. A review by the IARC in 1975 reported on 43 cases of angiosarcoma found in 10 different countries throughout the world, all of the patients having a history of working with vinyl chloride. Hepatic angiosarcoma is an extremely rare neo- plasm in humans, and thus the incidence seen in this group is far out of proportion to what might be expected in the general population. Cooper (1981) reviewed a population of 10,173 men whose jobs involved a probable exposure to vinyl chloride prior to January 1, 1973. In this
Figure 11.10 Structural chromosome alterations in chromosomes 8 and 21 in lymphocytes from blood of workers exposed to different levels of benzene in comparison with unexposed controls. The p values are significant for exposures at greater than 31 ppm. (Adapted from Smith et al., 1998, with permission of the authors and publishers.)
group, eight angiosarcomas were found, as well as a twofold increase in malignant neoplasms of the brain and other parts of the nervous system. Although the low incidence of these neoplasms in this large population suggests that vinyl chloride is a relatively weak carcinogen, the rarity of hepatic angiosarcoma in the general population strongly supports the causal relationship be- tween the organic monomer and the induction of this mesenchymal neoplasm. In addition, vinyl chloride exposure induces other, somewhat uncommon pathologies in the liver, including exten- sive fibrosis, which may lead to portal hypertension (increased pressure within the portal venous system), splenomegaly, and hypertrophy and hyperplasia of both hepatocytes and hepatic and splenic mesenchymal cells (Thomas et al., 1975). Vinyl chloride is rapidly absorbed after respi- ratory exposure and is metabolized in the liver by the cytochrome P450 2E1 system to electro- philic metabolites, particularly chloroethyleneoxide and chloroacetaldehyde. These electrophiles react with DNA bases to form adducts, of which the etheno-base compounds (Fig- ure 3.9) are highly persistent, with half-lives of more than 30 days (cf. Marion et al., 1996). Mutations in the K-ras and p53 genes occur in the livers of exposed individuals. The etheno
adducts cause transition and transversion mutations in bacterial systems (Cheng et al., 1991; Marion et al., 1996).
The United States government has sponsored publications suggesting that occupational causes of neoplasia in humans may be responsible for as much as 20% of total cancer mortality, but Doll and Peto (1981) have presented compelling arguments that such is not the case. Their review of the presently available statistical and epidemiological evidence indicates that only about 4% of all cancer deaths in the United States can be attributed to occupational causes. Fur- thermore, the strict governmental regulation of actual and potential industrial health hazards promises to reduce this figure even further in the future (Chapter 13).