In 1960 more than 100,000 young turkeys died in England. Studies showed that the death of the birds was due to a contaminant present in the peanut meal, obtained from South Africa, in their diet. Comparable problems arose in eastern Africa at the same time. Careful chemical studies demonstrated that the offending agent belonged to a group of toxic metabolites termed the afla- toxins, which are produced by some strains of the ubiquitous mold Aspergillus flavus. Shortly thereafter, the carcinogenicity of purified aflatoxin B1 was demonstrated in rodents, the most striking characteristic being its extremely high potency. Fifteen parts per billion of aflatoxin B1 in the diet for approximately 1¹⁄₂ years produced hepatomas in all treated rats. In trout, one part or less per billion was effective as a carcinogen (Newberne and Butler, 1969).
Epidemiological studies have shown that the geographical regions in which there is exten- sive contamination of foodstuffs by aflatoxin are also the areas where the incidence of human liver cancer is relatively high. However, for the most part in these same areas, there is substantial prevalence of hepatitis B and/or C infections. Thus, at least one study has suggested that aflatoxin exposure in these areas has relatively little influence on the primarily virally induced hepatic neoplasms (Campbell et al., 1990). On the other hand, several studies have shown a direct association between the aflatoxin content of food and the frequency of hepatomas in the popula- tion. In parts of Africa and East Asia where aflatoxin is found in the diet, levels of 100 to 1000 parts per billion in individual foodstuffs are not unusual. Thus, unlike most other environmental carcinogenic agents for the human, the dose of aflatoxin to which the human is exposed greatly exceeds that known to produce cancer in experimental animals. As yet, however, because of the major role played by hepatitis B and C in the induction of human hepatic neoplasia (see below), the role of aflatoxin exposure in the causation of human hepatic neoplasia is somewhat uncertain.
Further evidence for the role of aflatoxin in inducing hepatic neoplasia comes from two sources. Several “markers” (Chapter 17) have now been developed and demonstrated to be di- rectly related to aflatoxin exposure. In one study (Ross et al., 1992), a large number of Chinese patients were followed for several years. In those individuals exhibiting evidence both of hepati-tis B virus infection and aflatoxin exposure as measured by a urinary marker, the aflatoxin-N7-guanine adduct, a dramatic increase in the risk of developing liver cancer compared with the presence of either factor alone was noted. In addition, as we have noted in Chapter 6 (Figure 6.2), aflatoxin exposure has been related to a specific mutation at codon 249 in the p53 tumor suppressor gene (Hsu et al., 1991; Bressac et al., 1991). Further extension of these studies has demonstrated that patients from areas with high aflatoxin levels are more likely to exhibit the codon 249 mutation and other p53 mutations, especially transversion, than are patients from ar- eas with low aflatoxin exposure (Lasky and Magder, 1997). However, other studies in nonhuman primates, as in the rat (Chapter 6), did not exhibit codon 249 mutations in p53 or, for the most part, in any other part of the gene investigated when animals were exposed to aflatoxin B1 (Fujimoto et al., 1992). Furthermore, later investigations have suggested that many p53 gene mutations, which may be common in advanced hepatocellular carcinomas, occur as a late event during the stage of progression (Hsu et al., 1994). Thus, it appears that while some codon muta- tions in the p53 gene in hepatocytes may occur as an initiating event in hepatocellular carcinoma in the human, most mutations in this gene may be the result of selection during the stage of progression (Dragan and Pitot, 1994). The most likely scenario, however, is that high aflatoxin exposure to the human on a chronic basis may act as both an initiating and promoting stimulus which, in the presence of a viral infection involving clastogenic events (Hino et al., 1991), ulti- mately leads to the development of hepatocellular carcinoma. Although other mold toxins, such as sterigmatocystin (Gopalakrishnan et al., 1992), have been suggested as additional etiological factors in human liver cancer, their role in this disease has not been proven at the present time.
Other natural products—such as the pyrrolizidine alkaloids (McLean, 1970), which occur in extracts of various roots and leaves found in various parts of the world; cycasin, obtained from extracts of the cycad nut (Hoffmann and Morgan, 1984); and safrole (Borchert et al., 1973), a naturally occurring flavoring agent—have been shown to be carcinogenic for the liver in rodents. As yet there is no significant evidence that these materials cause hepatomas in humans, although the pyrrolizidine alkaloids have been implicated in the production of vascular disease of the liver in Jamaica and other West Indian countries (McLean, 1970). On the other hand, some natural products, as noted in Chapter 8 (Table 8.8), are inhibitory to the process of carcinogenesis, includ- ing even some naturally occurring mutagens (Stavric, 1984). A few edible plants have been found to contain or produce agents capable of inducing cancer in one or more species. Bracken fern is both a food delicacy and salad green in certain parts of the world. When fed to rats, this agent is a carcinogen for the bladder and intestine (Pamukcu et al., 1976). In areas of the world where cattle graze on bracken fern–containing pastures, these animals develop urinary bladder tumors.
An example of the induction of cancer in the human by naturally occurring dietary con- taminants is described in several provinces of the People’s Republic of China. In these locales, there is a positive correlation between the extremely high incidence of esophageal carcinoma and the consumption of pickled and otherwise moldy foodstuffs that contain carcinogenic nitro- samines (Singer et al., 1986; Li et al., 1986). Although the exact chemical structures of almost all of these dietary contaminants are unknown, at least one N-nitroso compound has been iso- lated from corn bread inoculated with fungi (Shixin et al., 1979) and from extracts of pickled vegetables. Moldy cornmeal fed to rats induces carcinoma and epithelial dysplasia of the fore- stomach. Since this latter structure is quite similar anatomically to the esophagus in the human, there is reason to argue for a causal relation of these contaminated foodstuffs to the high inci- dence of esophageal cancer in areas where such dietary contamination occurs.
Perhaps the most ubiquitous group of carcinogenic agents occurring “naturally” in the diet, primarily in cooked foods, are the heterocyclic amines, which have been described and studied by a number of investigators (cf. Sugimura et al., 1996). Figure 11.2, adapted from these authors (Sugimura et al., 1996), shows the structures of a number of mutagenic and carcinogenic hetero- cyclic amines that occur in heated foods. These substances apparently result from the pyrrolysis of amino acids and proteins. The chemistry of these reactions has been investigated and shown to occur in both crude and chemically characterized systems (Vuolo and Schuessler, 1985). The amount of these materials present in cooked food varies with the cooking method—i.e., broiling, frying, barbecuing, or microwaving—and the temperature of cooking above 150°C (Robbana- Barnat et al., 1996). Thus, one might expect that individuals consuming diets prepared in various ways will consume different levels of these carcinogenic amines; such a difference has been re- ported, for example, between Americans and Japanese (Nagao et al., 1996) as well as in specifi- cally and carefully studied cohorts (Augustsson et al., 1997). Cancer potency estimates for the human have been placed as high as 1 in 103 or 104 for an average lifetime of cooked beef intake of approximately 0.5 lb/day (Bogen, 1994). On the other hand, Gold et al. (1994), in a very ex- tensive study, have suggested that even the most potent of the heterocyclic amines offers a rela- tively low risk in the average American diet compared with other, less potent materials but which are consumed at much higher levels (e.g., ethyl alcohol). Epidemiological evidence suggests both a relationship between the methods of cooking meats and the daily consumption of meat in relation to the incidence of various cancers, especially colon cancer (cf. Felton et al., 1997).
A major area of concern in modern food technology is the addition of chemical agents to processed foods for flavor, coloring, and preservation. Also, additions to the human diet of po- tential carcinogens at low levels have been made inadvertently through the contamination of food sources by pesticides and industrial wastes. Examples of compounds that have been shown to be carcinogenic in experimental animals and that have entered the human diet are certain pesticides, DDT, and dieldrin as well as the industrial contaminant polychlorinated biphenyls. Specific addi- tions to the human diet of agents that have been found to be carcinogenic at very high levels in experimental animals include the synthetic antioxidant butylated hydroxytoluene (BHT), the col- oring agent red dye 2, and the artificial sweetener saccharin. Of these, saccharin created a na- tional controversy in relation to the legal constraints of the Delaney amendment (Chapter 13) and the wishes of Congress and the public to keep this material on the market. Sodium nitrite has been used as a preservative in a wide variety of prepared meat products for many years. Although nitrites themselves are not carcinogenic, they have been shown in experimental animals to react with secondary amines both in the food and in the intestine to produce carcinogenic nitrosamines (Leaf et al., 1989). The demonstration of low levels of dimethylnitrosamine in fish meal treated with nitrite and the presence in flour of trace amounts of diethylnitrosamine, produced from dry- ing the grain in a stream of exhaust gases containing oxides of nitrogen, are ample evidence that such compounds can be produced outside the living organism (cf. Gangolli et al., 1994). Further- more, some experimental studies have demonstrated the production of a number of different ma- lignancies of the gastrointestinal and respiratory tracts by the addition of nitrite and secondary or tertiary amines to the diet (Matsukura et al., 1977; Yoshida et al., 1993).
Studies have demonstrated that the major amount of nitrite in humans originates from en- dogenous sources, largely through the reduction of nitrates by oral bacteria ingested as compo- nents of vegetables and plant foodstuffs (Tannenbaum et al., 1978). The contribution of
Figure 11.2 Structures of some mutagenic and carcinogenic heterocyclic amines. Abbreviations: Trp-P-1, 3-amino-1,4-dimethyl-5H-pyrido[4,3-b]indole; Trp-P-2, 3-amino-1-methyl-5H-pyrido[4,3-b]indole; Glu-P-1, 2-amino-6-methyldipyrido[1,2-a:3′,2′-d]imidazole; Glu-P-2, 2-aminodipyrido[1,2-a:3′,2′- d]imidazole; Phe-P-1, 2-amino-5-phenylpyridine; Orn-P-1, 4-amino-6-methyl-1H-2,5,10,10b-tetraazafluoranthene; AαC, 2-amino-9H-pyrido[2,3-b]indole; MeAαC, 2-amino-3-methyl-9H-pyrido[2,3-b]indole; IQ, 2-amino-3-methylimidazo[4,5-f]quinoline; MelQ, 2-amino-3,4,-dimethylimidazo[4,5-f]quinoline; IQx, 2-amino-3-methylimidazo[4,5-f]quinoxaline; MelQx, 2-amino-3,8-dimethylimidazo[4,5-f]-quinoxaline; 4,8-DiMelQx, 2-amino-3,4,8-trimethylimi- dazo[4,5-f]quinoxaline; 7,8-DiMelQx, 2-amino-3,7,8-trimethylimidazo[4,5-f]quinoxaline; PhIP, 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine; 4′-OH- PhIP, 2-amino-1-methyl-6-(4-hydroxyphenyl)imidazo[4,5-b]pyridine; Cre-P-1, 4-amino-1,6-dimethyl-2-methylamino-1H,6H-pyrrolo[3,4-f]benzimidazole-
5,7-dione; 4-CH2OH-8-MelQx, 2-amino-4-hydroxymethyl-3,8-dimethylimidazo[4,5-f]quinoxaline; 7,9-DiMelgQx, 2-amino-1,7,9-trimethylimidazo[4,5- g]quinoxaline. The following ten compounds have been proven to be carcinogenic: Trp-P-1, Trp-P-2, Glu-P-1, Glu-P-2, AαC, MeAαC, IQ, MelQ, MelQx, PhIP. (From Sugimura et al., 1996, with permission of the authors and publishers.)
exogenous nitrite to the total internal nitrite pool is only about 10%, thus raising significant questions as to what role if any nitrite preservatives and coloring agents have in the genesis of human cancer. Hartman has reported that the daily levels of nitrate ingestion in the 1970s showed a strong positive correlation with gastric cancer mortality in at least 12 countries (Hartman, 1983), and human exposure is still quite extensive (Bartsch et al., 1989). However, Pobel et al. (1995). in a case-controlled study in France, were unable to demonstrate any associ- ation between the intake of nitrate and nitrite and the increased risk of stomach cancer.
A few edible plants have been found to contain or produce agents capable of inducing cancer in one or more species. Bracken fern is both a food delicacy and salad green in certain parts of the world. When fed to rats, this agent is a carcinogen for the bladder and intestine (Pa- mukcu et al., 1976). In areas of the world where cattle graze on bracken fern–containing pas- tures, these animals develop urinary bladder tumors.