Ranking Possible Carcinogenic Hazards BRUCE N. AMES,* RENAE MAGAW, Lois SWIRSKY GOLD doses to effects on humans exposed to low doses is routinely This review discusses reasons why animal cancer tests attempted by regulatory agencies when formulating policies at- cannot be used to predict absolute human risks. Such tempting to prevent future cancer. There is little sound scientific tests, however, may be used to indicate that some chemi- basis for this type of extrapolation, in part due to our lack of cals might be of greater concern than others. Possible knowledge about mechanisms of cancer induction, and it is viewed hazards to humans from a variety of rodent carcinogens with great unease by many epidemiologists and toxicologists (5, 9- are ranked by an index that relates the potency of each 11). Nevertheless, to be prudent in regulatory policy, and in the carcinogen in rodents to the exposure in humans. This absence of good human data (almost always the case), some reliance ranking suggests that carcinogenic hazards from current on animal cancer tests is unavoidable. The best use of them should levels of pesticide residues or water pollution are likely to Downloaded from www.sciencemag.org on March 3, 2010 be made even though few, if any, of the main avoidable causes of be ofminimal concern relative to the background levels of human cancer have typically been the types of man-made chemicals natural substances, though one cannot say whether these that are being tested in animals (10). Human cancer may, in part, natural exposures are likely to be of major or minor involve agents such as hepatitis B virus, which causes chronic importance. inflammation; changes in hormonal status; deficiencies in normal protective factors (such as selenium or p-carotene) against endoge- nous carcinogens (12); lack of other anticarcinogens (such as dietary fiber or calcium) (4); or dietary imbalances such as excess consump- E PIDEMIOLOGISTS ESTIMATE THAT AT LEAST 70% OF HUMAN tion of fat (3, 4, 12) or salt (13). cancer would, in principle, be preventable if the main risk There is a need for more balance in animal cancer testing to and antirisk factors could be identified (1). This is because emphasize the foregoing factors and natural chemicals as well as the incidence of specific types of cancer differs markedly in different synthetic chemicals (12). There is increasing evidence that our parts of the world where people have different life-styles. For normal diet contains many rodent carcinogens, all perfectly natural example, colon and breast cancer, which are among the major types or traditional (for example, from the cooking of food) (12), and that of cancer in the United States, are quite rare among Japanese in no human diet can be entirely free of mutagens or agents that can be Japan, but not among Japanese-Americans. Epidemiologists are carcinogenic in rodent systems. We need to identify the important providing important clues about the specific causes of human causes of human cancer among the vast number of minimal risks. cancer, despite inherent methodological difficulties. They have This requires knowledge of both the amounts of a substance to identified tobacco as an avoidable cause of about 30% of all U.S. which humans are exposed and its carcinogenic potency. cancer deaths and of an even larger number of deaths from other Animal cancer tests can be analyzed quantitatively to give an causes (1, 2). Less specifically, dietary factors, or their absence, have estimate of the relative carcinogenic potencies of the chemicals been suggested in many studies to contribute to a substantial tested. We have previously published our Carcinogenic Potency proportion of cancer deaths, though the intertwined risk and Database, which showed that rodent carcinogens vary in potency by antirisk factors are being identified only slowly (1, 3, 4). High fat more than 10 millionfold (14). intake may be a major contributor to colon cancer, though the This article attempts to achieve some perspective on the plethora evidence is not as definitive as that for the role of saturated fat in of possible hazards to humans from exposure to known rodent heart disease or of tobacco in lung cancer. Alcoholic beverage carcinogens by establishing a scale of the possible hazards for the consumption, particularly by smokers, has been estimated to con- amounts ofvarious common carcinogens to which humans might be tribute to about 3% of U.S. cancer deaths (1) and to an even larger chronically exposed. We view the value of our calculations not as number of deaths from other causes. Progress in prevention has providing a basis for absolute human risk assessment, but as a guide been made for some occupational factors, such as asbestos, to which to priority setting. One problem with this type of analysis is that few workers used to be heavily exposed, with delayed effects that still of the many natural chemicals we are exposed to in very large contribute to about 2% of U.S. cancer deaths (1, 5). Prevention may amounts (relative to synthetic chemicals) have been tested in animals also become possible for hormone-related cancers such as breast for carcinogenicity. Thus, our knowledge of the background levels cancer (1, 6), or virus-related cancers such as liver cancer (hepatitis of human exposure to animal carcinogens is fragmentary, biased in B) and cancer of the cervix (papilloma virus HPV16) (1, 7). favor of synthetic chemicals, and limited by our lack of knowledge of Animal bioassays and in vitro studies are also providing clues as to human exposures. which carcinogens and mutagens might be contributing to human cancer. However, the evaluation of carcinogenicity in rodents is expensive and the extrapolation to humans is difficult (8-11). We B. N. Ames is associated with the Department of Biochemistry, University of will use the term "possible hazard" for estimates based on rodent California, Berkeley, CA 94720. R. Magaw and L. Swirsky Gold are associated with the Biology and Medicine Division, Lawrence Berkeley Laboratory, Berkeley, CA 94720. cancer tests and "risk" for those based on human cancer data (10). Extrapolation from the results of rodent cancer tests done at high *To whom reprint requests should be sent. 17 APRIL I987 ARTICLES 271 Ranking of Possible Carcinogenic Hazards Pesticide residues. Intake of man-made pesticide residues from food in the United States, including residues of industrial chemicals such Since carcinogens differ enormously in potency, a comparison of as polychlorinated biphenyls (PCBs), averages about 150 ,ug/day. possible hazards from various carcinogens ingested by humans must Most (105 ,ug) of this intake is composed of three chemicals take this into account. The measure of potency that we have (ethylhexyl diphenyl phosphate, malathion, and chlorpropham) developed, the TD50, is the daily dose rate (in milligrams per shown to be noncarcinogenic in tests in rodents (16). A carcinogen- kilogram) to halve the percent of tumor-free animals by the end of a ic pesticide residue in food of possible concem is DDE, the principal standard lifetime (14). Since the TD50 (analogous to the LD50) is a metabolite (>90%) of DDT (16). The average U.S. daily intake of dose rate, the lower the TD50 value the more potent the carcinogen. DDE from DDT (HERP = 0.0003%) is equivalent to the HERP To calculate our index of possible hazard we express each human of the chloroform in one glass of tap water and thus appears to be exposure (daily lifetime dose in milligrams per kilogram) as a insignificant compared to the background of natural carcinogens in percentage of the rodent TD5o dose (in milligrams per kilogram) for our diet (Table 1). Even daily consumption of 100 times the average each carcinogen. We call this percentage HERP [Human Exposure intake of DDE/DDT or PCBs would produce a possible hazard that dose/Rodent Potency dose]. The TD50 values are taken from our is small compared to other common exposures shown in Table 1. ongoing Carcinogenic Potency Database (currently 3500 experi- Nature's pesticides. We are ingesting in our diet at least 10,000 ments on 975 chemicals), which reports the TD50 values estimated times more by weight of natural pesticides than of man-made from experiments in animals (14). Human exposures have been pesticide residues (12). These are natural 'toxic chemicals" that have estimated from the literature as indicated. As rodent data are all an enormous variety of chemical structures, appear to be present in calculated on the basis of lifetime exposure at the indicated daily all plants, and serve to protect plants against fungi, insects, and dose rate (14), the human exposure data are similarly expressed as animal predators (12). Though only a few are present in each plant lifelong daily dose rates even though the human exposure is likely to species, they commonly make up 5 to 10% of the plant's dry weight be less than daily for a lifetime. (12). There has been relatively little interest in the toxicology or Downloaded from www.sciencemag.org on March 3, 2010 It would be a mistake to use our HERP index as a direct estimate carcinogenicity of these compounds until quite recently, although of human hazard. First, at low dose rates human susceptibility may they are by far the main source of "toxic chemicals" ingested by differ systematically from rodent susceptibility. Second, the general humans. Only a few dozen of the thousands present in the human shape of the dose-response relationship is not known. A linear dose diet have been tested in animal bioassays, and only some of these response has been the dominant assumption in regulating carcino- tests are adequate for estimating potency in rodents (14). A sizable gens for many years, but this may not be correct. If the dose proportion of those that have been tested are carcinogens, and many responses are not linear but are actually quadratic or hockey-stick others have been shown to be mutagens (12), so it is probable that shaped or show a threshold, then the actual hazard at low dose rates many more will be found to be carcinogens if tested. Those shown might be much less than the HERP values would suggest. An in Table 1 are: estragole (HERP = 0.1% for a daily 1 g of dried additional difficulty is that it may be necessary to deal with basil), safrole (HERP = 0.2% for a daily natural root beer), sym- carcinogens that differ in their mechanisms of action and thus in phytine (a pyrrolizidine alkaloid, 0.03% for a daily cup of comfrey their dose-response relationship. We have therefore put an asterisk tea), comfrey tablets sold in health food stores (6.2% for a daily next to HERP values for carcinogens that do not appear to be active dose), hydrazines in mushrooms (0.1% for one daily raw mush- through a genotoxic (DNA damaging or mutagenic) mechanism room), and allyl isothiocyanate (0.07% for a daily 5 g of brown (15) so that comparisons can be made within the genotoxic or mustard). nongenotoxic classes. Plants commonly produce very much larger amounts of their Table 1 presents our HERP calculations of possible cancer natural toxins when damaged by insects or fungi (12). For example, hazards in order to compare them within several categories so that, psoralens, light-activated carcinogens in celery, increase 100-fold for example, pollutants of possible concem can be compared to when the plants are damaged by mold and, in fact, can cause an natural carcinogens in the diet. A convenient reference point is the occupational disease in celery-pickers and in produce-checkers at possible hazard from the carcinogen chloroform in a liter of average supermarkets (12, 17). (U.S.) chlorinated tap water, which is close to a HERP of 0.001%. Molds synthesize a wide variety of toxins, apparently as antibiotics Chloroform is a by-product of water chlorination, which protects us in the microbiological struggle for survival: over 300 mycotoxins from pathogenic viruses and bacteria. have been described (18). They are common pollutants of human Contaminated water. The possible hazards from carcinogens in food, particularly in the tropics. A considerable percentage of those contaminated well water [for example, Santa Clara ("Silicon") tested have been shown to be mutagens and carcinogens: some, such Valley, California, or Woburn, Massachusetts] should be compared as aflatoxin and sterigmatocystin, are among the most potent known to the possible hazard of ordinary tap water (Table 1). Of 35 wells rodent carcinogens. The potency of aflatoxin in different species shut down in Santa Clara Valley because of their supposed carcino- varies widely; thus, a bias may exist as the HERP uses the most genic hazard, only two have HERP values greater than ordinary tap sensitive species. The aflatoxin content of U.S. peanut butter water. Well water is not usually chlorinated and typically lacks the averages 2 ppb, which corresponds to a HERP of 0.03% for the chloroform present in chlorinated tap water. Water from the most peanut butter in an average sandwich (Table 1). The Food and Drug polluted well (HERP = 0.004% per liter for trichloroethylene), as Administration (FDA) allows ten times this level (HERP = 0.3%), indicated in Table 1, has a HERP value orders of magnitude less and certain foods can often exceed the allowable limit (18). Afla- than for the carcinogens in an equal volume of cola, beer, or wine. toxin contaminates wheat, corn (perhaps the main source of dietary Its HERP value is also much lower than that of many of the aflatoxin in the United States), and nuts, as well as a wide variety of common natural foods that are listed in Table 1, such as the average stored carbohydrate foodstuffs. A carcinogenic, though less potent, peanut butter sandwich. Caveats for any comparisons are given metabolite of aflatoxin is found in milk from cows that eat moldy below. Since the consumption of tap water is only about 1 or 2 liters grain. per day, the animal evidence provides no good reason to expect that There is epidemiologic evidence that aflatoxin is a human carcino- chlorination of water or current levels of man-made pollution of gen. High intake in the tropics is associated with a high rate of liver water pose a significant carcinogenic hazard. cancer, at least among those chronically infected with the hepatitis B 272 SCIENCE, VOL. 236 virus (19, 20). Considering the potency of those mold toxins that modem techniques of agriculture and storage, including use of have been tested and the widespread contamination of food with synthetic pesticides and fiunigants. molds, they may represent the most significant carcinogenic pollu- Preparation ofjfods and beverages can also produce carcinogens. tion of the food supply in developing countries. Such pollution is Alcohol has been shown to be a human carcinogen in numerous much less severe in industrialized countries, due to refrigeration and epidemiologic studies (1, 21). Both alcohol and acetaldehyde, its Table 1. Ranking possible carcinogenic hazards. Potenc ofcarcinogens: A number in parentheses indicates a TDm value not used in HERP calculation because it is the less sensitive species; (-) = negative in cancer test. (+) = positive for carcinogenicity in test(s) not suitable for calculating a TD50; (?) = is not adequately tested for carcinogenicity. TDIo values shown are averages calculated by taking the harmonic mean of the TD0's of the positive tests in that species from the Carcinogenic Potency Database. Results are similar if the lowest TDI, value (most potent) is used instead. For each test the target site with the low- est TD50 value has been used. The average TD50 has been calculated separately for rats and mice, and the more sensitive species is used for calculating the pos- sible hazard. The database, with references to the source of the cancer tests, is complete for tests published through 1984 and for the National Toxicology Program bioassays through June 1986 (14). We have not indicated the route of exposure or target sites or other particulars of each test, although these are re- ported in the database. Daiy human epoure: We have tried to use average or reasonable daily intakes to facilitate comparisons. In several cases, such as contaminated well water or factory exposure to EDB, this is difficult to determine, and we give the value for the worst found and indicate pertinent information in the References and Notes. The calculations assume a daily dose for a lifetime; where drugs are normally taken for only a short period we have bracketed the HERP value. For inhalation exposures we assume an inhalation of9,600 liters per 8 hours for the workplace and 10,800 liters per 14 hours for indoor air at home. Possibk hazrd: The amount of rodent carcinogen indicated under carcinogen dose is divided by 70 kg to give a milligram per kilogram of human exposure, and this human dose is given as the percentage of the TD50 dose in the rodent (in milligrams per kilogram) to calculate the Human Exposure/Rodent Potency index (HERP). Possible Potency of carcinogen: hazard: Daily human Carcinogen dose per TD50 (mg/kg) Refer- HERP (%) exposure 70-kgperson Rats Mce ences Downloaded from www.sciencemag.org on March 3, 2010 IEnvironmental pollution 0.001* Tap water, 1 liter Chloroform, 83 p.g (U.S. average) (119) 90 96 0.004* Well water, 1 liter contaminated Trichloroethylene, 2800 ,g (-) 941 97 (worst well in Silicon Valley) 0.0004* Well water, 1 liter contaminated, Wobum Trichloroethylene, 267 ,ug (-) 941 98 0.0002* Chloroform, 12 ,ug (119) 90 0.0003* Tetrachloroethylene, 21 Fg ' 101 (126) 0.008* Swimming pool, 1 hour (for child) Chloroform, 250 ,±g (average pool) (119) 90 99 0.6 Conventional home air (14 hour/day) Formaldehyde, 598 pg 1.5 (44) 100 0.004 Benzene, 155 ,ug (157) 53 2.1 Mobile home air (14 hour/day) Formaldehyde, 2.2 mg 1.5 (44) 28 Pe'atcide and other residues 0.0002* PCBs: daily dietary intake PCBs, 0.2 ,ug (U.S. average) 1.7 (9.6) 101 0.0003* DDE/DDT: daily dietary intake DDE, 2.2 itg (U.S. average) (-) 13 16 0.0004 EDB: daily dietary intake Ethylene dibromide, 0.42 ,ug 1.5 (5.1) 102 (from grains and grain products) (U.S. average) Naturi appetiides and dieay taxins 0.003 Bacon, cooked (100 g) Dimethylnitrosamine, 0.3 Fg (0.2) 0.2 40 0.006 Diethylnitrosamine, 0.1 1Lg 0.02 (+) 0.003 Sake (250 ml) Urethane, 43 j.g (41) 22 24 0.03 Comfrey herb tea, 1 cup Symphytine, 38 ,g 1.9 (?) 103 (750 ,ug of pyrrolizidine alkaloids) 0.03 Peanut butter (32 g; one sandwich) Aflatoxin, 64 ng (U.S. average, 2 ppb) 0.003 (+) 18 0.06 Dried squid, broiled in gas oven (54 g) Dimethylnitrosamine, 7.9 ,ug (0.2) 0.2 37 0.07 Brown mustard (5 g) Allyl isothiocyanate, 4.6 mg 96 (-) 47 0.1 Basil (1 g of dried leaf) Estragole, 3.8 mg (?) 52 48 0.1 Mushroom, one raw (15 g) (Agaricus bispors) Mixture of hydrazines, and so forth (?) 20,300 104 0.2 Natural root beer (12 ounces; 354 ml) Safrole, 6.6 mg (436) 56 105 (now banned) 0.008 Beer, before 1979 (12 ounces; 354 ml) Dimethylnitrosamine, 1 ,g (0.2) 0.2 38 2.8* Beer (12 ounces; 354 ml) Ethyl alcohol, 18 ml 9110 (?) 23 4.7* Wine (250 ml) Ethyl alcohol, 30 ml 9110 (?) 23 6.2 Comfrcy-pepsin tablets (nine daily) Comfrey root, 2700 mg 626 (?) 103 1.3 Comfrey-pepsin tablets (nine daily) Symphytine, 1.8 mg 1.9 (?) Food additive 0.0002 AF-2: daily dietary intake before banning AF-2 (fiuylfiuramide), 4.8 p.g 29 (131) 44 0.06* Diet Cola (12 ounces; 354 ml) Saccharin, 95 mg 2143 (-) 106 Dw7ffls [0.3] Phenacetin pill (average dose) Phenacetin, 300 mg 1246 (2137) 51 [5.6] Metronidazole (therapeutic dose) Metronidazole, 2000 mg (542) 406 107 [14] Isoniazid pill (prophylactic dose) Isoniazid, 300 mg (150) 30 108 16* Phenobarbital, one sleeping pill Phenobarbital, 60 mg (+) 5.5 50 17* Clofibrate (average daily dose) Clofibrate, 2000 mg 169 (?) 52 Ocupaional expoure 5.8 Formaldehyde: Workers' average daily intake Formaldehyde, 6.1 mg 1.5 (44) 109 140 EDB: Workers' daily intake (high exposure) Ethylene dibromide, 150 mg 1.5 (5.1) 55 *Astesks indicate HERP from carcinogens tiought to be nongenotoxic. 17 APRIUL 1987 ARTICLES 273 major metabolite, are carcinogens in rats (22, 23). The carcinogenic not clear how significant a risk these pose. Nitrosamines were potency of ethyl alcohol in rats is remarkably low (23), and it is ubiquitous in beer and ale (HERP = 0.008%) and were formed among the weakest carcinogens in our database. However, human from NO2 in the gas flame-heated air used to dry the malt. intake of alcohol is very high (about 18 g per beer), so that the However, the industry has switched to indirect heating, which possible hazards shown in Table 1 for beer and wine are large resulted in markedly lower levels (<1 ppb) of dimethylnitrosamine (HERP = 2.8% for a daily beer). The possible hazard of alcohol is (38). The dimethylnitrosamine found in human urine is thought to enormous relative to that from the intake of synthetic chemical be formed in part from NO2 inhaled from kitchen air (39). Cooked residues. If alcohol (20), trichloroethylene, DDT, and other pre- bacon contains several nitrosamines (HERP = 0.009%) (40). sumptive nongenotoxic carcinogens are active at high doses because Oxidation offats and vegetable oils occurs during cooking and also they are tumor promoters, the risk from low doses may be minimal. spontaneously if antioxidant levels are low. The result is the Other carcinogens are present in beverages and prepared foods. formation of peroxides, epoxides, and aldehydes, all ofwhich appear Urethane (ethyl carbamate), a particularly well-studied rodent car- to be rodent carcinogens (8, 12, 27). Fatty acid hydroperoxides cinogen, is formed from ethyl alcohol and carbamyl phosphate (present in oxidized oils) and cholesterol epoxide have been shown during a variety of fermentations and is present in Japanese sake to be rodent carcinogens (though not in tests suitable for calculating (HERP = 0.003%), many types of wine and beer, and in smaller a TD5o). Dried eggs contain about 25 ppm ofcholesterol epoxide (a amounts in yogurt and bread (24). Another fermentation product, sizable amount), a result of the oxidation of cholesterol by the NO2 the dicarbonyl aldehyde methylglyoxal, is a potent mutagen and was in the drying air that is warmed by gas flames (12). isolated as the main mutagen in coffee (about 250 ,ug in one cup). It Normal oxidation reactions in fruit (such as browning in a cut was recently shown to be a carcinogen, though not in a test suitable apple) also involve production of peroxides. Hydrogen peroxide is a for calculating a TD50 (25). Methylglyoxal is also present in a variety mutagenic rodent carcinogen that is generated by oxidation of of other foods, such as tomato puree (25, 26). Diacetyl (2,3- natural phenolic compounds that are quite widespread in edible butanedione), a closely related dicarbonyl compound, is a fermenta- plants. A cup of coffee contains about 750 jxg of hydrogen peroxide Downloaded from www.sciencemag.org on March 3, 2010 tion product in wine and a number of other foods and is responsible (25); however, since hydrogen peroxide is a very weak carcinogen for the aroma of butter. Diacetyl is a mutagen (27) but has not been (similar in potency to alcohol), the HERP for drinking a daily cup tested for carcinogenicity. of coffee would be very low [comparable to DDE/DDT, PCBs, or Formaldehyde, another natural carcinogenic and mutagenic alde- ethylene dibromide (EDB) dietary intakes]. Hydrogen peroxide is hyde, is also present in many common foods (22, 26-28). Formalde- also generated in our normal metabolism; human blood contains hyde gas caused cancer only in the nasal turbinates of the nose- about 5 pM hydrogen peroxide and 0.3 itM of the cholesterol ester breathing rodents and even though formaldehyde is genotoxic, the of fatty acid hydroperoxide (41). Endogenous oxidants such as dose response was nonlinear (28, 29). Hexamethylenetetramine, hydrogen peroxide may make a major contribution to cancer and which decomposes to formaldehyde in the stomach, was negative in aging (42). feeding studies (30). The effects of oral versus inhalation exposure Calork intake, which could be considered the most striking rodent for formaldehyde remain to be evaluated more thoroughly. carcinogen ever discovered, is discussed remarkably little in relation As formaldehyde is almost ubiquitous in foods, one can visualize to human cancer. It has been known for about 40 years that various formaldehyde-rich scenarios. Daily consumption of shrimp increasing the food intake in rats and mice by about 20% above (HERP = 0.09% per 100 g) (31), a sandwich (HERP of two slices optimal causes a remarkable decrease in longevity and a striking of bread = 0.4%) (22), a cola (HERP = 2.7%) (32), and a beer increase in endocrine and mammary tumors (43). In humans, (HERP = 0.2%) (32) in various combinations could provide as obesity (associated with high caloric intake) leads to increased levels much formaldehyde as living in some mobile homes of circulating estrogens, a significant cause of endometrial and gall (HERP = 2.1%; Table 1). Formaldehyde is also generated in bladder cancer. The effects of moderate obesity on other types of animals metabolically, for example, from methoxy compounds that human cancer are less clear (1). humans ingest in considerable amounts from plants. The level of Food additives are currently screened for carcinogenicity before use formaldehyde reported in normal human blood is strikingly high if they are synthetic compounds. AF-2 (HERP = 0.0002%), a (about 100 pM or 3000 ppb) (33) suggesting that detoxification food preservative, was banned in Japan (44). Saccharin mechanisms are important. (HERP = 0.06%) is currently used in the United States (the dose- The cooking offood generates a variety of mutagens and carcino- response in rats, however, is clearly sublinear) (45). The possible gens. Nine heterocyclic amines, isolated on the basis of their hazard of diethylstilbestrol residues in meat from treated farm mutagenicity from proteins or amino acids that were heated in ways animals seems miniscule relative to endogenous estrogenic hor- that occur in cooking, have now been tested; all have been shown to mones and plant estrogens (46). Some natural carcinogens are also be potent carcinogens in rodents (34). Many others are still being widely used as additives, such as allyl isothiocyanate (47), estragole isolated and characterized (34). An approximate HERP of 0.02% (48), and alcohol (23). has been calculated by Sugimura et al. for the daily intake of these Air pollution. A person inhales about 20,000 liters of air in a day; nine carcinogens (34). Three mutagenic nitropyrenes present in thus, even modest contamination of the atmosphere can result in diesel exhaust have now been shown to be carcinogens (35), but the inhalation of appreciable doses of a pollutant. This can be seen intake of these carcinogenic nitropyrenes has been estimated to be in the possible hazard in mobile homes from formaldehyde much higher from grilled chicken than from air pollution (34, 36). (HERP = 2.1%) or in conventional homes from formaldehyde The total amount of browned and burnt material eaten in a typical (HERP = 0.6%) or benzene (HERP = 0.004%; Table 1). Indoor day is at least several hundred times more than that inhaled from air pollution is, in general, worse than outdoor air pollution, partly severe air pollution (12). because of cigarette smoke. The most important indoor air pollutant Gas flames generate NO2, which can form both the carcinogenic may be radon gas. Radon is a natural radioactive gas that is present nitropyrenes (35, 36) and the potently carcinogenic nitrosamines in in the soil, gets trapped in houses, and gives rise to radioactive decay food cooked in gas ovens, such as fish or squid (HERP = 0.06%; products that are known to be carcinogenic for humans (49). It has Table 1) (37). We suspect that food cooked in gas ovens may be a been estimated that in 1 million homes in the United States the level major source of dietary nitrosamines and nitropyrenes, though it is of exposure to products of radon decay may be higher than that 274 SCIENCE, VOL. 236 received by today's uranium miners. Two particularly contaminated related and have short life-spans. Qualitative extrapolation of cancer houses were found that had a risk estimated to be equivalent to risks from rats or mice to humans, a very dissimilar long-lived species, receiving about 1200 chest x-rays a day (49). Approximately 10% of is unlikely to be as reliable. Conversely, important human carcinogens the lung cancer in the United States has been tentatively attributed may not be detected in standard tests in rodents; this was true for a to radon pollution in houses (49). Many of these cancers might be long time for both tobacco smoke and alcohol, the two largest preventable since the most hazardous houses can be identified and identified causes of neoplastic death in the United States. modified to minimize radon contamination. For many of the chemicals considered rodent carcinogens, there General outdoor air pollution appears to be a small risk relative to may be negative as well as positive tests. It is difficult to deal with the pollution inhaled by a smoker: one must breathe Los Angeles negative results satisfactorily for several reasons, including the fact smog for a year to inhale the same amount of bumt material that a that some chemicals are tested only once or twice, while others are smoker (two packs) inhales in a day (12), though air pollution is tested many times. The HERP index ignores negative tests. Where inhaled starting from birth. It is difficult to determine cancer risk there is species variation in potency, use of the more sensitive from outdoor air pollution since epidemiologists must accurately species, as is generally done and as is done here, could introduce a control for smoking and radon. tendency to overestimate possible hazards; however, for most Some common drugs shown in Table 1 give fairly high HERP chemicals that are positive in both species, the potency is similar in percentages, primarily because the dose ingested is high. However, rats and mice (57). The HERP may provide a rough correlate of since most medicinal drugs are used for only short periods while the human hazard from chemical exposure; however, for a given HERP index is a daily dose rate for a lifetime, the possible hazard chemical, to the extent that the potency in humans differs from the would usually be markedly less. We emphasize this in Table 1 by potency in rodents, the relative hazard would be different. bracketing the numbers for these shorter exposures. Phenobarbital Quantitative unceainties. Quantitative extrapolation from ro- (HERP = 16%) was investigated thoroughly in humans who had dents to humans, particularly at low doses, is guesswork that we taken it for decades, and there was no convincing evidence that it have no way of validating (1, 5, 10, 11, 58). It is guesswork because Downloaded from www.sciencemag.org on March 3, 2010 caused cancer (50). There is evidence of increased renal cancer in of lack of knowledge in at least six major areas: (i) the basic long-term human ingestion of phenacetin, an analgesic (51). Acet- mechanisms of carcinogenicity; (ii) the relation of cancer, aging, and aminophen, a metabolite of phenacetin, is one of the most widely life-span (1, 10, 42, 59); (iii) the timing and order of the steps in the used over-the-counter pain killers. Clofibrate (HERP = 17%) is carcinogenic process that are being accelerated; (iv) species differ- used as a hypolipidemic agent and is thought to be carcinogenic in ences in metabolism and pharmacokinetics; (v) species differences in rodents because it induces hydrogen peroxide production through anticarcinogens and other defenses (1, 60); and (vi) human hetero- peroxisome proliferation (52). geneity-for example, pigmentation affects susceptibility to skin Occupational eosures can be remarkably high, particularly for cancer from ultraviolet light. These sources of uncertainty are so volatile carcinogens, because about 10,000 liters of air are inhaled in numerous, and so substantial, that only empirical data will resolve a working day. For formaldehyde, the exposure to an average them, and little of this is available. worker (HERP = 5.8%) is higher than most dietary intakes. For a Uncertainties due to mechanism in multistage carcinogenesis. Several number ofvolatile industrial carcinogens, the ratio of the permitted steps (stages) are involved in chemical carcinogenesis, and the dose- exposure limit [U.S. Occupational Safety and Health Administra- response curve for a carcinogen might depend on the particular tion (OSHA)] in milligrams per kilogram to the TD50 has been stage(s) it accelerates (58), with multiplicative effects if several stages calculated; several are close to the TD50 in rodents and about two- are affected. This multiplicative effect is consistent with the observa- thirds have permitted HERP values >1% (53). The possible hazard tion in human cancer that synergistic effects are common. The three estimated for the actual exposure levels of the most heavily exposed steps of carcinogenesis that have been analyzed in most detail are EDB workers is remarkably high, HERP = 140% (Table 1). initiation (mutation), promotion, and progression, and we discuss Though the dose may have been somewhat overestimated (54), it these as an aid to understanding aspects of the dose-response relation. was still comparable to the dose causing cancer in half the rodents. Mutation (or DNA damage) as one stage of the carcinogenic An epidemiologic study of these heavily exposed EDB workers who process is supported by various lines of evidence: association of inhaled EDB for over a decade did not show any increase in cancer, active forms of carcinogens with mutagens (61), the changes in though because of the limited duration of exposure and the DNA sequence of oncogenes (62), genetic predisposition to cancer relatively small numbers of people monitored the study would not in human diseases such as retinoblastoma (63) or DNA-repair have detected a small effect (54, 55). OSHA still permits exposures deficiency diseases such as xeroderma pigmentosum (64). The idea above the TD50 level. California, however, lowered the permitted that genotoxic carcinogens might show a linear dose-response might level over 100-fold in 1981. In contrast with these heavy workplace be plausible if only the mutation step of carcinogenesis was acceler- exposures, the Environmental Protection Agency (EPA) has banned ated and if the induction of repair and defense enzymes were not the use of EDB for fumigation because of the residue levels found in significant factors (65). grain (HERP = 0.0004%). Promotion, another step in carcinogenesis, appears to involve cell proliferation, or perhaps particular types of cell proliferation (66), and dose-response relations with apparent thresholds, as indicated Uncertainties in Relying on Animal Cancer by various lines of evidence: (i) The work of Trosko et al. (67) on Tests for Human Prediction promotion of carcinogenesis due to interference with cell-cell com- munication, causing cell proliferation. (ii) Rajewsky's and other Species variation. Though we list a possible hazard if a chemical is a work indicating initiation by some carcinogenic agents appears to carcinogen in a rat but not in a mouse (or vice versa), this lack of require proliferating target cells (68). (iii) The work of Farber et al. agreement raises the possibility that the risk to humans is nonexis- (69) on liver carcinogenesis supports the idea that cell proliferation tent. Of 392 chemicals in our database tested in both rats and mice, (caused by partial hepatectomy or cell killing) can be an important 226 were carcinogens in at least one test, but 96 of these were aspect of hepatocarcinogenesis. They have also shown for several positive in the mouse and negative in the rat or vice versa (56). This chemicals that hepatic cell killing shows a toxic threshold with dose. discordance occurs despite the fact that rats and mice are very closely (iv) Work on carcinogenesis in the pancreas, bladder and stomach 17 APRIL 1987 ARTICLES 275 (70), and other tissues (58) is also consistent with results on the liver Experimentally, it is very difficult to discriminate between the (71, 72) though the effect of cell proliferation might be different in various extrapolation models at low doses (11, 58). However, tissues that normally proliferate. (v) The work of Mirsalis et al. (71) evidence to support the idea that a nonlinear dose-response relation- suggests that a variety of nongenotoxic agents are hepatocarcino- ship is the norm is accumulating for many nongenotoxic and some gens in the B6C3F1 mouse (commonly used in cancer tests) because genotoxic carcinogens. Dose-response curves for saccharin (45), of their toxicity. Other studies on chloroform and trichloroethylene butylated hydroxyanisole [BHA (84)], and a variety of other also support this interpretation (72, 73). Cell proliferation resulting nongenotoxic carcinogens appear to be nonlinear (85). Formalde- from the cell killing in the mouse liver shows a threshold with dose hyde, a genotoxic carcinogen, also has a nonlinear dose response (71). Also relevant is the extraordinarily high spontaneous rates of (28, 29). The data for both bladder and liver tumors in the large- liver tumors (21% carcinomas, 10% adenomas) in the male B6C3F1 scale study on acetylaminofluorene, a genotoxic chemical, could fit a mouse (74). These spontaneous tumors have a mutant ras oncogene, hockey stick-shaped curve, though a linear model, with a decreased and thus the livers in these mice appear to be highly initiated effect at lower dose rates when the total dose is kept constant (86), (mutated) to start with (75). (vi) Oncogenes: As Weinberg (62) has has not been ruled out. pointed out, "Oncogene-bearing cells surrounded by normal neigh- Carcinogens effective at both mutating and killing cells (which bors do not grow into a large mass if they carry only a single includes most mutagens) could be "complete" carcinogens and oncogene. But if the normal neighbors are removed ... by killing therefore possibly more worrisome at doses far below the MTD them with a cytotoxic drug . .. then a single oncogene often than carcinogens acting mainly by causing cell killing or prolifera- suffices." (vii) Cell killing, as well as mutation, appears to be an tion (15). Thus, all carcinogens are not likely to be directly important aspect of radiation carcinogenesis (76). comparable, and a dose of 1/100 the TD50 (HERP = 1%) might be Promotion has also been linked to the production of oxygen much more of a carcinogenic hazard for the genotoxic carcinogens radicals, such as from phagocytic cells (77). Since chronic cell killing dimethylnitrosamine or aflatoxin than for the apparently nongeno- would usually involve inflammatory reactions caused by neutrophils, toxic carcinogens trichloroethylene, PCBs, or alcohol (HERP values Downloaded from www.sciencemag.org on March 3, 2010 one would commonly expect chemicals tested at the maximally marked with asterisks in Table 1). Short-term tests for mutagenicity tolerated dose (MTD) to be promoters because of the chronic (61, 87) can have a role to play, not only in understanding inflammation. mechanisms, but also in getting a more realistic view of the Progression, another step in carcinogenesis, leading to selection background levels of potential genotoxic carcinogens in the world. for invasiveness and metastases, is not well understood but can be Knowledge of mechanism of action and comparative metabolism in accelerated by oxygen radicals (78). rodents and humans might help when estimating the relative Chronic cell toxicity caused by dosing at the MTD in rodent importance of various low-dose exposures. cancer bioassays thus not only could cause inflammation and cell Human cancer, except in some occupational or medicinal drug proliferation, but also should be somewhat mutagenic and clasto- exposures, is not from high (just subtoxic) exposures to a single genic to neighboring cells because of the release of oxygen radicals chemical but is rather from several risk factors often combined with from phagocytosis (12, 79, 80). The respiratory burst from phago- a lack of antirisk factors (60); for example, aflatoxin (a potent mutagen) cytic neutrophils releases the same oxidative mutagens produced by combined with an agent causing cell proliferation, such as hepatitis B radiation (77, 79). Thus, animal cancer tests done at the MTD of a virus (19). High salt [a possible risk factor in stomach cancer (13)] and chemical might commonly stimulate all three steps in carcinogenesis high fat [a possible risk factor in colon cancer (4)] both appear to be and be positive because the chemical caused chronic cell killing and effective in causing cell killing and cell proliferation. inflammation with some mutagenesis. Some of the considerable Risk from carcinogenesis is not linear with time. For example, human evidence for chronic inflammation contributing to carcino- among regular cigarette smokers the excess annual lung cancer genesis and also some evidence for and against a general effect of incidence is approximately proportional to the fourth power of the inflammation and cytotoxicity in rodent carcinogenesis have been duration of smoking (88). Thus, if human exposures in Table 1 are discussed (81). much shorter than the lifetime exposure, the possible hazard may be Another set of observations may also bear on the question of markedly less than linearly proportional. toxicity and extrapolation. Wilson, Crouch, and Zeise (82) have A key question about animal cancer tests and regulatory policy is pointed out that among carcinogens one can predict the potency in the percentage of tested chemicals that will prove to be carcinogens high-dose animal cancer experiments from the toxicity (the LD50) of (89). Among the 392 chemicals in our database that were tested in the chemical, though one cannot predict whether the substance is a both rats and mice, 58% are positive in at least one species (14). For carcinogen. We have shown that carcinogenic potency values are the 64 "natural" substances in the group, the proportion of positive bounded by the MTD (57). The evidence from our database results is similar (45%) to the proportion of positive results in the suggests that the relationship between TD50 and MTD has a synthetic group (60%). One explanation offered for the high biological as well as a statistical basis (57). We postulate that a just proportion of positive results is that more suspicious chemicals are sublethal level of a carcinogen causes cell death, which allows being tested (for example, relatives of known carcinogens), but we neighboring cells to proliferate, and also causes oxygen radical do not know ifthe percentage of positives would be low among less production from phagocytosis and thus chronic inflammation, both suspicious chemicals. If toxicity is important in carcinogenicity, as important aspects of the carcinogenic process (57). The generality of we have argued, then at the MTD a high percentage of all chemicals this relationship and its basis needs further study. might be classified as "carcinogens." If most animal cancer tests done at the MTD are partially measuring cell killing and consequent cell proliferation and phago- cytic oxygen radical damage as steps in the carcinogenic process, one The Background of Natural Carcinogens might predict that the dose-response curves would generally be nonlinear. For those experiments in our database for which life table The object of this artide is not to do risk assessment on naturally data (14) were available, a detailed analysis (83) shows that the dose- occurring carcinogens or to worry people unduly about an occasion- response relationships are more often consistent with a quadratic (or al raw mushroom or beer, but to put the possible hazard of man- cubic) model than with a linear model. made carcinogens in proper perspective and to point out that we 276 SCIENCE, VOL. 236 lack the knowledge to do low-dose "risk assessment." We also are For example, the aflatoxin in the average peanut butter sandwich, or almost completely ignorant of the carcinogenic potential of the a raw mushroom, are 75 and 200 times, respectively, the possible enormous background of natural chemicals in the world. For hazard of EDB. Before banning EDB, a useful substance with rather example, cholinesterase inhibitors are a common class of pesticides, low residue levels, it might be reasonable to consider whether the both man-made and natural. Solanine and chaconine (the main hazards of the alternatives, such as food irradiation, or the conse- alkaloids in potatoes) are cholinesterase inhibitors and were intro- quences of banning, such as increased mold contamination of grain, duced generally into the human diet about 400 years ago with the pose less risk to society. Also, there is a disparity between OSHA dissemination of the potato from the Andes. They can be detected in not regulating worker exposures at a HERP of 140%, while the the blood of almost all people (12, 90). Total alkaloids are present at EPA bans the substance at a HERP of 0.0004%. In addition, the a level of 15,000 ,ug per 200-g potato with not a large safety factor FDA allows a possible hazard up to a HERP of 0.3% for peanut (about sixfold) from the toxic level for humans (91). Neither butter (20 ppb), and there is no warning about buying comfrey pills. alkaloid has been tested for carcinogenicity. By contrast, malathion, Because of the large background of low-level carcinogenic and the main synthetic organophosphate cholinesterase inhibitor in our other (93) hazards, and the high costs of regulation, priority setting diet (17 jg/day) (16), is not a carcinogen in rodents. is a critical first step. It is important not to divert society's attention The idea that nature is benign and that evolution has allowed us away from the few really serious hazards, such as tobacco or to cope perfectly with the toxic chemicals in the natural world is not saturated fat (for heart disease), by the pursuit ofhundreds of minor compelling for several reasons: (i) there is no reason to think that or nonexistent hazards. Our knowledge is also more certain about natural selection should eliminate the hazard of carcinogenicity of a the enormous toll oftobacco-about 350,000 deaths per year (1, 2). plant toxin that causes cancer in old age past the reproductive age, There are many trade-offs to be made in all technologies. Trichlo- though there could be selection for resistance to the acute effects of roethylene and tetrachloroethylene (perchloroethylene) replaced particular carcinogens. For example, aflatoxin, a mold toxin that hazardous flammable solvents. Modem synthetic pesticides dis- presumably arose early in evolution, causes cancer in trout, rats, placed lead arsenate, which was a major pesticide before the modern Downloaded from www.sciencemag.org on March 3, 2010 mice, and monkeys, and probably people, though the species are not chemical era. Lead and arsenic are both natural carcinogens. There is equally sensitive. Many of the common metal salts are carcinogens also a choice to be made between using synthetic pesticides and (such as lead, cadmium, beryllium, nickel, chromium, selenium, and raising the level of plants' natural toxins by breeding. It is not clear arsenic) despite their presence during all of evolution. (ii) Given the that the latter approach, even where feasible, is preferable. For enormous variety of plant toxins, most of our defenses may be example, plant breeders produced an insect-resistant potato, which general defenses against acute effects, such as shedding the surface has to be withdrawn from the market because of its acute toxicity to lining of cells of our digestive and respiratory systems every day; humans due to a high level of the natural plant toxins solanine and protecting these surfaces with a mucin layer; having detoxifying chaconine (12). enzymes that are often inducible, such as cytochrome P-450, This analysis on the levels of synthetic pollutants in drinking conjugating enzymes, and glutathione transferases; and having water and of synthetic pesticide residues in foods suggests that this DNA repair enzymes, which would be useful against a wide variety pollution is likely to be a minimal carcinogenic hazard relative to the of ingested toxic chemicals, both natural and synthetic. Some human background ofnatural carcinogens. This result is consistent with the cancer may be caused by interfering with these normal protective epidemiologic evidence (1). Obviously prudence is desirable with systems. (iii) The human diet has changed drastically in the last few regard to pollution, but we do need to work out some balance thousand years, and most of us are eating plants (such as coffee, between chemophobia with its high costs to the national wealth, potatoes, tomatoes, and kiwi fruit) that our ancestors did not. (iv) and sensible management of industrial chemicals (94). Normal metabolism produces radiomimetic mutagens and carcino- Human life expectancy continues to lengthen in industrial coun- gens, such as hydrogen peroxide and other reactive forms of oxygen. tries, and the longest life expectancy in the world is in Japan, an Though we have defenses against these agents, they still may be extremely crowded and industrialized country. U.S. cancer death major contributors to aging and cancer. A wide variety of external rates, except for lung cancer due to tobacco and melanoma due to agents may disturb this balance between damage and defense (12, ultraviolet light, are not on the whole increasing and have mostly 42). been steady for 50 years. New progress in cancer research, molecular biology, epidemiology, and biochemical epidemiology (95) will probably continue to increase the understanding necessary for Implications for Decision-Making lengthening life-span and decreasing cancer death rates. For all of these considerations, our scale is not a scale of risks to REFERENCES AND NOTES humans but is only a way of setting priorities for concem, which 1. R. Doll and R. Peto, The Causes ofCancer (Oxford Univ. Press, Oxford, England, should also take into account the numbers of people exposed. It 1981). should be emphasized that it is a linear scale and thus may 2. Smoking and Health: A Repon of the Surgeon General, Departnent of Health, Education and Welfare Publication No. (PHS) 79-50066 (Office of the Assistant overestimate low potential hazards if, as we argue above, linearity is Secretary for Health, Washington, DC, 1979). not the normal case, or if nongenotoxic carcinogens are not of very 3. G. J. Hopkins and K. K. Carroll,J. Environ. Pathol. Tavicol. Oncol. 5, 279 (1985); J. V. Joossens, M. J. Hill, J. Geboers, Eds., Diet and Human Carinogenesis much concern at doses much below the toxic dose. (Elsevier, Amsterdam, 1985); I. Knudsen, Ed., Genetic Tasicology ofthe Diet (Liss, Thus, it is not scientifically credible to use the results from rodent New York, 1986); Committee on Diet, Nutrition and Cancer, Assembly of Life tests done at the MTD to directly estimate human risks at low doses. Sciences, National Research Council, Diet, Nutrition and Cancer (National Academy Press, Washington, DC, 1982). For example, an EPA "risk assessment" (92) based on a succession of 4. R. P. Bird, R. Schneider, D. Stamp, W. R. Bruce, Carcinogenesis 7, 1657 (1986); H. L. Newmark et al., in Large Bowel Cancer, vol. 3 in Cancer Research worst case assumptions (several of which are unique to EDB) Monographs, A. J. Mastromarino and M. G. Brattain, Eds. (Praeger, New York, concluded that EDB residues in grain (HERP = 0.0004%) could 1985), pp. 102-130; E. A. Jacobson, H. L. Newmark, E. Bright-See, G. cause 3 cases of cancer in 1000 people (about 1% of all U.S. cancer). McKeown-Eyssen, W. R. Bruce, Nutr. Rep. Int. 30, 1049 (1984); M. Buset, M. Lipkin, S. Winawer, S. Swaroop, E. Fridman, Cancer Ras. 46, 5426 (1986). A consequence was the banning of the main fiumigant in the 5. D. G. Hoe, R. A. Merrill, F. P. Perera, Eds., Banbury Report 19. Risk country. It would be more reasonable to compare the possible Quantato and Regulatoy Poliy (Cold Spring Laboratory, Cold Spring Har- bor, NY, 1985). hazard of EDB residues to that of other common possible hazards. 6. B. E. Henderson et al., Cancer Res. 42, 3232 (1982). 17 APRIL I987 ARTICLES 277 7. R. Peto and H. zur Hausen, Eds., Banbury Rot 21. Viral Etiology of Cerpical 38. T. Fazio, D. C. Havery, J. W. Howard, in N-Nitroso Compounds: Analysis, Cancer (Cold Spring Harbor Laboratory, Cold Spring Habor, NY, 1986); F.-S. Formation and Occumrnc, E. A. Walker, L. Griciute, M. Castegnaro, M. Yeh et al., Cancer Res. 45, 872 (1985). Borzsonyi, Eds. (IARC Scientific PubI. No. 31, International Agency for 8. International Agency for Research on Cancer, IARCMonographs on theEvaluation Research on Cancer, Lyon, France, 1980), pp. 419-435; R. Preussmann and G. ofthe CarcinogenicRisk ofChemicals to Humans (International Agency for Research Eisenbrand, in Chemical Carcinogenesis, C. E. Searle, Ed. (ACS Monograph 182, on Cancer, Lyon, France, 1985), vol. 39. American Chemical Society, Washington, DC, ed. 2, 1984), vol. 2, pp.829-868; 9. D. A. Freedman and H. Zeisel, FromMouse toMan: TheQuantitativeArssesment of D. C. Havery, J. H. Hotchkiss, T. Fazio, J. Food Sci. 46, 501 (1981). Cancer Risks (Tech. Rep. No. 79, Department of Statistics, University of 39. W. A. Garland at al., Cancer Res. 46, 5392 (1986). California, Berkeley, 1987). 40. E. A. Walker, L. Griciute, M. Castegnaro, M. Borzsonyi, Eds., N-Nitroso 10. R. Peto, in Assesmnt ofRisk from Low-Lcvel Expoure to Radiation and Chemical, Componds: Analysis; Formation and Occurrence (IARC Scientific Publ. No. 31, A. D. Woodhead, C. J. Shellabaer, V. Pond, A. Hollacnder, Eds. (Plenum, International Agency for Research on Cancer, Lyon, France, 1980), pp. 457- New York and London, 1985), pp. 3-16. 463; B. Spiegeihalder, G. Eisenbrand, R. Preussmann, Oncogy 37,211 (1980); 11. S. W. Samuels and R. H. Adamson,J. Natl. Cancer Int. 74, 945 (1985); E. J. R. A. Scaan and S. R. Tannenbaum, Eds., N-Nitroso Compounds (ACS Calabrese, Drg Metab. Rev. 15, 505 (1984). Symposium Series No. 174, American Chemical Society, Washington, DC, 12. B. N. Ames, Science 221, 1256 (1983); ibid. 224, 668, 757 (1984). 1981), pp. 165-180. Nitrosamines are formed in cured meats through reactions 13. H. Ohgaki et al., Gann 75, 1053 (1984); S. S. Mirvish,J. NatI. Cancer Inst. 71, of secondary amines with nitrites added during the manufacturing process. One 630 (1983); J. V. Joossens and J. Geboers, in Frontien in Gastointestina Cancer, survey of bacon commercially available in Canada identified N-nitrosodimethyla- B. Levin and R. H. Riddell, Eds. (Elsevier, Amsterdam, 1984), pp. 167-183; T. mine (DMN),N-nitrosodiethylamine (DEN), andN-nitrosopyrrolidine (NPYR) Hirayama,Jpn. J. Clin. Oncol. 14, 159 (1984); C. Furihata et al., Biochem. Biophys. in most samples tested, with average levels of 3.4, 1.0, and 9.3 ppb respectively. Res. Commun. 121, 1027 (1984). The cooked-out fat from the bacon samples contained DMN and NPYR at 14. R. Peto, M. C. Pike, L. Bernstein, L. S. Gold, B. N. Ames, Envimon. Health average levels of 6.4 and 21.9 ppb, respectively [N. P. Sen, S. Seaman, W. F. Perspect. 58, 1 (1984); L. S. Gold et al., ibid., p. 9; L. S. Gold et al., ibid. 67, 161 Miles, J. Agric. Food Chem. 27, 1354 (1979); R. A. Scanlan, Cancer Res. 43, (1986); L. S. Gold et al., ibid., in press. 2435s (1983)]. The average levels of NPYR in cooked bacon have decreased since 15. G. M. Williams and J. H. Weisburger, in Casarett and Doull's Tavicolo,y. The Basic 1971 because of reduced levels of nitrite and increased levels of ascorbate used in Science ofPoisons, C. D. Klaassen, M. 0. Amdur, J. Doull, Eds. (Macmillan, New bacon curing mixtures [D. C. Havery, T. Fazio, J. W. Howard,J. Assoc. Off Anal. York, ed. 3, 1986), chap. 5, pp. 99-172; B. E. Butterworth and T. J. Slaga, Eds., Chem. 61, 1379 (1978)]. Banbuiy Rtport 25. Non-Gcnotoocic Mechanisms in Carcinogenesis (Col Spring 41. Y. Yamamoto et al., AnaI. Biochem. 160, 7 (1987). Harbor Laboratory, Cold Spring Harbor, NY, 1987). 42. B. N. Ames and R. L. Saul, in Theoris of Carcinogenesis, 0. H. Iversen, Ed. 16. The FDA has estimated the average U.S. dietary intake of 70 pesticides, (Hermisphere, New York, in press); R. Catcart, E. Schwiers, R. L. Saul, B. N. herbicides, and industrial chemicals for 1981/1982 [M. J. Gartrell, J. C. Craun, D. Ames, Proc. NatI. Acad. Sci. U.S.A. 81, 5633 (1984). S. Podrebarac, E. L. Gunderson,J.Assoc. Off Anal. Chem. 69, 146 (1986)]. The 43. B. P. Yu, E. J. Masoro, I. Murata, H. A. Bertrand, F. T. Lynd,J. Gerontol. 37,130 negative test on 2-ethylhexyl diphenyl phosphate is in J. Treon, F. Dutra, F. (1982); F. J. C. Roe, Proc. Nutr. Soc. 40, 57 (1981); Nature (London) 303, 657 Downloaded from www.sciencemag.org on March 3, 2010 Cleveland, Arch. Ind. Hyg. Occup. Med. 8,170 (1953). (1983); M. J. Tucker, Int.J. Cancer 23, 803 (1979). 17. R. C. Beier ct al., Food Chem. Toxicol. 21, 163 (1983). 44. Y. Tazima, Environ. Health Perpect. 29, 183 (1979); M. Kinebuchi, T. Kawachi, 18. L. Stoloff, M. Castegnaro, P. Scott, I. K. O'Neill, H. Bartsch, Eds., Some N. Matsukura, T. Sugimura, Food Cosmet. Tavxico. 17, 339 (1979). Mycotavim, vol. 5 inEnvironmcntal Carcinogens. Selcted Methods ofAnalysis (IARC 45. F. W. Carlborg, FoodChem. Tacicol. 23, 499 (1985). Scientific Publ. No. 44, International Agency for Research on Cancer, Lyon, 46. T. H. Jukes, Am. Stat. 36, 273 (1982);J. Am. Med. Assoc. 229, 1920 (1974). France, 1982); H. Mori etal., CancerRes. 44, 2918 (1984); R. Roschenthaler, E. 47. Allyl isothiocyanate (AITC) is the major flavor ingredient, and natural pesticide, repy,G. Dirheimer,J. Toicol.-TainRev. 3,53(1984); W. F. O.Marasas, of brown mustard and also occurs naturally in varymg concentrations in cabbage, T N. P. Krick, J. E. Fincham, S. J. van Rensburg, Int. J. Cancer 34, 383 (1984)- Environmntal Health Critena 11: Mycotcins (World Health Organization, kale, broccoli, cauliflower, and horseradish [Y. M. Ioannou, L. T. Burka, H. B. Matthews, Toxicol. AppI. Pharmacol. 75, 173 (1984)]. It is present in the plant's Geneva, Switzerland, 1979), pp. 21-85; W. F. Busby et al., in Chemical volatile oil as the glucoside sinigin. (The primary flavor ingredient of yellow Carcinogens, C. E. Searle, Ed. (ACS Monograph 182, American Chemical mustard isp-hydroxybenzyl isothiocyanate.) The A1TC yield from brown mustard Society, Washington, DC, ed. 2, 1984), vol. 2, pp. 944-1136. is approximately 0.9% by weight, assuming all of the sinigrin is converted to 19. S. J. Van Rensburgetal.,Br.J. Cancer 51, 713 (1985); S. N. Zamanetal.,Lancet AITC [A. Y. Leung, Encydopcdia of Common Natural Ingrmdict Used in Food, 1985-I, 1357 (1985); H. Austin ct al., Cancer Res. 46, 962 (1986). Drugs and Cosmetics (Wiley, New York, 1980), pp. 238-241]. Synthetic AITC is 20. A. Takada, J. Nei, S. Takase, Y. Matsuda, Hepatology 6, 65 (1986). used in nonalcoholic beverages, candy, baked goods, meats, condiments, and 21. J. M. Elwood at al., Int. J. Cancer 34, 603 (1984). syrups at average levels rangg from 0.02 to 88 ppm [T. E. Furia and B. Nicolo, 22. Aldehydes and ketones are largely responsible for the aroma and flavor of bread Eds., Fenroli'sHandbook ofFlavorngredients, (CRC Press, Cleveland, OH, 2 ed., [Y. Y. Linko, J. A. Johnson, B. S Miller, Cereal Chemisty 39, 468 (1962)]. In 1975), vol. 1, p. 19]. freshly baked bread, formaldehyde (370 >g per two slices of bread) accounts for 48. Estragole, one of numerous safrole-like compounds in plants, is present in the 2.5% of the total carbonyl compounds [K. Lorenz and J. Maga, J. Apic. Food volatile oils of many edible plants, including basil, tarragon, bay, anise, and fennel, Chem. 20, 211 (1972)]. Acetaldehyde, which is present in bread at about twice as well as in pine oil and turpentine [A. Y. LeungyEncclopedia ofCommon Natural the level of formaldehyde, is a carcinogen in rats [R. A. Woutersen, L. M. Ingredints Used in Food, Drugs and Cometc (iley, New York, 1980)]. Dried Applan, V. J. Feron, C. A. Vanderheijden, Taxicol 31, 123 (1984)] and a basil has a volatile oil content of about 1.5 to 3.0%, which contains (on average) DNA cross-linking agent in human cells [B. Lambert, Y. Chen, S.-M. He, M. 25% estragole [H. B. Heath, Source Book of Flavors (AVI, Westport, CT, 1981), Sten, Mutat. Res. 146, 301 (1985)]. pp. 222-223]. Estragole is used commercially in spice, anise, licorice, and fruit 23. Ethyl alcohol contents of wine and beer were assumed to be 12% and 5%, flavors. It is added to beverages, candy, baked gods, chewing gums, ice creams, respectively. The TDn calculation is based on M. J. Radike, K. L. Stemmer, E. and condiments at average levels ranging from 2 to 150 ppm [NAS/NRC Food Bingham, Envion. Health Perpect. 41, 59 (1981). Rats exposed to 5% ethyl Protection Committee, Food and Nutrtion Board, Chemicals Used in Food alcohol in drnking water for 30 months had increased incidences of endocrine Procesn (NAS/NRC Publ. No. 1274, National Academy of Sciences, Washing- and liver tumors. ton, DC, 1965), p. 114]. 24. C. S. Ough,J. Agrc. Food Chem. 24, 323 (1976). Urethane is also carcinogenic in 49. The estimation of risk is from human data on uranium miners and estimates of hamsters and rhesus monkeys. intake. E. P. Radford, Environ. Health Perspect. 62, 281 (1985); A. V. Nero et al., 25. Y. Fujita, K. Wakabayashi, M. Nagao, T. Sugimura, Mutat. Res. 144, 227 Sacenc 234, 992 (1986); A. V. Nero, Technol. Rev. 89, 28 (1986); R. Hanley, (1985); M. Nagao, Y. Fujita, T. Sugmura, in IARC Workshop, in press. The New York Times, 10 March 1986, p. 17. 26. M. Petro-Turza and I. Szarfoldi-Szalma,ActaAlimentarza 11, 75 (1982). 50. The average daily adult dose of phenobarbital for sleep induction is 100 to 320 27. L. J. Marnett et al., Mutat. Res. 148, 25 (1985). mg (HERS = 26 to 83%), though its use is declining [AMA Division of Drugs 28. Formaldehyde in air samples taken from all the mobile homes examined ranged AAM Drug Evaluations (Amenrican Medical Association, Chicago, IL, ed. 5 from 50 to 660 ppb (mean, 167 ppb) [T. H. Connor, J. C. Theiss, H. A. Hanna, 1983), pp. 201-202]. The TDm data in the table is for phenobarbital, which, so D. K. Monteith, T. S. Matney, Taoicol. Lett. 25, 33 (1985)]. The important role far, has been shown to be carcinogenic only in mice; the sodium salt of of cell toxicity and cell proliferation in formaldehyde carcinogenesis is discussed in phenobarbital is carcinogenic in both rats and mnice. Human studies on phenobar- T. B. Starr and J. E. Gibson [Annu. Rev. Pharmacol. Taxicol. 25, 745 (1985)]. bital and cancer are reviewed in A. E. M. McLean, H. E. Driver, D. Lowe, I. 29. J. A. Swenberg at al., Carcino en 4, 945 (1983). Sutherland, Tavicl. Lett. 31 (suppi.), 200 (1986). 30. G. Della Porta, M. I. Colnaghi, G. Parmiani, Food Cosmet. Tacicol. 6, 707 (1968). 51. Phenacetin use has gradually decreased following reports of urinary bladder and 31. Formaldehyde develops postmortem in marine fish and crustaceans, probably kidney tumors in heavy users [J. M. Piper, J. Tonascia, G. M. Matanoski, N. EngI. through the metabolism of trimethylamine oxide. The average level found in J. Med. 313, 292 (1985)]. Phenacetin also induces urinary bladder and kidney shrimp from four U.S. markets was 94 mg/kg [T. Radford and D. E. Dalsis, J. tumors in rats and mice. Agric. Food Chem. 30, 600 (1982)]. Formaldehyde is found in remarkably high 52. The human dose of dofibrate is 2 gper day for many years [R. J. Havel and J. P. concentrations (300 ppm, HERP = 29% per 100 g) in Japanese shrimp that have Kane, Annu. Rev. Med. 33, 417 (1982)]. The role of clofibrate as a peroxisome been bleached with a sulfite solution [A. Yoshida and M. Imaida,J. Food Hygknic proliferator is reviewed in J. K. Reddy and N. D. Lalwani [CRC Cnt. Rev. Soc. Japan 21, 288 (1980)]. Tio. 12,1 (1983)]. An epidemiologic study is in World Health Organization 32. J. F. Lawrence and J. R. Iyengar, Int. J. Environ. Anal. Chem. 15, 47 (1983). Report, Lancet 1984-II, 600 (1984). 33. H. d'A. Heck t al.,Am. Ind. Hyg.Assoc. J. 46, 1 (1985). 53. L. S. Gold, G. Backman, N. K. Hooper, R. Peto, Lawrence Berkeky Laboratory 34. T. Sugimura at al., in Genetic Toxclogy of the Diet, I. Knudsen, Ed. (Liss, New Repwr23161 (1987); N. K. Hooper and L. S. Gold, inMonitoringofOccupational York, 1986), pp. 85-107; T. Sugimura, Scince 233, 312 (1986). Genotavicants, M. Sorsa and H. Norppa, Eds. (Liss, New York, 1986), pp. 217- 35. H. Ohgaki ct al., Cancer Lett. 25, 239 (1985). 228; K. Hooper and L. S. Gold, in Cancer Prsevntion: Strategies in the Workplace, 36. T. Kinouchi, H. Tsutsui, Y. Ohnishi, Mutat. Res. 171, 105 (1986). C. Becker, Ed. (Hemisphere, Washington, DC, 1985), pp. 1-11. 37. T. Kawabata et al., in N-Nitroso Componds:Analysis, Formation and Ocurrenc, E. 54. California Department of Health Services, EDB Criteria Document (1985). A. Walker, L. Griciute, M. Castegnaro, M. Borzsonyi, Eds. (IARC Scientific 55. M. G. Ott, H. C. Scharnweber, R. R. Langner, Br. J. Ind. Med. 37, 163 (1980); PubI. No. 31, International Agency for Research on Canccr, Lyon, France, J. C. Ramsey, C. N. Park, M. G. Ott, P. J. Gehring, Ticol. Appl. Pharmaco. 47, 1980), pp. 481-490; T. Maki, Y. Tamura, Y. Shimamura, and Y. Naoi [Bull. 411 (1978). This has been disputed (54). The carcinogen dose reported in the Environ. Comtam. Tacicol. 25, 257 (1980)] have surveyed Japanese food for table assumes a time-weighted average air concentration of 3 ppm and an 8-hour nitrosamines. workday 5 days per week for 50 weeks per year for life. 278 SCIENCE, VOL. 236 56. R. Magaw, L. S. Gold, L. Bernstein, T. H. Slone, B. N. Ames, in preparation. 89. J. E. Huff, E. E. McConnell, J. K. Haseman, Environ. Mutagenesis 7,427 (1985); 57. L. Bernstein, L. S. Gold, B. N. Ames, M. C. Pike, D. G. Hoel, Funmam. AppI. H. S. Rosenkranz, ibid., p. 428. Toxicol. 5, 79 (1985); L. Bernstein, L. S. Gold, B. N. Ames, M. C. Pike, D. G. 90. M. H. Harvey, B. A. Morris, M. McMillan, V. Marks, Human Taoicol. 4, 503 Hoel, Risk Anal. 5, 263 (1985). (1985). 58. D. B. Clayson, Taxicol. Pathol. 13, 119 (1985); D. B. Clayson, Mutat. Res., in 91. S. J. Jadhav, R. P. Sharma, D. K. Salunkhe, CRC Crit. Rev. Toxicol. 9, 21 (1981). press. 92. Environmental Protection Agency, Position Document4 (Special Pesticide Review 59. R. Peto, S. E. Parish, R. G. Gray, in Agc-Rclatcd Facton in Carcinq,gnesi, A. Division, Environmental Protection Agency, Arlington, VA, 1983). Likhachev, V. Anisimov, R. Montesano, Eds. (IARC Scientific Publ. No. 58, 93. R. Wilson and E. Crouch, RiskiBcnefit Analysis (Ballinger, Cambridge, MA, International Agency for Research on Cancer, Lyon, France, 1985), pp. 43-53. 1982); W. F. Allman Science 85 6, 30 (1985). 60. D. M. Shankel, P. Hartman, T. Kada, A. Hollaender, Eds., Antimutagenesis and 94. P. Huber, Regulation, 33 (March/April 1984); C. Whipple, ibid. 9, 37 Anticarcinogenesis: Mechanisms (Plenum, New York, 1986). (1985). 61. B. N. Ames and J. McCann, Cancer Res. 41, 4192 (1981). 95. B. A. Bridges, B. E. Butterworth, I. B. Weinstein, Eds., Banbury Report 13. 62. R. A. Weinberg, Science 230, 770 (1985). Indicators of Genotoxic Exposure. (Cold Spring Harbor Laboratory, Cold Spring 63. A. G. Knudson, Jr., Cancer Res. 45, 1437 (1985). Harbor, NY, 1982); P. E. Enterline, Ed., Fifth Annual Symposium on Environ- 64. J. E. Cleaver, in Genes and Cancer, J. M. Bishop, J. D. Rowley, M. Greaves, Eds. mental Epidemiology, Environ. Hcalth Pcrspcct. 62, 239 (1985). (Liss, New York, 1984), pp. 117-135. 96. A national survey of U.S. drinking water supplies identified the concentrations of 65. A. D. Woodhead, C. J. Shellabarger, V. Pond, A. Hollaender, Eds., Assessment of about 20 organic compounds. The mean total trihalomethane concentration was Risk fivm Low-Level Exposure to Radiation and Chemicals: A Critical Overvw 117 pLg/liter, with the major component, chloroform, present at a mean concen- (Plenum, New York, 1985). tration of 83 ug/liter (83 ppb). Raw water that is relatively free of organic matter 66. J. Cairns, Nature (London) 255, 197 (1975); C. C. Harris and T. Sun, results in drinking water relatively free of trihalomethanes after chlorination. Carcinogencsis 5, 697 (1984); A. M. Edwards and C. M. Lucas, Biochem. Biophys. These studies are reviewed in S. J. Williamson, The Scienc ofthe Total Environment Res. Commun. 131, 103 (1985); H. Tsuda ct al., Cancer Rcs. 39, 4491 (1979); 18, 187 (1981). W. H. Haese and E. Bueding, J. Pharmacol. Exp. Ther. 197, 703 (1976). 97. Public and private drinking water wells in Santa Clara Valley, California, have 67. J. E. Trosko and C. C. Chang, in Methods for Estimating Risk ofChemical Injury: been found to be contaminated with a variety of halogenated hydrocarbons in Human and Non-Human Biota and Ecosystems, V. B. Vouk, G. C. Butler, D. G. small amounts. Among 19 public water system wells, the most commonly found Hoel, D. B. Peakall, Eds. (Wiley, New York, 1985), pp. 181-200; J. E. Trosko contaminants were 1,1,1-trichloroethane (TCA), and 1,1,2-trichloro-1,2,2-tri- and C. C. Chang, in Assesment ofRisk from Low-Level Exposure to Radiation and fluoroethane (Freon-1 13). TCA was found in 15 wells generally at concentrations Chemicals: A Critical Overvw, A. D. Woodhead, C. J. Shellabarger, V. Pond, A. of less than 30 ppb, though one well contained up to 8800 ppb, and Freon- 113 Hollaender, Eds. (Plenum, New York, 1985), pp. 261-284; H. Yamasaki, was found in six wells at concentrations up to 12 ppb. Neither chemical has been Tavicol. Pathol. 14, 363 (1986). adequately tested for carcinogenicity in long-term bioassays. In addition to these 68. M. F. Rajewsky, in Age-Related Factors in Carcinogcncsis, A. Likhachev, V. compounds, three wells also contained carcinogenic compounds at low concentra- Anisimov, R. Montesano, Eds. (IARC Scientific Publ. No. 58, International tions. Water from public supply wells may be mixed with treated surface water before delivery, thus the concentrations of these compounds that people actually Downloaded from www.sciencemag.org on March 3, 2010 Agency for Research on Cancer, Lyon, France, 1985), pp. 215-224; V. Kinsel, G. Furstenberger, H. Loehrke, F. Marks, Carcinogenesis 7, 779 (1986). receive may be somewhat reduced. Thirty-five private drinking water supply wells 69. E. Farber, CancerRes. 44, 5463 (1984); E. Farber, S. Parker, M. Gruenstein, ibid. were examined; the major contaminant was the carcinogen trichloroethylene 36, 3879 (1976). (TCE), at levels up to 2800 ppb. TCA and Freon-113 were also found in some 70. A. Denda, S. Inui, M. Sunagawa, S. Takahashi, Y. Konishi, Gann 69, 633 wells, at maximum levels of 24 ppb and 40 ppb, respectively. Though fewer (1978); R. Hasegawa and S. M. Cohen, Cancer Lett. 30, 261 (1986); R. pep drink from private water wells, the contaminant concentrations may be Hasegawa, S. M. Cohen, M. St. John, M. Cano, L. B. Ellwein, Carcinogenesis 7, gher because the water is not mixed with water from other sources [California 633 (1986); B. I. Ghanayem, R. R. Maronpot, H. B. Matthews, ToxicolVy 6, 189 Department of Health Services, California Regional Water Quality Control (1986). Board 2, Santa Clara County Public Health Department, Santa Clara Valley 71. J. C. Mirsalis at al., Carcinogenesis 6, 1521 (1985); J. C. Mirsalis ct al., Envirmon. Water District, U.S. Environmental Protection Agency, Ground Water and Mutag. 8 (suppl. 6), 55 (1986); J. Mirsalis ctal., Abstract for Fourth Internation- Drinking Water in the Santa Clara Vally: A White Paper (1984), table 8]. al Conference on Environmental Mutagens, held 24-28 June in Stockholm, Trichloroethylene may not be a carcinogen in humans at low doses [R. D. Sweden (1985). Kirnbrough, F. L. Mitchell, V. N. Houk, J. Toxicol. Environ. Health 15, 369 72. W. T. Stott, R. H. Reitz, A. M. Schumann, P. G. Watanabe, Food Cosmet. Toxicol. (1985)]. 19, 567 (1981). 98. Contaminated drinking water in the area of Wobum, Massachusetts, was found to 73. D. H. Moore, L. F. Chasseaud, S. K. Majeed, D. E. Prentice, F. J. C. Roe, ibid. contain 267 ppb trichloroethylene, 21 ppb tetrachloroethylene, 12 ppb chloro- 20, 951 (1982). form, 22 ppb trichlorotrifluoroethane, and 28 ppb 1,2-trans-dichloroethylene [S. 74. J. K. Haseman, J. Huff, G. A. Boorman, Toxiol. Pathol. 12, 126 (1984); R. E. W. Lagakos, B. J. Wessen, M. Zelen,J. Am. Stat. Assoc. 81, 583 (1986)]. Tarone, K. C. Chu, J. M. Ward, J. Natl. Cancer Inst. 66, 1175 (1981). 99. The amount of chloroform absorbed by a 6-year-old child in a chlorinated 75. S. H. Reynolds, S. J. Stowers, R. R. Maronpot, M. W. Anderson, S. A. freshwater swimming pool has been estimated [J. A. Beech, Med. Hypotheses 6, Aaronson, Proc. NatI. Acad. Sci. U.S.A. 83, 33 (1986); T. R. Fox and P. G. 303 (1980)]. Table 1 refers to the chloroform in an average pool (134 pg/liter) Watanabe, Scienc 228, 596 (1985). and for a 37-kg child. Three other trihalomethanes were identified in these 76. T. D. Jones, Health Phys. 4, 533 (1984); J. B. Little, A. R. Kennedy, R. B. freshwater pools: bromoform, bromodichloromethane and chlorodibromometh- McGandy, Radiat. Res. 103, 293 (1985). ane. U. Lahl, J. Vondusze, B. Gabel, B. Stachel, W. Thiemann [Water Rcs. 15, 77. T. W. Kensler and B. G. Taffe,Adv. Free Radical Biol. Mcd. 2, 347 (1986); P. A. 803 (1981)] have estimated absorption in covered swimming pools. Cerutti, in UCLA Symposium on Mokcular and Biology Growth Factors, Tumor 100. J. McCann, L. Horn, J. Girman, A. V. Nero, in Short-Term Bioassays in theAnalysis Promotcrs and Cancer Genes, in press; P. A. Cerutti, in Biochemical and Mokcular of Compkx EnvironmentalMixtures, V. S. Sandhu, D. M. De Marini, M. J. Mass, Epidemiology of Cancer, vol. 40 of UCLA Symposium on Molecular and Cellular M. M. Moore, J. L. Mumford, Eds. (Plenum, New York, in press). This estimate Biology, C. Harris, Ed. (Liss, New York, 1986), p. 167; in Theoies of Carcino- (Table 1) for formaldehyde in conventional homes, excludes foam-insulated genesis, 0. H. Iversen, Ed. (Hemisphere, New York, in press); H. C. Bimboim, houses and mobile homes. The figure is a mean of the median or mean of the Carcinogenesis 7, 1511 (1986); K. Trenkel and K. Chrzan, ibid. 8, 455 (1987). reported samples in each paper. For benzene, the figure is a mean of all reported 78. J. Rotstein, J. 0. O'Connell, T. Slaga, Proc. Assoc. CancerRcs. 27, 143 (1986); J. median or mean samples. The level of benzene in Los Angeles outdoor air is S. O'Connell, A. J. P. Klein-Szanto, J. DiGiovanni, J. W. Fries, T. J. Slaga, Cancer similar (U.S. EPA Office of Air Quality Planning and Standards, EPA 450/4-86- Res. 46, 2863 (1986); J. S. O'Connell, J. B. Rotstein, T. J. Slaga, in Banbury 012, 1986). Report 25. Non-Genotoxic Mechanisms in Carcinogenesis, B. E. Butterworth and 101. The average adult daily PCB intake from food estimated by the FDA in fiscal years T. J. Slaga, Eds. (Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1981/1982 was 0.2 pg/day (16). Many slightly different PCB mixtures have been 1987). studied in long-term animal cancer bioassays; the calculation of TD50 was from a 79. M. A. Trush, J. L. Seed, T. W. Kensler, Proc. NatI. Acad. Sci. U.S.A. 82, 5194 test of Aroclor 1260 which was more potent than other PCBs (14). (1985); A. I. Tauber and B. M. Babior, Adv. Free-Radical Biol. Mcd. 1, 265 102. The average consumption of EDB residues in grains has been estimated by the (1985); G. J. Cheilman, J. S. Bus, P. K. Working, Prc. Natl. Acad. Sci. U.S.A. EPA for adults as 0.006 ,g kg-' day-' and for children as 0.013 Rg kg-' day-' 83, 8087 (1986). [U.S. EPA Office of Pesticide Programs, Ethykne Dibromide (EDB) Scientific 80. I. U. Schraufstatter ct al., Proc. NatI. Acad. Sci. U.S.A. 83, 4908 (1986); M. 0. Support and Deasion Document for Grain and Grain Milling Fumigation Uses (8 Bradley, in Basic and Applied Mutagcnesis, A. Muhammed and R. C. von Borstel, February 1984)]. Eds. (Plenum, New York, 1985), pp. 99-109. 103. The leaves and roots of Russian comfrey are widely sold in health food stores and 81. L. Diamond, T. G. O'Brien, W. M. Baird, Adv. Cancer Res. 32, 1 (1980); D. are consumed as a medicinal herb or salad plant or are brewed as a tea. Comfrey Schmahl,J. Cancer Res. Clin. Oncol. 109, 260 (1985); 0. H. Iversen and E. G. leaf has been shown to contain 0.01 to 0.15%, by weight, total pyrrolizidine Astrup, Cancer Invest. 2, 51 (1984); A. Hagiwara and J. M. Ward, Fundam.Appl. alkaloids, with an average level of 0.05% for intermediate size leaves [C. C. J. Tavicol. 7,376 (1986); J. M. Ward, in Carcinogenesis andMutagenesis Testing, J. F. Culvenor, J. A. Edgar, J. L. Frahn, L. W. Smith, Aust. J. Chem. 33, 1105 Douglas, Ed. (Humana, Clifton, NJ, 1984), pp. 97-100. (1980)]. The main pyrrolizidine alkaloids present in comfrey leaves are echimi- 82. L. Zeise, R. Wilson, E. Crouch, RiskAnalysis 4, 187 (1984); L. Zeise, E. A. C. dine and 7-acetyllycopsamine, neither of which has been tested for carcinogenic- Crouch, R. Wilson, ibid. 5, 265 (1985); L. Zeise, E. A. C. Crouch, R. Wilson,J ity. Almost all tested 1,2-unsaturated pyrrolizidine alkaloids have been shown to Am. College Toxicol. 5, 137 (1986). be genotoxic and carcinogenic [H. Mori at al., Cancer Res. 45, 3125 (1985)]. 83. D. Hoel, personal communication. Symphytine accounts for 6% of the total alkaloid in the leaves and has been shown 84. N. Ito, S. Fukushima, A. Hagiwara, M. Shibata, T. Ogiso, J. Natl. Cancer Int. to be carcinogenic [C. C. J. Culvenor at al., Experientia 36, 377 (1980)]. We 70, 343 (1983). assume that 1.5 g of intermediate size leaves are used per cup of comfrey tea 85. F. W. Carlborg, Food Chem. Toaic. 20, 219 (1982); Food Cosmet. Toxicol. 19, 255 (Table 1). The primary alkaloids in comfrey root are symphytine (0.67 g per (1981). kilogram of root) and echimidine (0.5 g per kilogram of root) [T. Furuya and M. 86. K. G. Brown and D. G. Hoel, Fundam. Appl. Toxicol. 3, 470 (1983); N. A. Hiluchi, Phytochemitry 10, 2217 (1971)]. Comfrey-pepsin tablets (300 mg of Littlefield and D. W. Gaylor, J. Toxicol. Environ. Health 15, 545 (1985). root per tablet) have a recommended dose of one to three tablets three times per 87. J. Ashby, Mutagcnesis 1, 3 (1986). day. Comfrey roots and leaves both induce liver tumors in rats [I. Hirono, H. 88. R. Doll, Cancer Res. 38, 3573 (1978); and R. Peto, J. Epidemiol. Mon, M. Haga,J. NatI. CancerInst. 61,865(1978)], and the TD50 value is based Community Health 32, 303 (1978). on these results. Those pyrrolizidine alkaloids tested have been found to be at least 17 APRIL I987 ARTICLES 279 as potent as carcinogens such as symphytine. If the other pyrrolizidine alkaloids in Food, Drugs and Cosmetics (Wiley, New York, 1980)]. In 1960, safrole and safrole- comfrey were as potent carcinogens as symphytine, the possible hazard of a dailv containing sassafras oils were banned from use in foods in the United States [Fed. cup of tea would be HERP = 0.6% and that of a daily nine tablets would be Regist. 25, 12412 (1960)]. Safrole is also naturally present in the oils of sweet HERP= 7.3%. basil, cinnamon leaf, nutmeg, and pepper. 104. Agarcus bisporfs is the most commonly eaten mushroom in the United States with 106. Diet cola available in a local market contains 7.9 mg of sodium saccharin per fluid an estimated annual consumption of 340 million kilograms in 1984-85. Mush- ounce. rooms contain various hvdrazine compounds, some of which have been shown to 107. Metronidatole is considered to be the drug of choice for trichomonal and cause tumors in mice. Raw mushrooms fed over a lifetime to male and female Gardnerella infections [AMA Division of Drugs, AMA Drug Evaluations (Ameri- mice induced bone, forestomach, liver, and lung tumors [B. Toth and J. Erickson, can Medical Association, Chicago, IL, ed. 5, 1983), pp. 1717 and 1802]. Cancer Res. 46, 4007 (1986)]. The 15-g raw mushroom is given as wet weight. 108. Isoniazid is used both prophylactically and as a treatment for active tuberculosis. The TD50 value based on the above report is expressed as dry weight of The adult prophylactic dose (300 mg daily) is continued for 1 year [AMA mushrooms so as to be comparable to other values for TD50 in Table 1; 90% of a Division of Drugs, AMA Drug Evaluations (American Medical Association, mushroom is assumed to be water. A second mushroom, Gyromitra escuknta, has Chicago, IL, ed. 5, 1983), pp. 1766-1777]. been similarlv studied and found to contain a mixture of carcinogenic hvdrazines 109. D M. Siegal, V. H. Frankos, M. A. Schneiderman, Reg. Toxicol. Pharmacol. 3, [B. Toth,J. Environ. Sci. Health C2, 51 (1984)]. These mushrooms are eaten in 355 (1983). considerable quantities in several countries, though less frequently in the United 110. Supported by NCI Outstanding Investigator Grant CA39910 to B.N.A., NIEHS States. Centc Grant ES01896, and NIEHS/DOE Interagency Agreement 222-YOl-ES- 105. Safrole is the main component (up to 90%) of oil ofsassafras, formerly used as the 10066. We are indebted to numerous colleagues for criticisms, particularly W. main flavor ingredient in root beer [J. B. Wilson, J. Assoc. Off Anal. Chem. 42, Havender, R. Peto, J. Cairns, J. Miller, E. Miler, D. B. Clayson, J. McCann, and 696 (1959); A. Y. Leung, Encyclopedia of Common Natural Ingredients Used in F. J. C. Roe. Perception of Risk Downloaded from www.sciencemag.org on March 3, 2010 PAUL SLOVIC experience with hazards tends to come from the news media, which Studies of risk perception examine the judgments people rather thoroughly document mishaps and threats occurring make when they are asked to characterize and evaluate throughout the world. The dominant perception for most Ameri- hazardous activities and technologies. This research aims cans (and one that contrasts sharply with the views of professional to aid risk analysis and policy-making by (i) providing a risk assessors) is that they face more risk today than in the past and basis for understanding and anticipating public responses that future risks will be even greater than today's (2). Similar views to hazards and (ii) improving the communication of risk appear to be held by citizens of many other industrialized nations. information among lay people, technical experts, and These perceptions and the opposition to technology that accompa- decision-makers. This work assumes that those who pro- nies them have puzzled and frustrated industrialists and regulators mote and regulate health and safety need to understand and have led numerous observers to argue that the American how people think about and respond to risk. Without public's apparent pursuit of a "zero-risk society" threatens the such understanding, well-intended policies may be inef- nation's political and economic stability. Wildavsky (3, p. 32) fective. commented as follows on this state of affairs. How extraordinary! The richest, longest lived, best protected, most resourceful civilization, with the highest degree of insight into its own technology, is on its way to becoming the most frightened. T HE ABILITY TO SENSE AND AVOID HARMFUL ENVIRONMEN- Is it our environment or ourselves that have changed? Would people like tal conditions is necessary for the survival of all living us have had this sort of concern in the past? . . . Today, there are risks from numerous small dams far exceeding those from nuclear reactors. Why is the organisms. Survival is also aided by an ability to codify and one feared and not the other? Is it just that we are used to the old or are some learn from past experience. Humans have an additional capability of us looking differently at essentially the same sorts of experience? that allows them to alter their environment as well as respond to it. This capacity both creates and reduces risk. During the past decade, a small number of researchers has been In recent decades, the profound development of chemical and attempting to answer such questions by examining the opinions that nuclear technologies has been accompanied by the potential to cause people express when they are asked, in a variety of ways, to evaluate catastrophic and long-lasting damage to the earth and the life forms hazardous activities, substances, and technologies. This research has that inhabit it. The mechanisms underlying these complex technolo- attempted to develop techniques for assessing the complex and gies are unfamiliar and incomprehensible to most citizens. Their subtle opinions that people have about risk. With these techniques, most harmful consequences are rare and often delayed, hence researchers have sought to discover what people mean when they say difficult to assess by statistical analysis and not well suited to that something is (or is not) "risky," and to determine what factors management by trial-and-error learning. The elusive and hard to underlie those perceptions. The basic assumption underlying these manage qualities of today's hazards have forced the creation of a new efforts is that those wvho promote and regulate health and safety need intellectual discipline called risk assessment, designed to aid in to understand the ways in which people think about and respond to identifying, characterizing, and quantifying risk (1). risk. Whereas technologically sophisticated analysts employ risk assess- ment to evaluate hazards, the majority of citizens rely on intuitive The author is prcsident of Decision Research, 1201 Oak Street, Eugene, OR 97401, risk judgments, typically called "risk perceptions." For these people, and professor of psychology at the University of Oregon. 280 SCIENCE, VOL. 236 referring to these intangible constraints as that there is universal human exposure to the "outrage factor." It is outrageous for my rodent carcinogens of natural origin, it is neighbor to dispose of minute amounts of first important to prioritize among the Risk Assessment hazardous waste in my hot tub without my plethora of possible hazards in order to permission. Sophisticated decision analysts avoid being distracted from working on the With regard to the article by Bruce Ames know this and take it into consideration as a more important problems. The enormous et al. (17 Apr., p. 271), consider the follow- constraint. Ames et al. ignore this factor and uncertainties in the use of animal data to ing parable: I am steaming in my Berkeley the decision-analysis literature that has tried assess human risk and our lack of knowledge hot tub when my neighbor leans over the to deal with it. Although helpful in overall about the mechanisms of carcinogenesis redwood fence with a long spoon and sprinm perspective, the information in the article by make policy-making especially difficult; kles some TCE (trichloroethylene)into the Ames et al. provides little guidance in help- however, we do not imply that all problems hot tub. "What are you doing," I ask in ing us to decide if we should initiate a should be passed over until the last smoker some constemation. "It's so expensive to program to prevent underground tanks from lays down his cigarette. dispose of this legally, I thought I'd dispose leaking or how polluted a well needs to be BRUCE N. AMES of it this way," he replies. When I start to before we shut it down. RENAE MAGAW protest he points out that the "HERP" It is one thing to say that the degree of Department of Biochemisy, [Human Exposure dose/Rodent Potency ground-water contamination to date does Univertity of California, dose] from the TCE is negligible when not warrant the kind of sensational treat- Berkeley, CA 94720 compared with the chloroform from the hot ment it has received in the press. It is Lois S. GOLD tub, the aflatoxin from my half-eaten peanut another thing to ignore the "outrage factor" Biology and Medicine Division, butter sandwich, and the basil in my herb and the potential for worsening ground- Lawrence Berkeley Laboratory, Downloaded from www.sciencemag.org on February 8, 2010 salad. Although this has a reassuring effect water pollution and to imply that scientific Berkeley, CA 94720 on me, it does not prevent me from sloshing data suggest that the problem should be off to call my lawyer to obtain an injunction. passed over until the last smoker lays down This parable illustrates the strength and the his cigarette! weakness of the article by Ames et al. It is RAYMOND NEUTRA Public Health Service Revitalization reassuring to assess exposures and risks in a 956 Evelyn Avenue, Albany, CA 94706 larger context. But the decision to choose I would like to comment on Gina Kolata's between action options (stay in the tub or Response: Neutra's hot tub parable is not article about the tempest in a teapot at the call the lawyer) is governed by more than germane to the issues raised by our article. National Institutes of Health (NIH) over mere risk considerations. First, one must We did not imply that cost-benefit-risk con- the plan to revitalize the commissioned also consider the tangible and intangible siderations should be the sole basis of public corps of the U.S. Public Health Service costs of tolerating or replacing an exposure. policy. Our intention was not to provide a (News & Comment, 29 May, p. 1055). This means that my neighbor should not new regulatory policy but rather to contrib- Surgeon General Koop's prerogatives and count on convincing me to automatically ute scientific information and perspective. initiatives are clearly stated in the Public accept risks comparable to those previously Neutra's parable leaves out the benefits to Health Laws of the United States and are accepted on the basis of specific cost-benefit everyone (including health) of modem tech- just as he says they are. There is an old saw trade-offs made in other settings. Thus the nology. Every industry pollutes to some in Washington that "If it ain't broke, don't fact that the Environmental Protection extent, and reduction of exposure to pollu- fix it." It became clear at the meeting Agency, after considering the benefits of tants usually involves trade-offs, including described incompletely by Kolata that the water chlorination, accepted a particular risk loss of some benefits. Neutra's car pollutes corps was "broke" and that Koop is trying to from trihalomethanes, does not mean that I the air for those of us who walk to work, but "fix it." or the proverbial rational decision-maker, modem automotive technology benefits all Commissioned officers in the Public would allow my neighbor to continue of us, even those without cars, in many Health Service are not paid more than civil spooning TCE into my hot tub until the risk ways. A decision on whether or how much servants. Persons with medical degrees conveyed the same HERP as did the chlori- to increase the costs of transportation in (whether they treat patients or not) receive a nation! Since there are no benefits from order to reduce the pollution of cars and physician's bonus similar to physicians in bathing in TCE I will predictably tolerate trucks, depends in part on understanding other uniformed services. Nonphysicians are less risk from it than I would tolerate from the true health costs of each option. paid decidedly less than equivalent ranks in the chlorination that prevents skin infection As we pointed out, modem technologies the civil service. The value of perquisites and unsightly algal blooms! There is a sec- are constantly replacing older, more hazard- available to commissioned officers has been ond class of considerations that is most ous technologies. The reason billions of steadily diminishing in recent years. In addi- important. These are societal and ethical pounds of the solvents TCE and PCE tion, the corps promotion lists have been considerations that override cost-benefit- (perchloroethylene-the main dry-cleaning stagnant for a long time. risk considerations. Our society tends to be solvent in the United States) are used is The commissioned corps has never been intolerant of situations in which exposures because of their low acute toxicity and the other than as described in the law. That are involuntary or when one party derives dangers of the flammable solvents they re- people might have joined it for their person- the benefit and the other party bears the placed. We have also pointed out that con- al benefit does not change that, and Surgeon risk. We fear some illnesses and some ways sideration of alternative substances and pos- General Koop should get some credit for his of dying more than others. Slovic's article in sible preventative measures should be part of return to the will of Congress and the the same issue of Science (17 Apr., p. 280) the public policy decision-making process. people who elected them. emphasizes the public concem with dread In the modem context of being able to CECIL H. Fox disease and unknown outcomes. Peter Sand- measure parts-per-billion and parts-per-tril- U.S. Public Health Servie, man at Rutgers University has been publicly lion levels of substances and the realization Bethesda, MD 20205 17 JULY I987 LETTERS 235 criticized (4, 7), and a confounding factor and R. Vleggaar, Eds. (Elsevier, Amsterdam, 1986), pp. 457-471. has since been determined to be chronic 5. W. F. Busby, Jr., and G. N. Wogan, in Chemical infection with hepatitis B virus (HBV). Carcinogens, C. E. Searle, Ed. (American Chemical There is a strong association-an odds ratio Society, Washington, DC, 1984), pp. 945- 1136. of 223 for liver cancer in HBV carriers (8) 6. lARC Monorphs on the Evaluation of the Carcino- compared with an odds ratio of 10 for lung genie Risk of Chem4als to Humans, Supkment 4 to Carcinogenicity of Aflatoxins cancer in cigarette smokers (9)-between IARCMonog s, Vols. 1-29 (International Agency for Research on Cancer, World Health Organiza- liver cancer, the putative hazard from afla- tion. Lyon, France, 1982), pp. 11 and 31. * The generally well-presented articles and toxin ingestion, and chronic infection with 7. D. J. Wagstaff, Regul. Tavicol. Pharmacol. 5, 384 (1985). editorial in the "Risk Assessment" issue of HBV (10) in areas of the world where liver 8. R. P. Beasley, L-Y. Hwang, C-C. Lin, C-S. Chien, Science (17 April) contain, by my count, 12 cancer is encountered. The conclusion that Lancet 1981-11, 1129 (1981). references to aflatoxin (a mold toxin, or aflatoxin is not a likely human carcinogen is 9. R. Doll and R. Peto,J. NatI. Canccr Inst. 66, 1191 (1981). mycotoxin) and one generalization about supported by other independent studies of 10. B. S. Blumberg and W. T. London, ibid. 74, 267 mycotoxins. Each reference is presented as liver cancer (7, 11) and other cancers (12) in (1985); TecJmical Rept Series No. 691 (World an illustration of a point, but unfortunately Health Organization, Geneva, 1983). the United States. The current contention is 11. L. Stoloff, Nutr. Cancer 5, 165 (1983). much ofthe key information given is inaccu- that aflatoxin intoxication may interact with 12. T. J. Mason, F. W. McKay, R. Hoover, W. J. Blot, rate and the reader may be left with an chronic HBV infection to produce liver J. F. Fraumeni, Jr., HEW Puhl. No. (NIH) 75-780 (Dcpartment of Health, Education, and Welfare, incorrect impression of the risk from afla- cancer (13), but the evidence is not persua- Washington, DC, 1975). toxin and other mycotoxins and the manage- sive. 13. S. J. Van Rensburg et al., Br. J. Cancer 51, 713 ment of that risk. Ames et al. state (p. 273) that "[c]onsider- (1985). 14. L. Stoloff, in Carcinogens and Mutagen in the Envi- Richard Wilson and E. A. C. Crouch (p. ing the potency of those mold toxins that ronment, vol. 1, Food Products, H. F. Stitzh, Ed. Downloaded from www.sciencemag.org on February 8, 2010 267) and Lester B. Lave (p. 291) imply a have been tested and the widespread con- (CRC Press, Boca Raton, FL, 1982), pp. 97-119. toxicological basis for the Food and Drug tamination of food with molds, they repre- Administration (FDA) "action level" of 20 sent the most significant carcinogenic pollu- parts per billion of aflatoxins. In fact, that tion of the food supply in developing coun- Reponse: We and Stoloff are apparently in concentration was established in 1969, with tries." This subject has been reviewed (14). agreement that aflatoxin is a carcinogen in no toxicological basis, as the lowest at which Of those mycotoxins likely to be contami- several species, and that species differ in the identity of aflatoxin could be confirmed nants of foods, only aflatoxin, ochratoxin A, their sensitivity. Although, as we indicated by the then available methods (1). Although patulin, penicillic acid, zearalenone, T-2 tox- in our table, there are no positive experi- improved methods now allow confirmation in, and deoxynivalenol have been studied ments in mice that are suitable for calcula- of identity (a prerequisite for legal action) at with any degree of thoroughness. Aflatoxin tion of TD50, our "+" in mice is based on much lower concentrations, the "action lev- and T-2 toxin have been implicated in acute the evaluation of the International Agency el" has not been reduced. human toxicoses; no mycotoxin has been for Research on Cancer that aflatoxin in- Wilson and Crouch (table 3, p. 270), and linked with a specific cancer in humans. duces tumors in that species. The epidemio- Bruce N. Ames et al. (p. 271) state with There has been speculation that one or more logical data suggest that it is a human carcin- varying degrees of certitude that aflatoxin is trichothecenes (for example, T-2 toxin) may ogen in combination with hepatitis B virus, a human carcinogen, relying on outdated be related to esophageal cancer in some although we agree with Stoloff that the (Wilson and Crouch) or incomplete (Ames areas of Africa and Asia and that ochratoxin evidence is not of the same certainty as that et a!.) information; and Ames et al. (table 1, A may be a factor in the endemic nephritis linking smoking and cancer (1). What our p. 273) list aflatoxin as a carcinogen for observed in the Balkans. However, the risk HERP (Human Exposure dose/Rodent Po- mice, an interpretation of the data that is of human injury from patulin, penicillic tency dose) ranking points out is that at questionable. The positive observations of acid, and zearalenone has been found to be current levels of human exposure and given liver malignancies in mice were from experi- insignificant. Another 28 mycotoxins have the potency in rats, the possible hazard of ments in which large interperitoneal doses been shown to produce a cellular aberration aflatoxin in a peanut butter sandwich is were used (2). Large doses given orally by some type of mutagen screening test. I greater by 10 to 100 times than possible produced no tumors (3) (mice are generally believe that jumping to conclusions from hazards from several environmental pollut- considered to be refractory to aflatoxin car- such evidence is hazardous. Interest and ants, including trichloroethylene in contam- cinogenesis). Ames et al. could have dis- enthusiasm can easily affect the unwary to inated well water and ethylene dibromide cussed the considerable information on afla- the point that speculation changes to in- residues in grain. Yet those synthetic con- toxin metabolism and pharmacodynamics creasing degrees of certainty, with no taminants are given greater regulatory scru- (4, 5) in rats, mice, other susceptible and change in material evidence. Scientists are tiny on the basis of the results of animal resistant species, and humans (in vitro) that not immune to this disease. experiments and even in the absence of points to between-species differences. The LEONARD STOLoFF epidemiological data, indicating that they epidemiological evidence on which they rely 13208 Belepue Street, might be carcinogenic in humans. In ex- for their conclusion "that aflatoxin is a hu- Silver Sping, MDl) 20904 treme cases in the United States HERP man carcinogen" allowed a select committee values for aflatoxin reached levels of 6% of of the Intemational Agency for Research on REFERENCES the TD50 dose, which seems to us reason for Cancer, meeting in 1982, to conclude (6) 1. L. Stoloff, J. Assoc. Off Anal. Cbem. 63, 1067 concem. We also stand by our statement on only that the evidence for carcinogenicity in (1980). pollution by molds in developing countries. humans was limited, that is "a causal inter- 2. S. D. Vesselinovitch, N. Mihailovich, G. N. Wogan, In addition, new mutagenic mold toxins in L. S. Lombard, K. V. N. Rao, CancerRes. 32, 2289 pretation is credible, but altemate explana- (1972). food are constantly being found when they tions such as chance, bias, or confounding 3. G. N. Wogan, Method Cancer Res. 7, 309 (1973); 1). B. Louria, G. Finkel, J. K. Smith, M. Buse, are looked for, and it is reasonable to sup- could not be excluded." The studies on Sabouraudia 12, 371 (1974). pose many will be found to be carcinogenic which this conclusion was based can be 4. L. Stoloff, in Mycovcins and Phycotavins, P. S. Steyn (2). II SEPTEMBER I987 LETrERS 1283 We stress that it is important to view the possible hazard of aflatoxin from the per- spective of the many everyday possible haz- New! Micro-g® Table Top ards of life and with the knowledge that there are a great many uncertainties in the use of animal bioassay data in extrapolation Vibration Isolators to humans. As we discussed at length, the * High-performance portable, able in two sizes: 24" square and promotional aspects of cancer are also criti- low-profile isolator plafform for 24" x 30't, with three or four isolator cal, and it is likely that the hazard from small instruments units. Stainless tops are available in aflatoxin will be much lower in the absence * Superior attenuation efficiency three sizes: 243/4" x 35"n 293/4" of some toxicity in the liver such as from both horizontally and vertically square, and 293/4" x 35", with four * Granite or stainless steel tops in isolators. hepatitis virus, alcoholic cirrhosis, or the several standard sizes-plus See how this exclusive TMC inno- maximum tolerated dose in rodents. Since custom configurations vation can advance the precision and the HERP values for synthetic pollutants, * Gimbal Piston® isolator units can efficiency of your work-with unpar- including pesticides, are usually an order of be used with 2" thick optical alleled economy. Wrte or call today magnitude less than that from aflatoxin, breadboards - for new PRODUCT BULLETIN concem over them should be even less. Granite tops are avail- NOV and price list. BRUCE N. AMES RENAE MAGAW Department of Biochemisty, Univerity of Califoma, Downloaded from www.sciencemag.org on February 8, 2010 Brkeley, CA 94720 Lois SwiRsiy GOLD Lawrence Berkely Laboratory, Berkley, CA 94720 REFERENCES 1. H. Autrop, T. Seremet, J. Wakhisi, A. Wasunna, GCrRes. 47,3430 (1987); S. V. Thonmson etal.,J. *MC Appi. Environ. Miaobiol. 35, 1150 (1978); S. J. Technical Manufacturing Corporation Cheng ct al., Caracinogenais 6, 903 (1985). 15 Centennial Drive * Peabody, MA 01960, USA * Telephone: 617-532-6330 * Telex: 951408 Response: We generally agree both with Circle No. 112 on Readers' Service Card Stoloffs letter and the response of Ames et r.. al. However, we were aware that the reliabil- INTRODUCING THE JOURNAL OF LIPOSOME RESEARCH ity of the connection between human can- cers and exposure to aflatoxin B1 has been Call for Papers called into question by the realization that a The Journal of Liposome Research is a new publication.by Marccel Dekker, I0c. whose mission is to more important risk factor is infection with resent high quality original liposne research and a smal number of selected reviews. The subjects will be broad, ranging from biophysical analysis of liposome membranes to cliical applicabons of liposome- hepatitis B virus, which inevitably con- encapsulated drugs. Only papers focused on som aspect of liposome research will be consklered. Dr. founds the data. Nonetheless, we believe Marc J. Ostro, Vice Chairman and Chief Science Offioer of The Uposorfe Company, Inc. will be the editor- ifl-chief and to whom all manuscripts should be submitted. The Journal has attracted an outstanding that the certainty for human carcinogenesis international editorial board detailed below. It is anticipated that the first issue will be published in the is high, although not absolute; it is certainly Summer of 1987 and wil initially appear quarterly. EDITOR-IN-CHIEF superior to the evidence for cancers caused Marc J. Ostro, Ph.D. by dioxin. The 20 parts-per-billion action Vice Chairman and Chief Science Officer The Uposome Cbmpany, Inc. level for aflatoxin in peanut butter may One Research Way indeed have been set at a detection limit Princeton, New Jersey 08540 (although we do not like this practice). EDITORIAL BOARD Dr. Carl R. Alving Dr. Enrico Mihich However, as Stoloff himself points out, it Waiter Reed Army Institute of Research Roswell Park Memorial Institute has not been reduced, although a modest, in Dr. John D. Baldeschwielbr Dr. Richard E. Pgano our view inadequate, proposal to reduce it Caifornia Institute of Technoiogy Carnegie Insitute Dr. Yechezkel Barenholz Dr. Demetmios Papahajopoulos to 15 ppb was made in 197,7 long after more Hadassah Medical School University of California, San Francisco sensitive detection equipment was available. Dr. Gerald P. Bodey M.D. Anderson Hospital & Tumor Institute Dr. Bengt SamuieLsson Kitrolinska Institute The proposal was abandoned. Dr. Denis J. Chapman Dr. Alan C. Sartorell RICHARD WILSON University of London Yale School of Medicine E. A. C. CROUCH Dr. Pieter R. Cullis Dr. Tstigio Shimamoto University of British Columbia Takeda Chemical Industries, Ltd., Japan Department of PAysics and Energy and Dr. Gregory Gregoriadis Dr. Junzo Sunamoto Environmental Policy Center, The Royal Free Hospital Dr. So M. Gruner University of Nagasaki, Japan Dr. Frank Szoka Harvard Univmity, Princeton University University of California, San Francisco Cambrdc, MA 02138 Dr.. Leaf Huang Univeity of Tennessee Dr. Andre Trouet IRE-Celitarg Dr. Keizo Inoue Dr. Moseley Waite Erratum: In table 1 of the article "Changes in the University of Tokyo, Japan The Bowman Gray School of Medicine ditribution of Amencan fa inmcomes, 1947 to 1984" Dr. Maurice Kates Dr. John N. Weinstein by Frank Levy (22 May, p. 923), the first quintile (%) University of Ottawa National Institute of Health for 1949 was inadvertently omitted. It should have becn Dr. Gabriel Lopez-Berestein Dr. Gerald Weissman 4.5. M.D. Anderson Hospital & Tumor Institute N.Y.U. Medical Center 128+ Circle No. 151 on Readers' Service Card provide a fresh examination of issues, in the United States and then move to a con- large part because the authors selected have sideration of which impairments of health familiar and entrenched positions. Instead, might reasonably be associated with expo- it reinforces three persistent fallacies: First, sure to chemical agents, with the use of such that the only primary concern is cancer; techniques as biological markers to support Cost of International Congresses second, that the data on exposure are reli- proposed linkages (2). After such an analy- able; and third, that bare calculations of sis, rational ranking might occur. Recently I received the first circular of the health risk can be expected to guide human This method would revise our current 28ti International Geological Congress, to behavior. practice of going from the chemical by be held in Washington, D.C., in 1989. Richard Wilson and E. A. C. Crouch (p. means of its toxicology to the estimation of Preregistration costs $250 (U.S.), and the 267) have long lamented the failure of the health impact, the Environmental Protec- cost of the technical excursions (probably public to rationalize their "risk portfolios," tion Agency dogma ofhazard identification, the most informative and useful activity at which suggests that the authors rather than risk characterization, exposure assessment, geological congresses) ranges from between the public are slow to learn that no one and then to risk assessment, as explicated by $300 and $2000. This means that the mini- makes choices solely on the basis of simple Milton Russell and Michael Gruber (p. mum cost of attending the congress and one equations or point estimates. Physicist-soci- 286). Such an approach, while radically excursion is $550, which is equivalent to ologists of risk need to note that some ofthe different from current science policy, could approximately 1 month of my salary. If one recent work in the study of economic behav- avoid some of the silliness of current regula- takes into account the cost of air travel to ior has provided a framework for a more tory practice, which provokes not only the and from Washington (approximately complex analysis of consumer choice in the amusement of scientists but also the disgust $500) and a 10-day stay in Washington (at marketplace in place of simple comparisons of the public as it observes continued failure Downloaded from www.sciencemag.org on February 8, 2010 least $1500), the total cost of attending the of marginal benefit and cost. The proposal to deal efficiently, at the source, with obvi- Congress is approximately $2550, or the by Bruce N. Ames etal. (p. 271) for ranking ously significant environmental risks like equivalent of about 8 months of my salary. risk of carcinogens, while elegant in struc- lead, sulfur dioxide, radon, formaldehyde, The total official allowance currently avail- ture, is not realistic or implementable. First, and asbestos. able for foreign travel at our institute is as a basis for the HERP (Human Exposure ELLEN K. SILBERGELD $500. These figures clearly indicate that dose/Rodent Potency dose), it relies heavily Environmental Defense Fund) many Venezuelan and Latin American geol- on the assumption that there are reliable 1616 P Street,NWj, ogists will not be able to attend the most data on exposure. Assessment of exposure Washington, DC 20036 important intemational meeting in their remains the weakest aspect of evaluating profession. And this situation is likely to risks for regulatory purposes. The failure to REFERENCES worsen in the future. require meaningful information on new 1. E. K. Silbergeld, Nat. Rts. Environ. 2, 17 (1986). Therefore I would like to urge the orga- chemicals and overreliance on models rather 2. Board on Environmental Sciences and Toxicology, National Academy of Sciences-National Research nizing committees of internationat meetings than on monitoring have resulted in a void Council, Biolgical Marken and Environmental Medi- to take these considerations into account of information for calculating human expo- cinc (National Academy Press, Washington, DC, in and to seek to provide facilities for Third press). sure. When this lack of data is factored into World participants. Otherwise, internation- an equation already burdened by the range al congresses will just be regional, rich- ofunresolved issues and uncertainties of risk Risponse: Silbergeld does not emphasize country meetings. assessment (1), it is doubtfiul how much thie importance of settmng pnonties m re- CARLOS SCHUBERT practical use the approach ofAmes et al. can search and regulation, so that efforts to Instituto Venezolano de be. Second, any comprehensive system rank- protect public health are not diverted from Invest,qaciones Cicntificas, ing risk should be capable of devolution to the most important issues. Since regulation Ministerio de Sanidady Asistencia Social, deal with risk control decisions at the mar- of carcinogens has been based largely on Apartado 21827, Caracas, Venezuela gin. That is, it is important to be able to results of rodent bioassays, it is necessary to determine how to deal with, for instance, recognize that about half of all chemicals risks of dioxin from incinerator emissions in tested at the maxiu m tolerated dose are populations who smoke, eat certain foods, carcinogens in rodents, whether the chemi- Risk Assessment sunbathe, or otherwise engage in risky busi- cals are natural or man-made. We believe ness. It is hard to know how to use the that our attempts to provide a framework Risk assessment may have its fimny side, approach of Amnes et al. for this critical for setting priorities among human expo- as noted by Daniel E. Koshland, Jr. (Edito- assessment. sures to rodent carcinogens is of practical rial, 17 Apr., p. 241), but current misman- Finally, the approach of Ames et al. and use. One contribution is to show that possi- agement of risk by regulatory agencies is no much of the discussion of risk assessment in ble carcinogenic hazards to humans from laughing matter. Identifying, controlling, Scince and elsewhere continues to confine current levels of pesticide residues or water and setting priorities for risks within the our national debate to one end point- pollution are likely to be of miniimal concern areas that Congress has designated for feder- cancer risk. While evaluating the potential relative to the background levels of natural al activity has been extraordinarily inconsist- risks of chemicals as carcinogens is impor- substances, although one cannot say wheth- ent and unprotective. Koshland's reaction is tant, the human disease and dysfunction that er these natural exposures are likely to be of not unlike that of most environmentalists, can reasonably be associated with impacts of major or minor importance. Another contri- who have long worried that the practice of chemical exposure and environmental modi- bution is to examine the many uncertainties risk assessment to date has not improved fications are likely to be expressed in many in relying on animal cancer tests for human health or advanced policy. other outcomes. The debate on risk assess- prediction given our current understanding Unfortunately, the special Risk Assess- ment needs to be radically revised; it should of the mechanisms of carcinogenesis. ment issue of Science (17 April) does not start with an assessment of health status in Silbergeld states that it is a fallacy to treat i8 SEPTEMBER I987 LEmTRS 1399 cancer as "the only primary concern." We geld's example, our highest priority should Drinking Water Standards of the Environ- agree: it is also desirable to set priorities for be to warn people about the carcinogenic mental Protection Agency, in a discussion of chemicals that cause other toxicological and teratogenic hazards of smoking and risks of organic hydrocarbons (1), omits any problems. In both cases it is counterproduc- alcohol and of the carcinogenic hazards of mention of chloroform, thereby withholding tive to focus on quantities that are minute sunbathing and to investigate the dietary from the Administrator and from the public relative to their toxic level. Although our imbalances that appear likely to be major the instructive comparison with risks of work focused on cancer, our methods are causes of cancer. trichloroethylene in our table 2 and on page also relevant to other biological end points, Silbergeld laments the quality of exposure 269 of our article. including reproductive damage. Ranking data. Yet our society has made an enormous We agree that no one makes choices solely priorities among possible teratogenic haz- effort to measure exposures to man-made on the basis of simple equations or point ards is important, especially since fully one- pollutants and to regulate them at a large estimates and have said so in almost all of third of the 2800 chemicals tested in labora- economic cost. We have tumed up remark- our writings, including the last paragraph of tory animals have been shown to induce ably little of public health interest aside from our article in Science. However, that is no birth defects at maximum tolerated doses occupational hazards. Additional measure- excuse for not accurately determining the (1). Humans are ingesting enormous ex- ments of parts per billion or per trillion of point estimate-and the uncertainty of that cesses of natural chemicals compared with man-made pollutants do not seem likely to estimate-and for putting these numbers man-made ones. For example, we ingest make a major contribution. into perspective by comparison. about 10,000 times more of nature's pesti- Silbergeld states that the public is con- Public health officials, both in private and cides than man-made pesticide residues (2). cerned with more than "bare" calculations of public, have in the last century emphasized Thus, one priority should be to estimate health risks. That may be, but it is the job of acute effects that occur as a result of a short, whether their toxicological effects might be scientists to provide the best estimates that high exposure. For these it is generally Downloaded from www.sciencemag.org on February 8, 2010 in about the same proportion. There is no they can about possible hazards. This in- assumed that a low exposure means a risk convincing evidence, either epidemiological dudes putting worst-case estimates ofhypo- dose to zero. Risk assessors follow public or toxicological, to suggest that pollution is thetical human risks in perspective. Our demand in addressing the risk of cancer-a likely to be of great teratogenic interest work suggests that traces of pollutants are chronic effect arising from long exposure, relative to the background of natural chemi- likely to be of only minimal concern relative often at lower levels. For these it is often cals. to the background of natural chemicals. assumed that there is linearity between re- Silbergeld's reference to dioxin pollution Epidemiological evidence indicates that sponse (probability of cancer) and dose. seems to imply that new incinerators should there is no epidemic of cancer (other than However, as we emphasized, the risk calcu- not be built until we know that dioxin poses that due to smoking) or of birth defects. lations for cancer can be a surrogate for no harm "to people who smoke, eat certain The biological understanding of the other end points also. foods, sunbathe, or otherwise engage in causes of cancer and birth defects is pro- Since for chrotiic effects risk is approxi- risky business." Such an approach is imprac- gressing remarkably rapidly, considering the mately dose times potency, dose informa- tical toxicologically and is an invitation to complexity of the problem. Silbergeld's sug- tion is vital. When it is available, a direct paralysis. To attempt to avoid all exposures gestions are not likely to change the prior- comparison such as, for example, for the that might cause some type of harm to ities of the many accomplished scientists radiation doses in our table 1, is less uncer- someone under some circumstances ignores working in this area. tain, and we find that people are helped by the background of natural hazards, the BRUCE N. AMES tiis. Again, however, we find that regula- benefits of technology, and the hazardous Department ofBiochemistry, tory agencies and newspapers often omit side effects of the alternatives when some Univerity of California, this comparison, thereby failing to ade- technology is eliminated. Is dioxin of impor- Berkeky, CA 94720 quately inform the public of the risk and its tance at the tiny levels people are exposed to Lois SwIRsKY GOLD meaning. This makes the risk assessment from incinerators when compared with the Biology and Medicine Division, useless and any decision less well based than "risky business" people are already engaged Lawrence Berkeley Laboratoy, it need be. in? Silbergeld's letter has prompted us to Berkeley, CA 94720 We would also like to note, as kindly compare dioxin and alcohol in terms of the RENAE MAGAW pointed out by Ertiest V. Anderson, that in exposures to humans relative to the dose Department of Biochemistry, the discussion in our article of "Expression levels that have been shown to be teratogen- Univenity of California, Berkeley of risks" (p. 270, paragraph 2, line 24), an ic to mice in laboratory experiments. Unlike arithmetic error occurred: 0.0047% should dioxin, alcohol is a known, and important, REFERENCES have been 0.023%. human teratogen. The teratogenic dose of 1. J. L. Schardein, B. A. Schwatz, M. F. Kenal, Environ. RicHARD WILSON alcohol for mice is more than a million times Heath Perspea. 61, 55 (1985). E. A. C. CROUCH 2. B. N. Ames, R. Magaw, L. S. Gold, Sciewc 236, 271 greater than the teratogenic dose of dioxin, (1987). Department of Physics and similar to the difference in carcinogenic 3. D. J. Paustenbach, H. P. Shu, F. J. Murray, Regul. Tweicol. Phamacd. 6, 284 (1986). Energy and Environmwntal Policy Center, doses for the two chemicals. However, be- Harvard Univerity, cause the dose of alcohol in a bottle of beer Cambrde, MA 02138 is very high, drinking a daily beer would Respoe: The criticism by Silbergeld REFERENCES pose a possible teratogenic hazard about the should primarily be addressed to the risk 1. Fed. Regist. 50, 46880 (13 November 1985). equivalent of eating a daily kilogram of dirt management procedures of the federal gov- contaminated with 1 part per billion of ernment and society in general. One possi- Eratum: In the Research News artide "Taking a dioxin. Soil ingestion is considered by gov- ble reason that risk management has been June,doser look at AIDS virus relatives" by Jean L. Marx (19 p. 1523), Beatrice Hahn was incorrectly identified enment regulatory agencies to be the main inconsistent is a failure of regulatory agen- as a mcmber of the Galo-Wong-Staal group. Although possible route of exposure (3). Given the Hahn collaborates with Gallo and Wong-Staal of the cies to properly inform the managers in the National Cancer Institute, she is in the Department of information available concerning Silber- same agencies. For example, the Office of Medicinc of the University of Alabama at Birmingham. I4-00 SCIENCE, VOL. 237 ago. (2). Ames et al. note that some pyroly- cers, such as breast cancer, appear closely sis products are potent carcinogens. Howev- related to patterns of dietary fat consump- er, fire-cooked wild game meats have been tion (7). But several cancers, with no known Paleolithic Diet, Evolution, and consumed by humans for at least 700,000 or suspected nutritional basis, have been Carcinogens years; for example, in Lantian, China (3), increasing. Moreover, some food-related along with a variety of plants (4). cancers, including stomach cancer have been Philip H. Abelson (Editorial, 31 July, p. A recent visit with my son Aaron to the declining in many industrial countries (8). 473) and Bruce N. Ames et al. (Articles, 17 expanded exhibit at the Hall of Fossils ofthe In the United States cancers in persons Apr., p. 271) observe that cancer is a com- Smithsonian Institution's Museum of Natu- under age 45 have also declined markedly in plex of diseases with multiple causes, rang- ral History provided some relevant informa- recent years (9). In contrast, multiple myelo- ing from carcinogens and hormonal factors tion. Reconstructions of the earliest archeo- ma, lung cancer, and brain cancer have to chronic infectious diseases and dietary logical sites ofhuman ancestors indicate that increased at least 50% from 1968 to 1978 in patterns. Moreover, Ames et al. advise that the larger, more robust form of Australo- white and nonwhite persons aged 75 to 84. naturally occurring carcinogens in the food pithecus, Homo robustus, died out about 1 (9, 10). From 1975 to 1984, the age-adjust- supply are generally more toxic than indus- million years ago and probably depended on ed U.S. cancer mortality rate rose from trial carcinogens, excepting workplace expo- vegetable foods, as its huge molar teeth and 162.2 to 170.7 per 100,000 individuals,; sures. This interpretation of greater toxicity massive jaws are well adapted for such a during this same time, the death rate per of food-borne carcinogens derives from the rough diet. A sagittal crest (bony ridge of 100,000 for nonlung cancer changed from HERP [Human Exposure dose/Rodent Po- the top of the skull) and protruding cheek 125.4 to 125.1 (11). tency dose] index ofAmes et al., which uses bones anchored the strong chewing muscles. In light of these complex patterns, serious data from animal studies of carcinogenicity The hominids from which we evolved had research needs to be done on possible Downloaded from www.sciencemag.org on February 8, 2010 and finds alcohol and peanut butter more teeth that were adapted for an omnivorous changes in the environment in the past that potent than pesticide residues. diet of vegetables and meat and lived about could account for these patterns. Whether While the work of Ames et al. presents an 1.2 to 3. million years ago. Moreover, the recent chemical exposures are linked with interesting use of toxicological data, it range of early diets was extensive, from changing cancer patterns in the elderly re- should not be construed as the final word on protein rich diets of far northern peoples to mains an open question. However, in the the role of synthetic organic carcinogens in the vegetable-laden diets of the Australian past three decades, production of synthetic producing cancer patterns in humans. The Kalahari. organic chemicals grew exponentially (Fig. 1). relative contribution of different synthetic To be sure, materials causing chronic ill- This older cohort includes persons who have and natural toxicants to human evolution nesses that are commonly expressed in post- lived long enough to experience cancers that and to current cancer and other disease reproductive persons would not have a selec- may be associated with such exposures. patterns is a complex matter. A National tive influence on the evolution of human As Ames et al. point out, the range of Research Council (NRC) report (1) noted genotypes. However, such materials could variation m worldwide cancer patterns is that many of the nondietary toxicants in have had major effects on human develop- substantial, running at least sixfold, and foods are not known to be harmful to ment. Experimental data suggest that few many cancers occur with even greater varia- normal healthy human beings when the carcinogens are not also toxic to reproduction tion (8). Diet alone is unlikely to explain all foods are prepared in time-honored ways. (5). Thus, exposure to food-bome toxicants in of this variation, nor are changes in diet Adequate cooking reduces or destroys the early humans may have selected out genotypes likely to be involved with some of the harmful properties of the cyanogenetic gly- that produced spermatocytes, oocytes, embry- specific changes noted above. cosides in the lima bean, the goitrogens in os, and fetuses with susceptibility to toxic The relative roles of food and nonfood certain vegetables, thiaminase in fish, and constituents of foods. Early pregnant humans carcinogens are unclear. It is highly likely avidin in the egg. After ripening, the ackee may have expenenced spontaneous abortions that the impact of the latter may differ fruit and grapefruit lose their toxic compo- due to prenatal and other exposures to carcin- qualitatively from that of the former. Also nents. ogens in the food supply, which would have synergies may occur between them, with Some observations from studies of Paleo- produced genetic resistance in the human newer compounds enhancing the toxicity of lithic nutrition may also be relevant, as genome. longer established compounds. In light of widely varying foods were available to Nearly four decades ago, J. B. S. Haldane the relatively recent increase in the volume evolving hominids at least 4 million years argued that diseases are responsible for of production of some carcinogenic and much of the observed biochemical and ge- other hazardous substances, it is not now netic variability of wild populations, insofar possible to determine the extent to which 0340- as the struggle against disease plays an im- exposures to such chemicals will influence portant evolutionary role (6). Reasoning future cancer rates. Prudent public policy 300 that a small biochemical change provides a dictates that additional research be conduct- 260 host species a substantial degree of resist- ed on the relative potencies of these materi- X 220 ance, Haldane argued that it is an advantage als for humans. o ~180- to a species to be biochemically diverse. DEVRA LEE DAviS 140- Whatever the role of evolution may prove Board on Environmental Studies and 100- to be, humans have been eating complex Toxicology, 0~60- foods far longer than they have been ex- National Research Council, 201945 ...1955...1965. 1975.. 1985 posed to synthetic, organic carcinogens. 2101 Constitution Avenue, NW, 1945 1955 1965 1975 1985 Moreover, some cancer patterns in the Unit- Washington, DC 20418 Year Fig. 1. Production of synthetic organic chemicals, ed States have changed markedly and recent- including tar and primary products from petro- ly in ways that are unlikely to be related to REFERENCES leum and natural gas, 1945 to 1986. changes in food consumption. Other can- 1. National Research Council, Taicants Occut7in i8 DECEMBER I987 LETTERS 1633 Naturaly in Food (National Academy of Sciences, few of nature's pesticides that we are eating human cancers, convincingly point out why Washington, DC, 1966). have been tested for carcinogenicity, but such apparent increases may be due to recent 2. S. B. Eaton and M. J. Konner, N. Engl. J. Med. 312, 283 (1985). about half of the naturally occurring sub- improvements in diagnosis (6). Peto con- 3. J. D. Clark and J. W. K. Harris, Afr. Arrhaeol. Rev. stances that have been tested in rats and cluded, in commenting on this matter (7, p. 3, 3 (1985). 4. A. B. Stahl, CGru. Anthropol. 25, 151 (1984). mice are carcinogens. We also pointed out 283), that "Future trends may differ sub- 5. D. L. Davis, Tavic Subst. J. 1, 205 (1979). that the modern diet is vastly different from stantially from recent trends, of course, but 6. J. B. S. Haldane, On Being the Robt Size and Other that of a few thousand years ago or of at present the U.S. data contain no clear Essays (Oxford Univ. Press, Oxford, England, 1985). 7. National Research Council, Diet, Nutrition and primitive man (3). Davis dismisses dietary evidence for any generalized increase in can- Caner (National Academy of Sciences, Washing- and other life-style factors too readily as cer over and above that due to the delayed ton, DC, 1983). potential causes of cancer that do not effects of tobacco. Opposite conclusions by 8. Cancer Incidewc on Five Continents (International Agency for Research on Cancer, Lyon, France, change; they do change all of the time. For other commentators appear to derive chiefly 1976), vol. 3. example, as part of the back-to-nature move- from methodological oversights." 9. S. S. Devesa et al. J. Natl. Cancer Inst. 79, 701 (1987). ment we are eating canavanine in alfalfa From a policy perspective, we discussed in 10. D. L. Davis, A. D. Lilienfeld, A. Gittelsohn, M. E. sprouts, carcinogenic hydrazines in raw our article that it is prudent to consider the Scheckenbach, Tasicol. Ind. Health 2, 127 (1986). mushrooms, and carcinogens in herb teas. benefits of modern technology and also the 11. J. C. Bailar, III, Isues Sci. Technol. 4, 16 (1987). Cooking food does destroy some carcino- alternative substances that might replace Respone: Davis takes issue with our docu- gens but also makes others, such as the regulated compounds. Modern chemicals mentation that carcinogenic hazards from variety of nitrosamines and nitropyrenes commonly replaced more hazardous sub- current levels of pesticide residues or water formed when food is cooked in gas ovens, a stances, for example, chlorinated solvents pollution are likely to be of minimal concern relatively recent invention. Davis' argument replaced flammable solvents. Modern tech- relative to the background levels of natural that natural selection eliminated all hazards nology, which concomitantly causes the in- Downloaded from www.sciencemag.org on February 8, 2010 substances. She indicates that humans, as from carcinogens acting late in life because crease in production of synthetic chemicals, opposed to rats or mice, may have devel- they are reproductive toxins is not support- has contributed in important ways to our oped secific resistance to these natural chem- ed by good evidence and appears unlikely. steadily increasing life-span. Currently, as a icals, since we have been selected by evolu- We have discussed why "risk assessment" society our expenditures on pollution abate- tion to deal with plant toxins or cooked based on worst-case scenarios may not have ment and control are more than $80 billion food. This is unlikely, because, as we dis- much to do with biological reality for either annually (Fig. 1), despite the uncertainty of cussed in our article, both rodents and hu- synthetic or natural chemicals. Linear ex- whether environmental pollutants at parts- mans have developed many types of,general trapolations from results at the maximum per-billion levels have public health signifi- defenses against the large amounts and enor- tolerated dose may enormously exaggerate cance. We believe that the potential carcino- mous variety of toxic chemicals in plants risks at low dose if, as appears to be true, an genic hazards of pollutants should be evalu- (nature's pesticides). These defenses include important aspect of carcinogenesis is cel ated in the context of background level the constant shedding of the surface layer of proliferation, which may frequently result exposures to natural substances until science cells of the digestive system, the glutathione from the high (maximally tolerated) doses of makes the further understanding of mecha- transferases for detoxifying alkylating test chemicals administered in rodent bioas- nisms clearer, as we emphasized in our agents, the active excretion of hydrophobic says (4). Concern with very low doses is article. toxins out of liver or intestinal cells (1), even more likely to be misplaced for agents BRUCE N. AMES numerous defenses against oxygen radicals suspected of causing birth defects, because Department of Biochemisty, (2), and DNA excision repair. The fact that of a threshold effect. In this respect it would Univetsity of California, Berkely, CA 94720 defenses appear to be mainly general, rather be useful to compare rodent data for partic- Lois SwiRsKY GOLD than specific for each chemical, makes good ular synthetic chemical pollutants with those Lawrence Berkeley Laboratory, evolutionary sense and is supported by vari- for a representative set of natural chemicals, Berkeley, CA 94720 ous studies. Experimental evidence indicates analogous to our HERP index comparisons. that these general defeinses will work against One important comparison to be made REFERENCES both natural and synthetic compounds, would be that between alcohol and other 1. F. Thiebaut et al., Proc. Natl. Acad. Sci. U.S.A. 84, since basic mechanisms of carcinogenesis are rodent teratogens. Alcohol is a leading cause 7735 (1987). 2. B. N. Ames, in Detection Methods for DNA-Damag- not unique to either. of mental retardation in humans (fetal alco- ing Agents in Man (International Agency for Re- We also pointed out that humans ingest hol syndrome), and such a comparison search on Cancer, Lyon, Frapce, in press). 3. L. A. Cohen, Sci. Am. 257, 42 (November 1987). about 10,000 times more of nature's pesti- would put possible teratogenic hazards into 4. J. AK Swenberg ct al., Environ. Hcalth Perpect., in cides than man-made pesticides. Relatively perspective. press. The key issue is not that production of 5. L. S. Gold, G. M. Backman, N. K. Hooper, R. Peto, ibid., in press. synthetic chemicals has gone up markedly in 6. R. Doll and R. Peto, The Causas of Cancer (Oxford 70 9000 recent years, but whether the tiny amounts Univ. Press, Oxford, England, 1981). National of pesticide residues or water pollutants we 7. R. Peto, in Quant#ication ofOccupational Cancer, R. X ' environmental 7000 O Peto and M. Schneiderman, Eds. (Banbury Report are ingesting are likely to be important in 9, Cold Spring Harbor Laboratory, Cold Spring 0 50 spending human cancer. In our ranking, such expo- Harbor, NY, 1981), pp. 269-284. 0 ~~~~~~5000~ 8. J. Hirschhorn, SeiousReduction ofHazardous Waste Pages of .8 sures are very low compared with the back- (Office of Technology Assessment, Washington, 30 X federal 3000 ground of natural carcinogens, but we also DC, 1986), figure 1, p. 8. regulations pointed out that workplace exposures often 10 1000 rank high (5). 1972 1976 1980 1984 Davis contends that the incidence of brain Definition Required Year tumors and multiple myelomas in the elderly Fig 1. Expenditures for environmental protec- has clearly increased. However, Doll and Concerning "Science and mutual self-in- tion (8). Peto, in a detailed analysis of the causes of terest' by David Dickson and Colin Nor- 1634. SCIENCE, VOL. 238