Document 3KvMggQRe24L67pR83R9yR80

Carcinogens and Human Health: Part 1 Bruce N. Ames and Lois Swirsky Gold (Perspective, 31 Aug., p. 970) posit that most human exposure to synthetic chemicals that are rodent carcinogens poses little or no risk of cancer. They argue that the high doses used in rodent bioassays cause tumors largely by inducing cytotoxicity with resultant compensatory cell proliferation (mitogenesis) that converts DNA damage (mostly caused by endogenous compounds in food) into mutations. In their view, mitogenesis dominates the carcinogenic process with the result that thresholds exist for "nongenotoxic" rodent carcinogens, and the doseresponse curve for genotoxic carcinogens is sublinear. They conclude that the current U.S. regulatory policy, which calls for controlling involuntary exposures to industrial chemicals and pesticides identified as carcinogenic in the laboratory, imposes unnecessary costs on society and conveys no benefit in terms of health protection. Thus they dismiss the potential risks from the more than 1 billion pounds of pesticides and related products produced annually in the United States and the estimated 22.5 billion pounds of toxic chemicals released or disposed of each year in this country ( 1 ) . These arguments are not new (2). Thus far, however, lengthy deliberations by U.S. and international health protection agencies, scientific advisory boards, and panels of experts have rejected proposals to relax standards for carcinogens and have supported the use of animal tests as predictors of effects in humans (3).U.S. agencies involved in risk assessment policy have adopted the general assumption of low-dose linearity for carcinogens-regardless of their presumed mechanism of action. The rationale for these decisions is threefold: (i) the lack of adequate understanding of mechanisms by which carcinogens (especially those termed "nongenotoxic") exert their effect; (ii) the absence of an identifiable threshold or safe level of exposure for a diverse human population; and (iii) the desirability of preventing cancer through the use of testing in model systems, obviating the reliance on epidemiologic data in humans. This rationale remains valid in light of current knowledge. First, a large body of data on chemical carcinogenesis and the molecular biology of cancer supports a far more intricate mechanistic explanation of tumor induction by both nongenotoxic and genotoxic carcino- gens than otic which is dominated by mito- gcncsis. Available rodent bioassay data d o not show a consistent correlation between organ toxicity at the target site and carcinogenicity ( 4 ) . Moreover, there are few cases of rodent carcinogens that are positive only at the high dose (4-5). In addition, not only does epidemiology fail to show a threshold at the lower bound of exposure to carcinogens in the workplace, but low-level community exposures to "occupational carcinogens" such as arsenic have resulted in increased incidence of cancer (6). On another level, the multistage process of cancer development is known to involve complex biology and etiology of the disean itself. FREDERICAP.PERERA Division of Environmental Sciences, Columbia University School of Public Health, 60 Haven Avenue, E-109, New York, NY 10032 REFERENCES AND NOTES 1. Synthetic O p n i c Chemicals: U.S. Produrtwn am' Sales (International Trade Commission, Washing ton, DC, 1988);E P A T& Release Inventory: A Natiomf Penoective ( 0 6 c c of Toxic Substmca. Environmental Pro&on Agency, Washingma) DC, 1989). I (clilorcrmctliyl) have Iwii avoi 12. J, Huff and 13. Medirine, 1. M ton-Ccntury-(: Futun Risk: RI ronmcntal Prc Board, Washin Rev. Public Hrl 13. F. Percra, P. BI es in Onco/cgy? Rnscnberg, E4 1989). pp. 249 Response: Pel views accurate11 dence that con the "toxic cheo both mutagenic and nonmutagenic mecha- tion" view of ( nisms. These result in the induction of multiple direct and indirect genetic changes at target oncogenes or tumor suppressor genes as well as alterations in signal transduction pathways involved in growth control (7). There is no evidence that these molecular events occur only at high, toxic doses (8). Despite recent exciting advances in the molecular biology of cancer, many uncertainties Cold Spring Harbor, NY, 1977),pp. 187-205;I. case, low-dose r H. Weisburger and G . M. Williams, Science 214, 401 (1981);W. T. Ston, R. H. Reitz, A. M. Schumann, P. G. Watanabc, Food Cosmef. Toxicd. Perera's state1 causing the en1 18 567(1981);R. Kroa, N.Y. Acad. Sci. 407,398 damage rate frw (1983). 3. Carcinogen Identiation Policy: A Statement $Scirnn ' from normal ml as a Basis $Policy (Department of Health Service, dogenous coml State of California, Sacramento, CA, 1982); dative damage "Guidelines for carcinogen risk -sment;" Fd. R q . 51,33992(24 Scptcmer 1986);Appmchrr b logical findings antioxidants frc remain. In light of the uncertainty about mecha- pears to be a n types of cancer, nisms and human dose-response, the as- other degenerat sumption of low-dose linearity for carcin- aging (5). Oxic ogens continues to be a reasonable one (9). large amounts It is consistent with the fact that humans oxygen radicals are exposed to multiple carcinogens, capa- eration, that is, I ble of additive and even multiplicative ef- Antioxidants p fects. It is also a prudent assumption given the striking interindividual variation in the biologic response to carcinogens. Recent studies show an impressive range of human response to xenobiotics in terms of the effects. We thi such as folate (I cies are major cc The natural u of chemicals tha activation and detoxification of carcino- in both weight gens, covalent binding to DNA, and DNA repair (10). Such findings argue against the concept of a single population threshold for a carcinogen. The large and growing burden of cancer in the United-States (now at 500,000 cancer. deaths per year) vividly demonstrates the need for prevention. Prevention has always been the guiding principle in toxicology and public health policy and now merits increas- ing emphasis (II), Prevention means not grams of natural milligrams of ch I7. I. B. Weinstein, Cdt&rRes. 48,4135 (1988);J.C Rarrett, in Mechanism ofEnvironmenfal Carrinogrnr. sir, J. C. Rarrett, FA. (CRC Prcss, Boca Raton, FI. 1987)vol. 2, pp. 117-126;S. H. Rcynolds et PI., compared with pesticide water pollutants .%&e 237,ljb9 (1987);P.A. Ceru&, ibid. 227, 375 (1985);K.Frenkel and K.C h m , Carrinogm. esis 8,455 (1987). chemicals is not 8.J. C. Bamtt, Proceedinfs dfheCold Sprinf Harbor1 chemicals ar IMeeting on Origins $Human Cancer (GB Spring1 Harbor Laboratory, Cold Spring Harbor, NY, in press). dose, not just b! About half tl 9.b. Hod, N. Kaplin, M. Anderson, Sriencc 219.1 chronically in ral only addressing those cancer risks already established as "major" contributors to the disease burden (such as smoking) and researching new potential "major" risks, as Ames and Gold suggest, but also reducing 10. C. C.Harris, Envimn. Health Persp. 62,185(1985); K. Vahakangasand 0.Pclkonen, in GeneticEpidemi. 3). These tests c ology o/Cancer, H . T. Lynch and T. Hirayama, Ed$` control for the (CRC PmS, Baca Raton, FL, 1989);F. Pcreranal.. Enuiron. Healfh Persp., in press. a high _. current involuntary exposures to identified industrial carcinogens (12). Indeed, on the basis of a highly simplified (called "HEW") system for ranking carcinogens developed by Ames and Gold, the estimated range of risk for "natural" and man-made carcinogens is comparable (13). While it is tempting to ations 4Carcinogenuity:An Updafing4ZARCMono- eve.,, natural ,-hc g r a p h (InternationalAgency for Racarch on Can. cer, Lyon, France, 1988),vo~s. 1-42, supp~.71. less. In roasted ( IIndeed, the animal data often predated human cpi- tested. 16 were demiologic evidence. Had privcntivc action be& taken on the basis of the preexisting cxwrimcntd one A p ofcoffet . OS2 "1011 `!I: