Pesticides Aaron Blair, Ph.D.*	Concern over long-term hazardous effects of pesticides has been a major force behind the environmental movement in the United States, raising serious questions about potential carcinogenic and other toxic effects of these chemicals. The potential for voluntary exposures from agricultural and home use and for involuntary exposures from food residues, water contamination, community and neighbor spraying, or military service has precipitated a major scientific and political controversy. The evaluation of pesticides for carcinogenicity, which began with animal bioassays in the 1960s (Innes et al., 1969), is now pursued in earnest through a battery of toxicologic and epidemiologic investigations. The National Toxicology Program tested some 50 pesticides in animals, usually using males and females in two species (Ashby and Tennant, 1988). Of these, 17 were positive for carcinogenicity in at least two of the sex/species groups: chlordane, chlordecone, chlorobenzilate, dieldrin, heptachlor, toxaphene, dichlorvos, tetrachlorvinphos, aminotriazole, nitrofen, ozadiazon sulfallate, captan, chlorthalonil, dibromochloropropane, dichloropropane, ethylene dibromide, and ethylene oxide (Blair et al., 1990). An additional six pesticides were positive in one sex in one species: aldrin, dicofol, piperonyl sulphoxide, chloramben, monuron, and trifluralin. The International Agency for Research on Cancer (1987, 1991) has concluded that several pesticides should be considered as probable human carcinogens: amitrole, arsenic, chlordane, chlorophenols, chlorophenoxy herbicides, DDT, 1,2-dibromochloropropane, ethylene dibromide, ethylene oxide, Mirex, and toxaphene. Studies of human populations exposed to pesticides are also available (Blair et al., 1990). Many of these studies evaluated cancer risks from use of pesticides in general, without attempting to focus on specific chemicals. For example, excesses of lung cancer have been observed in some studies of agricultural (Barthel, 1981) and urban applicators (Blair et al., 1983, MacMahon, 1988), but these excesses could not be related to individual pesticides. Surveys in a number of developed countries have noted excesses for several cancers among farmers, including leukemia, non-Hodgkin's lymphoma, multiple myeloma, soft-tissue sarcoma, and cancers of the skin, lip, stomach, brain, and prostate (Blair et al., 1992). Farmers represent an occupation that may have frequent contact with a variety of pesticides, which underscores the need for additional investigations of agricultural populations. A few recent epidemiologic studies have attempted to evaluate specific pesticides. Lung cancer has been associated with blood levels of DDT among residents of South Carolina (Austin et al., 1989), and pancreatic cancer risk was excessive among workers employed in the manufacture of DDT (Garabrant et al., 1992). Levels of DDT and its metabolites in blood and adipose tissue have been associated with breast cancer in recent studies in the United States (Falck et al., 1992; Wolff et al., 1993; Krieger et al., 1994). These epidemiologic findings, coupled with experimental evidence of the carcinogenicity of DDT in laboratory animals, suggest that this pesticide may present a cancer hazard to humans. Increased risks for several lymphatic and hematopoietic cancers have been found among individuals exposed to insecticides in the United States (Hoar et al., 1986; Boffetta et al., 1989; Cantor et al., 1992; Zahm et al., 1990; Brown et al., 1990), Sweden (Flodin et al., 1988), and Italy (Corrao et al., 1989). Several groups of pesticides may be involved, including organochlorines, organophosphates, and carbamates, but additional research is necessary to determine definitively which--if any--are human carcinogens. Epidemiologic investigations have also linked herbicides with some cancers. Phenoxyacetic acid herbicides have been associated with non-Hodgkin's lymphoma in Sweden (Hardell et al., 1981; Persson et al., 1989), Canada (Wigle et al., 1990), and the United States (Hoar et al., 1986; Zahm et al., 1990) and with soft-tissue sarcomas in Sweden (Hardell et al., 1979; Eriksson et al., 1981), Denmark (Lynge, 1985), and Italy (Vineis et al., 1987). Studies in New Zealand, however, did not find such associations (Smith and Pearce, 1986; Pearce, 1989). Use of phenoxyacetic acid herbicides on lawns was also associated with malignant lymphoma in dogs (Hayes et al., 1991). There is no clear evidence from bioassays, however, that phenoxyacetic acid herbicides cause cancer in animals. Concerns have recently been raised about triazine herbicides, another widely used group of weed killers. Some triazine herbicides cause breast cancer in rodents (IARC, 1992) and have been associated with ovarian cancer among Italian women engaged in agricultural activities (Donna et al., 1989). A study of manufacturers of phenoxyacetic acid herbicides exposed to dioxin showed they experienced excesses for several cancers, including lung cancer, and soft-tissue sarcoma (Fingerhut et al., 1991). Experimental data indicate that several pesticides can cause cancer in animals, thus raising concerns about human exposures. Epidemiologic studies suggest that occupational exposure to some pesticides may present a carcinogenic hazard. Although evidence that pesticides cause cancer in humans is not conclusive, a prudent course would be to minimize exposure through use of protective practices and appropriate personal hygiene. Research is continuing to clarify cancer risks from specific pesticides and to determine mechanisms of action.
continue

* From the Occupational Studies Section, Division of Cancer Etiology, National Cancer Institute, Bethesda, Maryland