Prevention of environmentally related diseases.

The National Institute of Environmental Health Sciences is dedicated to reducing human diseases through a multidisciplinary approach focused on prevention of environmentally related illnesses. A central question is how best to focus our research and other activities to have the greatest impact on human health. Current research activities at the NIEHS are directed toward these goals, but improvements in certain areas, together with an enhanced emphasis on environmentally relevant diseases, can greatly expedite the achievement of our public health mission. A successful program for prevention of environmentally related diseases should have three components that must be integrative and interactive: 1) identification ofenvironmental hazards, 2) elucidation of mechanisms of environmental agents and environmentally related diseases, and 3) development and refinement of risk assessment methodologies. The National Toxicology Program has unquestionably the world's premier program for carcinogen identification. The NTP bioassay should continue to be the cornerstone of the NIEHS program of prevention of environmentally related diseases by identifying agents that have the potential to cause human cancer and other chronic disorders. A question exists, however, about whether the program identifies the major environmental causes of human cancers. It is imperative to ask whether the strategies used to select chemicals for NTP testing are the most relevant and have the largest possible impact on the identification of causes of human cancer and other environmentally related diseases. It would be interesting (if it were possible) to estimate the total number of potentially preventable human cancers attributable to the carcinogens so far identified by the NTP; the number could be a very small percentage of all cancers. Therefore, new principles should be adopted to improve the selection of chemicals and testing procedures. Some examples of these principles are as follows. * Exploit what is known. We should take advantage of the advances in our knowledge about mechanisms of environmentally related diseases that may allow more chemicals to be evaluated and predicted (tested) by alternative means. For example, the role ofmutagenesis in carcinogenesis is clearly established; thus, there is little need to continue to study mutagens in two-year carcinogenesis bioassays for carcinogen identification. Even though some chemical mutagens are inactive in the standard bioassay, these are potentially carcinogenic in other contexts, and perhaps mutagens per se should receive greater attention from regulatory agencies. Internationally accepted procedures for mutagen evaluation in vitro and in vivo should be more uniformly and widely adopted, and chemicals that are clearly positive should be considered reasonably anticipated to be carcinogenic to humans. This would allow evaluation of a larger number of chemicals by alternative, short-term tests. Alternative approaches to doseresponse relationships for risk assessment could be based on mutagenic or preneoplastic endpoints. In addition, evaluation of chemicals by structure-activity relationships is another alternative to the two-year bioassay (1). For those dasses of chemicals where our knowledge is limited, bioassays should be performed to test hypotheses related to structure-activity relationships. Furthermore, where our knowledge is substantial, bioassays are not necessary and certain chemicals or dasses of chemicals should be assumed to be carcinogenic to humans (e.g., nitrosamines, anthraquinones, and benzidine congeners) (2. The recent success ofTennant and co-workers (1) in predicting the outcomes of rodent carcinogenicity tests demonstrates that alternative approaches to carcinogen identification are feasible. Although this approach is not possible with all chemicals, there are certain chemicals and classes of chemicals for which predictive toxicology is possible based on knowledge ofmechanisms ofaction. Additional alternatives to the two-year bioassay indude shortand mid-term in vivo assays (3), transgenic animals (4), alternative species including aquatic animals (5), and cell culture models (6). These are valuable for evaluation of chemicals for noncancer as well as carcinogenic effects, for identifying possible mechanisms, and for establishing priorities in the selection of chemicals that should be tested in two-year bioassays. Another practical consideration would be simply to reduce the size of the two-year studies by using only male rats and female mice and to adopt modified protocols (2,1). This has been shown, in a retrospective evaluation of standard twospecies, both-sex studies, to have "correctly" identified 96% of the "positive" and "negative" studies (2,7,8). Theoretically, this would allow evaluation of nearly twice as many chemicals with the same resources. Further, expanded bioassays of transgenerational carcinogenesis may be warranted (19). The NTP bioassay has identified several potent rodent carcinogens (2,10). It is important to identify human cohorts exposed to these chemicals for epidemiological studies to determine if there is evidence of their carcinogenicity in humans. * Use multidisciplinary approaches to focus on important causes of human cancers based on epidemiological clues. Hopes for prevention of human cancer are based on the high percentage of cancers that are environmentally related. Yet these hopes are not being realized. Industrial chemicals are likely to contribute to fewer human cancers than several other causes, including tobacco, diet, alcoholic beverages, hormones, UV light, and viruses. These also represent preventable causes ofhuman cancer, and a program of prevention should focus on all environmental agents. Of course, human cancers probably result from interactions between multiple etiological factors. Diet appears to represent a major environmental determinant of human cancer, but the scientific knowledge of what dietary factors are important and how they influence carcinogenesis is still rudimentary (11,12). The NTP/NIEHS could make a major impact in this important area of cancer prevention. The role of diet in the NTP bioassay should continue to be addressed (13). Evaluation of dietary factors for carcinogenic, co-carcinogenic, and anticarcinogenic activity, coupled with studies of mechanisms and epidemiology, would have a significant scientific and public health impact. For example, there is a growing body of evidence that consumption of fresh fruit and vegetables is associated with a decreased risk of cancer, particularly for the digestive tract and hormone-related cancers (14). Advances in molecular biology of cancer could rapidly improve our ability to identify and understand dietary/nutritional factors in carcinogenesis. Another area that could benefit considerably from multidisciplinary approaches is the role of hormones in cancer (15). Our new initiative in receptor-mediated pathobiology may provide an example of how to identify environmental estrogens using an alternative approach (16). The incidence of non-Hodgkin's lymphoma is increasing faster than any other cancer in the United States (12). Increases in this cancer in agricultural communities suggest a pesticide, herbicide, or other agricultural exposure as a causative factor (18-20). The environmental cause(s) for this cancer should be better established. Again, interactive studies using animal models combined with epidemiological investigations and molecular biology research should be pursued. * Develop models for important human cancers. Relatively few carcinogens have been identified for important human cancers, including breast, ovarian, prostatic, and colon cancers (15). Development of better animal and cell culture models for these cancers could lead to establishment ofnew tests for environmental causes of these cancers. * Develop and define models for human predisposition and suscepti-