Conflicting views have been expressed frequently on assessments of human cancer risk of environmental agents based on animal carcinogenicity data; this is primarily because of uncertainties associated with extrapolations of toxicologic findings from studies in experimental animals to human circumstances. Underlying these uncertainties are issues related to how experiments are designed, how rigorously hypotheses are tested, and to what extent assertions extend beyond actual findings. National and international health agencies regard carcinogenicity findings in well-conducted experimental animal studies as evidence of potential carcinogenic risk to humans. Controversies arise when both positive and negative carcinogenicity data exist for a specific agent or when incomplete mechanistic data suggest a possible species difference in response. Issues of experimental design and evaluation that might contribute to disparate results are addressed in this article. To serve as reliable sources of data for the evaluation of the carcinogenic potential of environmental agents, experimental studies must include a) animal models that are sensitive to the end points under investigation; b) detailed characterization of the agent and the administered doses; c) challenging doses and durations of exposure (at least 2 years for rats and mice); d) sufficient numbers of animals per dose group to be capable of detecting a true effect; e) multiple dose groups to allow characterization of dose-response relationships, f) complete and peer-reviewed histopathologic evaluations; and g) pairwise comparisons and analyses of trends based on survival-adjusted tumor incidence. Pharmacokinetic models and mechanistic hypotheses may provide insights into the biological behavior of the agent; however, they must be adequately tested before being used to evaluate human cancer risk.
Keywords: dose selection; maximally tolerated dose; mode of action; rodent cancer bioassay; statistical power; tumor pathology.