Doses in rodent cancer studies: sorting fact from fiction.

The belief that rodent cancer bioassays predict for human cancers is a fundamental public health precept based on sound biological principles. Nonetheless, it is appropriate to periodically debate this point as scientific understanding of cancer causation advances. This presentation addresses one of the many factors that determines the predictive value of rodent tumor bioassay results for human health. This is the issue of dose. Examination of several recent National Toxicology Program (NTP) studies demonstrates that the applied dose often far overestimates the actual effective dose, or maximum blood concentration attained in a rodent, when compared with similar relationships in humans. Further examination of the NTP database on rodent toxicity and carcinogenicity studies revealed summary information on factors that were pivotal in prechronic studies for selecting doses for chronic studies. Contrary to popular belief, target organ toxicity was a determining factor in only about half of the studies. The typically minimal nature of the lesions which limit doses for chronic studies is described for several common target sites. Taken together, these facts paint a far different picture than the common public perception of the "massive" doses used in chronic rodent studies and suggest that, in some cases, dose limitations are actually so severe as to limit the sensitivity of a chronic bioassay to detect a carcinogenic effect.