Carcinogen testing. Fact and fallacy

In the absence of human information on the carcinogenicity of chemical substances, one must rely primarily on information from long-term animal testing. Although far from perfect, animal studies seem to be reasonable predictors of the human experience, both qualitatively and quantitatively. Short-term tests for genotoxicity may be helpful for establishing priorities for chemical testing, but they are not as strong indicators of potential carcinogenicity as had been previously thought. New directions in toxicologic research hold the promise for scientists being able to perform more reasoned assessments of carcinogenic risk.