Advances in the design and reporting of conventional carcinogenicity tests.

It is generally acknowledged that animal experiments provide the most important and best reliable service of information for assessing the possible carcinogenic activity of chemicals. In the past, experimental procedure has mainly concentrated on the choice of species, improved animal husbandry, the modes of exposure, test conditions and pathological examination. Although experience gained thereby indicated the necessity for further scientific exploitation of bioassay models, the rather rigid stereotyped procedures--proposed for convenience--were generally accepted and adopted. Nevertheless, continuous improvements have been made in the design of carcinogenicity tests in rodents. Current developments in test design tend to be primarily concerned with the predictive significance of carcinogenicity studies. Modern protocols call for optimum facility operation as much as for computerized support of study monitoring and data evaluation. Of particular significance are the automated handling of autopsy and histopathological data and their proper statistical evaluation. By the economical use of modern technology, cost-saving protocols can be designed for the production of the data which are essential for the reliable assessment of risk.