[Evaluation studies of experimental carcinogenesis].

In many cases, epidemiological studies demonstrate carcinogenic risk of occupational exposures and industrial processes, but do not allow to assess carcinogenic potential of the various substances involved in exposures (speciation). Expert judgment combining human and animal data is needed for classifying and labelling of the pure substances. For decision aiding purpose, a method was established for reviewing critically experimental carcinogenesis: 1. Standardized toxicological profiles are established according to pure substances, species, strains and administration routes. 2. Methodological evaluation is made using a scheme based on 3 critical parameters: study duration which should extend to the life expectancy of the species under study; sample size, the statistical significance of negative experiments being ascertained depending on size of samples; survival rates of groups, since sufficient numbers of animals should be at risk during most of their life span. 3. Classification of the results for risk assessment, taking into account the relevance of the administration routes.