[Toxicology of reproduction: predictive value of experimental models. Round Table No 1 at Giens XIII]

The carrying out of clinical trials with a view to the marketing of drugs for human use is directly related to results of some animal studies. This workshop was devoted to evaluation of the quality and interest of these experimental models in reproductive toxicology. The predictive ability of preclinical trials to make extrapolations from animals to man decreases from foetotoxic to tetratogenic risks respectively and from the effects on fertility in both sexes to postnatal risks. As a result of this workshop, we propose the following improvements: (1) standardization and generalization of fertility test evaluations, especially the spermogram, in order to improve animal and human correlations; (2) development of knowledge and standardization of the follow up of the oestral cycle; (3) improvement of standardization, harmonization and diffusion of postnatal tests that prove relevant in animals; (4) increase in initiatives aimed at better mutual understanding of all drug partners; (5) creation of registers for new drugs, as soon as possible during clinical trials, to study their effects on the whole reproductive process; (6) recommendations for the creation of guidelines for International Conference on Harmonisation (ICH) to enable classification of observed effects in experimental models. This could lead to specific (potentially for each phase of the reproductive cycle) guidelines, precautions for use and/or contraindications which are listed in the summary of product characteristics.