Can therapeutic efficacy be predicted from phase I data?

The aim of phase I studies per se is to explore the tolerance of new compounds which have demonstrated a certain level of activity in animals at sufficiently low non-toxic doses. In most cases, these studies are conducted in 2 steps in a limited number of healthy male volunteers: a single rising dose study followed by a repeated dose study in which the pharmacokinetic features of the drug are explored. In such a context, it would be quite presumptuous to ascertain the therapeutic efficacy of a drug from those initial human studies. At best, these trials can provide pharmacological and/or biological indications which are related to some degree to the expected efficacy of the drug: hypnotics, anticancer drugs, antibiotics, platelet antiaggregants, beta-blockers, etc. In these examples it is recommended to design controlled study protocols so as to better investigate these potentially interesting signs of activity. In the majority of cases, however (psychoactive drugs, analgesic, anti-inflammatory drugs, gastro-intestinal compounds), phase I studies will unfortunately not provide much information regarding the expected therapeutic activity.