Methodology of phase I study in Japan.

This paper reviews the methodology of phase I study in Japan. Most leading institutions participated in a phase I study when the decision was made to proceed with clinical trials after detailed analysis of experimental results of a new antitumor agent. An initial dose in phase I study used to be calculated according to the dose comparable to the minimum effective dose, 1/100 of LD 50 (mg/kg) in rodents, but the dose recently has been estimated based on one-tenth of mouse LD 10 (mg/m2) or one-third TLD in sensitive large animals. Qualitative prediction through preclinical toxicology indicated better correlation with clinically observed toxicities in gastrointestinal and hematopoietic systems, but renal and hepatic toxicities were being overestimated, and neurologic and skin toxicities were being underestimated. A recent trend in the development of new antitumor agents has been to find a less toxic and more effective compound. This has made phase I study more complicated, because metabolites produced in the midway of metabolic pathway might induce unprediced toxicities. Several approaches to conduct a phase I study more efficiently are discussed.