Le gigantisme des essais cliniques est-il inévitable.

The size of randomised phase III trials has greatly increased over the last two decades. Several national or multi-national trials, including thousands of patients, are now ongoing, as it is the case in the field of adjuvant therapy for breast cancer patients. We analyse the different parameters conditioning the size of the trial: type I (alpha) and type II (beta) error rates, the expected response (survival rate, recurrence rate...) in the control arm, and the difference in treatment responses between the two arms where detection needs to be considered just as important. In addition we shall discuss the limitations of such very large trials, and the possible alternatives for today and for the future. Very large trials could be considered as a consequence of slow progress compared to what we expect from anti-cancer strategies today. [ABSTRACT FROM AUTHOR]