Multistage designs for phase II clinical trials: statistical issues in cancer research

Abstract

The main objective of phase II clinical trials is to estimate treatment efficacy on a relatively small number of patients in order to decide whether the treatment ought to be studied in large-scale comparative trials. They play a key role in the drug development process, since the results determine whether or not to proceed to phase III trials. Multistage designs for phase II clinical trials proposed by Gehan, Fleming, Simon and Ensign are described and compared. Gehan's and Simon's designs have two stages, Fleming's designs can have two or more stages, and Ensign's three-stage design combines the first stage of Gehan with the two stages of Simon. Phase II clinical trial protocols and reports should include a description of the design selected with a justification for the particular choice. The present practice is very far from this ideal.