Confirmatory Seamless Phase II/III Clinical Trials with Hypotheses Selection at Interim: General Concepts

摘要: Traditional drug development consists of a sequence independent trials organized in different phases. Full typically involves (i) learning phase II trial and (ii) one or two confirmatory III trial(s). For example, the several doses new compound might be compared to control and/or placebo with goal deciding whether stop continue and, latter case, selecting "best" carry forward into phase. The are then conducted as stand-alone studies, not incorporating their statistical analyses data collected previous Seamless II/III designs aimed at interweaving phases full by combining them single, uninterrupted study stages. In dose-finding example above, (or more) dose(s) selected after first stage based on available interim, observed further second stage. final analysis includes patients from both stages is performed such that overall type I error rate controlled prespecified level regardless dose selection rule used interim. adequacy interim obviously critical step for success seamless trial. this paper we focus description flexible test procedures allowing adaptively hypotheses thus combination confirming single We review background, introduce compare power study. subsequent (Schmidli et al., 2006) give applications our daily practice discuss related implementation issues conducting adaptive designs.