A Phase II Design for Combination Codevelopment Based on the Probability of Correct Selection

• Activity Number: 35
• Type: Contributed
• Date/Time: Sunday, July 29, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Xinyu Tang*+ and Xinyu Tang
• Companies: University of Arkansas for Medical Sciences and University of Arkansas for Medical Sciences
• Keywords: Phase II design ; Investigational drugs ; Factorial design ; Time-to-event endpoint

Abstract

The United States Food and Drug Administration (FDA) (2010) has recently issued a draft guidance entitled "Guidance for industry - codevelopment of two or more unmarketed investigational drugs for use in combination." In the case where each of the two investigational drugs has some activity and both can be given individually, the guidance recommends that a four-arm Phase II factorial design should be used to compare the combination of two investigational drugs to each drug alone and to placebo or standard of care. The aim of the design is to establish contribution of each component in Phase II without exposing a large number of patients to relatively ineffective treatments, enabling a simple two-arm randomized Phase III trial to be conducted. We used Green's criteria (Green 2006) for selecting the best treatment in a factorial design using a time-to-event endpoint as the basis of a Phase II strategy relying on the probability of correct selection concept applied to a progression-free survival endpoint. We claim that selection of the combination as the best agent using Green's criteria may provide evidence of component contribution in Phase II trials.