Study Designs in Induction/Maintenance Trials

• Activity Number: 177
• Type: Contributed
• Date/Time: Monday, August 5, 2013 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Lixia Pei*+ and Kevin Liu
• Companies: Janssen Research & Development and Janssen Reseach & Development
• Keywords: maintenance ; induction ; two-stage randomization design ; factorial design ; inverse probability weight ; clinical trials

Abstract

With oncology drug development, trials with a maintenance treatment after an induction treatment are more on demand. In such trials, the comparison of interest includes both the induction treatment effect (i.e., new treatment vs. standard treatment) and maintenance treatment effect (i.e., maintenance treatment vs. no maintenance treatment or new maintenance treatment vs. standard maintenance treatment). Two types of trial designs can be used. One is the standard 2x2 factorial design, where patients are randomized upfront to a single randomization designs (SRDs), which is one of the four possible induction/maintenance combinations. Another is the two-stage randomization design (TSRD), where patients are first randomized to one of the induction treatments, and then the responders will be randomized to one of the maintenance treatments. For TSRD, inverse probability weighted (IPW) method needs to be used to remove biased comparison. In this paper, these two types of study designs will be compared. Simulation study will be conducted to compare the 2x2 factorial design with TSRD in terms of power.