Adaptive Randomization Designs for Oncology Trials with Both Short-Term and Long-Term Outcomes

• Activity Number: 552
• Type: Topic Contributed
• Date/Time: Thursday, August 6, 2009 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Hao Liu*+ and Xuelin Huang
• Companies: Baylor College of Medicine and The University of Texas M.D. Anderson Cancer Center
• Keywords: adaptive randomization ; clinical trial design ; generalized Friedman's urn ; multiple endpoints ; randomized play-the-winner

Abstract

We propose an urn-based adaptive randomization for oncology trials to incorporate both short-term tumor response and long-term survival outcome. As a generalization of the randomized play-the-winner rule of Wei and Durham (1978), the design can update the randomization probability quickly using short-term response, but also using long-term survival outcome to favor the treatment arm with the definitive therapeutic benefit. Using the method of generalized Friedman's urn, we derive the explicit formula for the limiting distribution of the number of subjects assigned to each arm. We perform careful simulation study to evaluate the properties of the proposed adaptive design including the statistical power. The simulation results show that its performance is satisfactory with the desired properties, and the loss of power due to imbalance is minimal.