A Dose-Finding Design with Time-to-Toxicity Endpoint

• Activity Number: 619
• Type: Contributed
• Date/Time: Thursday, August 2, 2012 : 8:30 AM to 10:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Rui Qin*+ and Xian-Jin Xie
• Companies: Mayo Clinic and The University of Texas Southwestern Medical Center at Dallas
• Keywords: dose finding ; late-onset toxicity ; continual reassessment method ; parametric survival distribution

Abstract

Testing novel regimen in radiation oncology usually requires a much longer follow-up time for toxicity evaluation. Conventional cohort designs including continuous reassessment method do not work well in this type of trials, because enforcing repeated temporary closure of cohort for toxicity evaluation not only disrupts the conduct of the trial, but also makes the trial duration prohibitively long. We have proposed a dose-finding design (survCRM) that can accommodate longer follow-up time for late-onset toxicities. This new dose-finding design treats time-to-toxicity as primary endpoint and models it using widely applicable parametric survival distribution. Simulations are conducted to study the operating characteristics under various common trial scenarios. Comparisons are also made with respect to time-to-event continuous reassessment method (TITE-CRM). The survCRM design performs as well as TITE-CRM design and warrants further evaluation in more real trial settings where longer follow-up time is required.