With Keaven M. Anderson, Ph.D., Merck Research Laboratories

11:00AM-12:00PM

Location: Medical School Office Building (MSOB) x303

Abstract: Several issues have arisen in designing clinical trials for immunotherapies to maximize the chance of success. Progression free survival benefit has not been a great surrogate for an overall survival benefit. Immunotherapy may not provide efficacy early either due to a delay in treatment effect or to a possibly unidentified subgroup where treatment is less effective. Thus, the trials may evaluate multiple endpoints in multiple populations at multiple points in time during the course of the trial. In addition to this large multiple testing issue, standard methods assuming proportional hazards are likely inefficient for detecting a treatment benefit. Despite all of these challenges, many trials have proven very successful. An FDA-initiated effort has considered alternatives to logrank tests. The current status of these efforts will be discussed.

About the Speaker: Dr. Keaven Anderson is a Distinguished Scientist working in Late Development Statistics at Merck Research Laboratories where he has performed in various roles since the end of 2003. Keaven is a fellow of the ASA. His primary research interests are in group sequential design, particularly for studies with time-to-event endpoints. Multiple testing and non-proportional hazards are current areas of interest. Keaven has developed the gsDesign R package and associated Web interface for group sequential design. His PhD in Mathematical Statistics is from Stanford University. Since graduating, he has been at the Harvard School of Public Health, the Framingham Heart Study and Centocor/J&J prior to coming to Merck.