Institution: CDER, FDA
Street: 5600, Fishers Lane
Phone: 301-594 5739
Fax: 301-594 6593
Authors: Rajeshwari Sridhara, Gang Chen, Kun He, George H. Chi
Title: Problem of Alpha Adjustment In Studiess With Multiple Primary Endpoints - Evaluation of Relationship Between Endpoints
Abstract: In randomized oncology trials comparing efficacy between two treatment groups, overall survival (OS) has been the gold standard primary efficacy endpoint of interest. However, drug manufacturers often argue that time to progression (TTP) should be used in evaluating drug efficacy comparisons. Their argument stems from the conjecture that patients who progress during the study are given different treatment options at progression and thus potentially confounding treatment differences with respect to overall survival after progression. On the other hand some studies have suggested that TTP may not be a good surrogate marker of overall survival. One possibility is that we can design trials with both OS and TTP as co-primary endpoints that are likely to be correlated. Thus in order to design such a trial evaluation of the relationship between OS and TTP becomes essential to adjust type I error for the two endpoints maintaining the overall type I error at a fixe! d level. In this presentation we try to explore a method for evaluating relationship between OS and TTP by considering OS = TTP + TPD, where TPD is time from progression to death and a random variable.