Optimal Design for A Series of Clinical Trials for New Agents

Abstract

In exploratory vaccination studies for cancer therapeutics, toxicity is usually less a concern than the ability of the vaccine in inducing immunological responses which may lead to therapeutic effect. In practice, there may be many potential vaccines constructed one after one, and the resources are limited. Therefore, the goal is to identify a promising vaccine using as few patients as possible in total. A new approach is introduced for determining the appropriate sample sizes for a series of screening trials to identify promising new therapeutic agents. The formulation of the problem is motivated by recognition of the fact that screening of new agents is a continuing process. We fix the composite error rates and optimize the individual sample sizes to minimize the time to identify a promising agent, using an empirical Bayes formulation. A simple stopping rule is available to allow early termination of a trial without changes in Type I error. This approach is further improved to a two-stage design which offers extra savings. When the procedure was applied to the historical data of exploratory vaccination trials at Memorial Sloan-Kettering Cancer Center, New York, the method demonstrates that relatively small individual screening trials are optimal in this setting. The reliability of the results is evaluated using bootstrapping techniques.