|
Abstract:
|
Phase I drug-combination trials primarily aim at identifying the maximum tolerated dose (MTD) combination(s). In contrast to single-agent phase I trials, drug-combination phase I trials face many challenges, especially in oncology where the sample size is often limited. One important challenge is the existence of multiple MTDs in the multi-dimensional dose searching space. As drug-drug interactions may cause varying treatment effects among multiple MTDs, it is of intrinsic interest to find multiple MTDs such that their efficacy can be further evaluated in subsequent phase II trials. Regardless of its importance, research in finding multiple MTDs is particularly limited. In this talk, we consider representative dual-agent drug-combination designs for finding multiple MTDs including the product of independent beta probabilities escalation (PIPE) design and the waterfall design. We briefly review the methodology, conduct a comprehensive simulation to compare their performance in terms of accuracy and safety. Last but not the least, we list and compare pros and cons of the selected designs, to arrive at a recommendation and promote the use of novel trial designs in real applications.
|
