A Comparative Study on the SWOG Two-Stage Design Extension to Stop Early for Efficacy in Single Arm Phase II Trials

Abstract

The research landscape in the era of personalized medicine is shifting its focus to smaller trials due to increasing reliance on predictive genomic biomarkers to guide more targeted therapies. This is especially true for Phase II single arm trials conducted in rare cancer settings that are challenging in terms of selecting, recruiting, and treating sufficient numbers of patients in a reasonable time-frame. To monitor ethical/safety concerns of the patients, interim analyses are needed to allow early termination of the trial in the case of treatment futility. However, with the availability of more promising targeted and immuno-oncology agents, these nondefinitive Phase II trials should also be stopped for early indication of efficacy to accelerate drug development. In this article, we extend the commonly used methodology within the cancer clinical trials group SWOG for a single-arm two-stage design with early stopping for futility to accommodate early stopping for efficacy. We then compare the operating characteristics of the proposed design with earlier work by Fleming, and rationalize the validity of the approach in more prevalent disease settings.