Keywords: immuno-oncology, adaptive design, biomarker, interim analysis
The recent advance in Immuno-Oncology highlights the importance of identifying the patient population that truly benefit from this new class of therapies. Although the initial biomarker validations based on tumor responses in chemo-refractory patients provide some clues, the uncertainty on the utility of biomarkers often carries over to Phase III studies in the first line or adjuvant settings.
When more than one potential biomarker cut-points are considered in a Phase III trial, population selection based on the interim data is possible. In addition, specific low biomarker-expressing groups of concern may need to cross futility hurdle in order to continue in the trial. These practical needs call for an adaptive strategy that can eliminate potential non-performing subgroups while maintaining the statistical validity of the trial. In this roundtable discussion, we will explore the various regulatory and statistical issues involved in this type of adaptive trial with fellow statisticians who face similar challenges in their line of work.