Independent drug action and its statistical implications for development of combination therapies

Researchers have long sought to find combinations of cancer drugs that might achieve synergy. However, while observed in some preclinical tumor models, synergistic effects are rarely seen in clinical trials. In fact, growing evidence in clinical trial data shows that the treatment effect of most approved combination therapies can be largely explained by the independent drug action model at the patient level. Previous statistical research on drug combinations mainly centered on experimental designs for dose-finding followed by measure of combination efficacy. In this paper, we introduce the independent drug action model to those working in late stage clinical development, propose a new approach to predict the progression-free survival of combination therapies, and discuss its statistical implications for trial design and monitoring. The discussion is enriched with real data examples.