Advances in science and technology are creating unprecedented opportunities to improve patient outcomes through therapies targeted to a highly specific patient population. However, translating our scientific progress to rapidly deliver more effective treatments to the right patients requires updating the clinical and regulatory frameworks used to evaluate new biomedical products. One promising and innovative approach is using adaptive designs that allow for evidence accrued throughout the course of an ongoing clinical trial to shape the trial. These types of trial designs are already being used with greater frequency in exploratory trials, but there is an even more significant opportunity to incorporate adaptive features into phase III confirmatory trials where the need for cost savings, efficiencies, and improved regulatory success rates is the greatest. As with any new approach, adaptive trials also introduce a number of statistical and operational considerations that have so far impeded the use of complex adaptive designs in phase III efficacy trials necessary for regulatory approval.
On March 27th, in partnership with the U.S. Food and Drug Administration, the Engelberg Center for Health Care Reform at Brookings convened an expert workshop, “Pioneering Statistical Approaches to Accelerate Drug Development through Adaptive Trial Designs.” The workshop focused on addressing the statistical and operational challenges adaptive designs pose for ensuring adequate and well-controlled phase III trials. A number of theoretical designs developed by workshop participants served to stimulate discussion around these issues. Participants included experts from industry, contract research organizations (CROs), and government as well as statistical consultants, clinicians, and academics. Drawing on their extensive experience, experts critically evaluated the relative advantages of the proposed designs and shared their insights into current regulatory, statistical, and operational challenges facing adaptive clinical trials, as well as potential solutions.