Consistent with the Food and Drug Administration’s (FDA) commitment to streamlining the prescription drug and biologics approval processes by adopting innovative approaches to clinical trial design, the agency recently published its draft guidance for industry, “Use of Bayesian Methodology in Clinical Trials of Drug and Biological Products.” The draft guidance signals a significant shift in regulatory thinking about statistical methods in drug and biologic development. Should this nonbinding draft guidance become final, it would align FDA’s approach with other global regulatory frameworks, including the International Council for Harmonisation (ICH) guidelines. This would prevent redundancy in industry efforts to get products approved internationally. Such an approach may be especially useful for global innovators submitting safety and efficacy data to FDA related to treatments for unmet medical needs and rare and pediatric diseases or when conducting large-scale clinical trials that pose ethical concerns. This flexibility is designed to help sponsors leverage relevant data more efficiently, while maintaining FDA’s well-established standards for evaluating safety and effectiveness.
Traditionally, FDA has predominantly required data collected during a clinical trial to support safety and efficacy without any outside information. In contrast, the Bayesian approach permits the borrowing or combining of relevant clinical data to determine the probability that a treatment is safe and/or efficacious. For example, in limited instances, FDA has permitted innovators to formally incorporate data from a previous Phase 2 placebo-controlled study into a Phase 3 study, and it has also allowed incorporation of borrowed data from adult clinical trials to support efficacy in the pediatric context.
The draft guidance details recommendations for how Bayesian methods may be incorporated in various aspects of clinical trial design and analysis, including:
- Using Bayesian probability estimates to draw conclusions about safety and efficacy
- Enabling interim decisions like early stopping for efficacy or futility
- Incorporating data from previous studies, real-world evidence (RWE), historical controls, and other scientifically justified sources
The use of these statistical methods have the potential to enhance trial efficiency and shorten development timelines. This means that sponsors may reach reliable conclusions with fewer patients, especially as it relates to trials involving rare diseases, pediatric populations, or small subgroups where recruitment is challenging. Incorporation of Bayesian methodology in FDA-regulated clinical trials may also reduce the need for large concurrent control groups, and because implementation of the Bayesian framework can facilitate early stopping for futility or efficacy, innovators may be able to be more nimble and can direct resources toward more-promising candidates and away from ineffective ones. Permitting statistically efficient and flexible designs supports sponsors’ efforts to pursue adaptive platform trials, use of RWE, and other modern evidence strategies, all of which are becoming central to 21st-century clinical development.
Key actions for life sciences industry clients include the following:
- Review the draft guidance and consider submitting comments before the deadline.
- Evaluate where Bayesian designs could enhance your ongoing and future programs.
- Engage early with FDA in the drug and biologic development process to discuss Bayesian proposals.
As FDA’s interpretation of regulatory science evolves to align with ICH guidelines and other global regulatory frameworks, Bayesian methods may become an increasingly important tool in the evidence gathering toolkit — enabling sponsors to bring safe and effective drugs and biologics to patients faster and with greater efficiency.