OBJECTIVES: Randomized clinical trials are crucial for assessing new treatments. Bayesian borrowing (BB) utilized external data to increase power for within trial comparisons with modest sample sizes, comparisons beyond the trial, and to allow for a reduced control arm cohort for ethical and patient availability reasons. The acceptability of BB approaches from a regulatory and HTA perspective is evolving. Anticipatedly, the FDA will release draft guidance on employing Bayesian methods in clinical trials for drugs and biologics by the end of 2025. This targeted literature review provides an overview and examples of the implementation of BB in clinical trials, highlighting its potential benefits, methodological advancements, and challenges.

METHODS: The review searched MEDLINE and Embase in the past five years, and identified peer-reviewed articles, industry reports, and regulatory and HTA guidelines. Sources were selected based on relevance and methodological rigor.

RESULTS: Although limited, there has been increasing adoption of BB in clinical trials in recent years. BB is particularly valuable in pediatric trials or rare disease research where patient populations are limited. For example, the FDA recommended that GlaxoSmithKline (GSK) use BB to borrow adult trial data into a pediatric trial analysis, and these results were subsequently accepted. In another example, FDA approved heart failure-targeted Optimizer Smart System, following a trial that utilized BB to include external data. In 2020, NICE recommended Bayesian hierarchical models (BHMs), in the technical appraisal for larotrectinib in NTRK-fusion positive solid tumors. Regulatory and HTA agencies have increasingly recognized BB and its potential benefits.

CONCLUSIONS: BB has the potential to significantly improve the precision and robustness of treatment effect assessments. This review underscores the necessity for continued research in Bayesian methodologies to address existing limitations and expand their applicability in health research.