The REFRACT trial: implementation of Bayesian power priors in a randomised, sequential phase II adaptive platform...

Area of Science:

• Hematology
• Clinical Trial Design
• Oncology

Background:

• Follicular lymphoma (FL) is a common non-Hodgkin lymphoma with relapsed or refractory (rrFL) disease urgently needing new therapies.
• Traditional clinical trials for rrFL require large patient cohorts, delaying the evaluation of novel treatments.
• The REFRACT trial addresses this by employing an adaptive, sequential design to rapidly assess multiple novel therapies.

Purpose of the Study:

• To rapidly evaluate multiple novel therapeutic agents for relapsed or refractory follicular lymphoma (rrFL).
• To utilize a prospective, adaptive, and sequentially randomized clinical trial design.
• To compare novel therapies against investigator choice standard therapy (ICT) using minimal patient numbers.

Main Methods:

• The REFRACT trial employs a Bayesian power priors approach for efficient data sharing across control arms.
• Adaptive randomization ratios are implemented, starting at 1:1 and shifting to 1:4 in favor of experimental arms in later rounds.
• Control arm data from previous rounds are incorporated, weighted at 75%, to enhance statistical power.

Main Results:

• The trial design, validated through simulations, includes three sequential treatment rounds.
• Each round assesses a novel experimental arm against a control arm, with primary outcome being complete metabolic response (CMR) at 24 weeks.
• Bayesian power priors allow earlier control data to inform current round operating characteristics, improving efficiency.

Conclusions:

• The adaptive design and Bayesian power priors enable sequential evaluation of three novel treatments for rrFL.
• This innovative approach accelerates the assessment of much-needed treatment options for follicular lymphoma.
• The REFRACT trial commenced recruitment in July 2023 (EudraCT: 2022-000677-75; NCT05848765).