A surrogate endpoint-based provisional approval causal roadmap, illustrated by vaccine development

Area of Science:

• Biostatistics
• Clinical Trial Design
• Rare Disease Research

Background:

• Many rare diseases lack approved preventive interventions, despite promising treatment candidates.
• Pivotal Phase 3 trials are challenging due to small patient populations and difficulty demonstrating direct benefit on target outcomes.
• Surrogate endpoints offer a potential pathway for provisional approval when direct evidence is limited.

Purpose of the Study:

• To describe a surrogate endpoint-based provisional approval causal roadmap.
• To outline a method for analyzing combined observational and Phase 3 trial data.
• To enable conservative inference on treatment effects for rare disease interventions.

Main Methods:

• Utilizing observational data to estimate a surrogate index (target outcome regression on surrogate endpoints and covariates).
• Employing Phase 3 trial data, despite being underpowered for direct assessment, to infer treatment effect (TE) based on the surrogate index.
• Incorporating two bias functions to ensure conservative inference: surrogate index imperfection and imperfect transportability.

Main Results:

• Development of plug-in and nonparametric efficient one-step estimators for TE.
• Inclusion of inferential procedures for robust statistical analysis.
• Evaluation of estimator performance through simulation studies.

Conclusions:

• The causal roadmap provides a framework for leveraging surrogate endpoints in rare disease intervention approval.
• This approach allows for conservative estimation of treatment effects by combining diverse data sources.
• Illustrated by Group B Streptococcus vaccine development, this methodology can accelerate the availability of crucial therapies.