A Review of Causal Inference for External Comparator Arm Studies

Area of Science:

• Clinical Trial Methodology
• Regulatory Science
• Health Economics and Outcomes Research

Background:

• Randomized controlled trials (RCTs) are the standard for drug efficacy but are increasingly supplemented by single-arm trials (SATs).
• The rise of precision medicine and rare disease treatments has led to more SATs, necessitating external data for comparison.
• External comparator arm (ECA) studies integrate SATs with external patient cohorts to provide context and improve interpretability.

Purpose of the Study:

• To provide a methodological overview of External Comparator Arm (ECA) studies.
• To outline emergent best practice recommendations for ECA study design and analysis.
• To identify key areas for future methodological research in ECA studies.

Main Methods:

• Review of existing scientific and regulatory guidance for ECA studies.
• Discussion of appropriate causal inference methods for analyzing ECA data.
• Consideration of sample size, estimands, and handling of differential data sources (covariates, endpoints, timing).

Main Results:

• ECA studies offer a viable approach to contextualize SAT results, particularly when RCTs are not feasible.
• Methodological challenges include managing differential baseline covariates, endpoint measurements, and timings between trial and external data.
• ECA studies can potentially address bias from unmeasured covariates if these are available in the SAT.

Conclusions:

• ECA studies are an increasingly important tool in drug development, especially for rare diseases and precision medicine.
• Adherence to methodological best practices and careful consideration of causal inference are crucial for robust ECA study results.
• Further research is needed to refine methods for handling data complexities and maximizing the utility of ECA studies.