Harmonized Estimation of Subgroup-Specific Treatment Effects in Randomized Trials: The Use of External Control Data

Area of Science:

• Biostatistics
• Clinical Trial Design
• Pharmacometrics

Background:

• Subgroup analyses are crucial in precision medicine drug development, guiding trial design and treatment approval.
• Small sample sizes in early-stage trials create uncertainty in subgroup-specific treatment effect estimates.

Purpose of the Study:

• To explore the use of external control (EC) data to augment subgroup analyses in randomized controlled trials (RCTs).
• To define and discuss "harmonized estimators" for subpopulation-specific treatment effects using EC data.

Main Methods:

• Leveraging EC data to augment subgroup analyses within RCTs.
• Developing "harmonized estimators" that modify existing subgroup estimates (e.g., from linear regression) to align with overall RCT results.
• Ensuring the weighted average of harmonized subgroup estimates matches the RCT-only overall effect estimate.

Main Results:

• The proposed harmonized estimators provide a method to incorporate EC data into RCT subgroup analyses.
• Analytic results, simulations, and a case study in oncology demonstrate the performance of these estimators.
• The weighted average of harmonized subgroup estimates is coherent with the overall RCT effect estimate.

Conclusions:

• Harmonized estimators offer a robust approach to address small sample size limitations in RCT subgroup analyses by incorporating external control data.
• This method enhances the reliability of subpopulation-specific treatment effect estimates, supporting precision medicine initiatives.
• The approach is applicable to various statistical methods and has been validated through simulations and a real-world oncology case study.