Testing superiority at interim analyses in a non-inferiority trial.

Area of Science:

• Biostatistics
• Clinical Trial Design
• Pharmaceutical Research

Background:

• The pharmaceutical industry is shifting towards developing differentiated medical products (first-in-class, best-in-class, bio-betters) aiming for superior efficacy over standard care.
• While non-inferiority trial designs are appropriate in certain contexts, there's a growing interest in demonstrating superiority even within these frameworks.

Purpose of the Study:

• To propose a novel group sequential design that enables superiority testing at interim analyses within a non-inferiority clinical trial.
• To evaluate the impact of interim superiority tests on the statistical properties of the primary non-inferiority hypothesis at the final analysis.

Main Methods:

• The proposed design incorporates conventional group sequential methods for interim superiority testing.
• If superiority is not established, the trial proceeds to a final non-inferiority analysis without alpha adjustment for the primary hypothesis.
• The impact on the false positive rate and statistical power of the non-inferiority test is assessed using Brownian motion boundary-crossing properties and numerical evaluations.

Main Results:

• The study demonstrates that interim superiority tests have a negligible impact on the false positive rate and statistical power of the final non-inferiority test.
• This effect remains minimal even with substantial alpha spending on interim superiority tests, particularly with O'Brien-Fleming type alpha-spending functions.
• Numerical evaluations confirm these findings across various alpha-spending functions.

Conclusions:

• The proposed group sequential design offers a valid method to test for superiority during interim stages of a non-inferiority trial.
• This design facilitates early stopping for superior efficacy without compromising the validity of the subsequent non-inferiority conclusion.
• The methodology provides a robust framework for optimizing clinical trial efficiency and decision-making in drug development.