Indirect comparisons in the comparative efficacy and non-inferiority settings

Area of Science:

• Clinical trial design
• Regulatory science
• Biostatistics

Background:

• International Conference on Harmonization E10 guidelines address non-inferiority trials and comparative efficacy assessments.
• Indirect comparisons are frequently employed in both settings, analyzing relationships between treatments not directly tested against each other.

Purpose of the Study:

• To analyze the sample size requirements for indirect versus direct comparisons in clinical trials.
• To investigate the logical consistency between preference criteria for experimental treatments and current regulatory approval pathways based on indirect comparisons.

Main Methods:

• Comparative analysis of sample size calculations for direct and indirect superiority trials.
• Introduction of the "preservation of effect" concept to equate hypotheses in different trial settings.
• Logical assessment of existing regulatory approval criteria against a proposed preference criterion.

Main Results:

• Trials demonstrating superiority via indirect comparison necessitate larger sample sizes than direct comparisons.
• The hypothesis tested in both non-inferiority and comparative efficacy settings using indirect comparisons is fundamentally the same.
• A logical inconsistency exists where a preferred experimental treatment may not meet current regulatory approval criteria due to the burden of standard treatment uncertainty.

Conclusions:

• Current regulatory guidance for experimental treatment approval via indirect comparison may not align with reasonable preference criteria.
• The experimental treatment must overcome the uncertainty associated with the standard treatment's effect, creating potential approval discrepancies.
• An alternative approval criterion is proposed to resolve the identified logical inconsistency.