Sequential bioequivalence approaches for parallel designs.

Area of Science:

• Pharmacokinetics and Biopharmaceutics
• Clinical Trial Design
• Statistical Analysis in Drug Development

Background:

• Regulatory agencies in the EU, USA, and Canada permit two-stage designs for bioequivalence assessments.
• Crossover designs are commonly used but may be unsuitable for certain drug formulations.
• Evaluating alternative bioequivalence methodologies is crucial for pharmaceutical development.

Purpose of the Study:

• To assess the performance of two-stage bioequivalence designs adapted for parallel group trials.
• To investigate the impact of equal versus unequal variances on the reliability of these designs.
• To provide guidance on the application of parallel group designs in bioequivalence studies.

Main Methods:

• Adaptation of existing two-stage methods (Potvin et al.) for parallel group trial simulations.
• Simulation of trials considering both equal and unequal variances between treatment groups.
• Evaluation of statistical performance, focusing on type I error rate control.

Main Results:

• Adapted Methods B and C demonstrated robust protection against type I error rate inflation across all simulated scenarios.
• The performance of the parallel design methods showed minimal sensitivity to the assumption of equal test and reference variances.
• This research presents the first evaluation of two-stage approaches specifically for parallel bioequivalence designs.

Conclusions:

• Two-stage bioequivalence evaluation using parallel designs is a statistically sound approach.
• These methods maintain reliability irrespective of variance equality assumptions, offering flexibility.
• The findings support the use of these parallel designs when crossover bioequivalence studies are not practical.