Bioequivalence for highly variable drugs: regulatory agreements, disagreements, and harmonization

Area of Science:

• Pharmacokinetics and Drug Development
• Regulatory Science
• Pharmaceutical Statistics

Background:

• Regulatory authorities established bioequivalence (BE) evaluation procedures for highly variable drugs in the past decade.
• Existing methods from agencies like the Food and Drug Administration (FDA) and European Medicines Agency (EMA) differ in regulatory constants and statistical approaches.
• These differences can lead to variations in BE assessment, particularly concerning within-subject variability and Type I error rates.

Purpose of the Study:

• To analyze the discrepancies in bioequivalence evaluation procedures for highly variable drugs among regulatory authorities.
• To evaluate the impact of differing regulatory constants (FDA vs. EMA) on BE limit interpretation and Type I error inflation.
• To explore novel approaches, such as Exact algorithms, designed to harmonize and improve BE assessment.

Main Methods:

• Comparative analysis of FDA and EMA bioequivalence guidelines and their recommended regulatory constants.
• Statistical evaluation of Type I error rates associated with different regulatory constants.
• Review of recently proposed procedures, including Exact algorithms, for BE assessment.
• Examination of additional regulatory requirements, such as constraints on point estimates and upper limits for procedure application.

Main Results:

• The FDA's recommended constant creates discontinuity in BE limits and inflates Type I error (16-17%) around a 30% within-subject variation threshold.
• The EMA's constant avoids this discontinuity, with Type I error rates around 6-7%.
• Additional constraints on point estimates and upper limits lack scientific basis and complicate interpretation.
• Inconsistencies exist in applying BE criteria to AUC and Cmax across different authorities.

Conclusions:

• Significant disharmonization exists in regulatory approaches to bioequivalence for highly variable drugs.
• The EMA's approach is computationally simpler and more interpretable than the FDA's approximation-based method.
• Exact algorithms show promise in integrating the benefits of existing methods while mitigating their drawbacks.
• A unified and scientifically grounded approach to bioequivalence assessment is necessary to resolve current discrepancies.