Comparison of Dissolution Profiles: A Statistician's Perspective

Area of Science:

• Pharmaceutical Sciences
• Biostatistics
• Drug Quality Assurance

Background:

• Dissolution profile comparisons are crucial for post-approval changes, ensuring product quality.
• Current guidelines and statistical practices often conflict, leading to lengthy discussions.
• Existing statistical methods for highly variable dissolution profiles may require updates.

Purpose of the Study:

• To address statistical interpretation conflicts in dissolution profile comparisons for post-approval changes.
• To propose statistical improvements for analyzing and interpreting dissolution data.
• To advocate for updated regulatory guidance on statistical equivalence testing.

Main Methods:

• Review of current statistical principles and guideline interpretations.
• Analysis of existing methods for dissolution profile comparison, including the similarity factor and multivariate equivalence procedures.
• Evaluation of the T² test for equivalence as a suitable method for highly variable dissolution profiles.

Main Results:

• Current practices often misinterpret sample size recommendations and lack clear variability criteria.
• Pooled comparisons are preferred over batch-to-batch comparisons for multiple batches.
• The T² test is identified as an appropriate and reliable method for comparing highly variable dissolution profiles, with available software implementations.

Conclusions:

• Statistical interpretation of dissolution profile comparisons requires clearer guidelines and updated methodologies.
• Implementing the T² test for equivalence can improve the reliability of decisions for highly variable drug products.
• Updating FDA guidelines on multivariate equivalence procedures is necessary to reflect current statistical knowledge.