DISSOLUTION PROFILE SIMILARITY ANALYSES-STATISTICAL PRINCIPLES, METHODS AND CONSIDERATIONS.

Area of Science:

• Pharmaceutical Sciences
• Biostatistics
• Regulatory Science

Background:

• Dissolution similarity testing is vital for pharmaceutical product lifecycle management and regulatory decisions.
• Current regulatory guidance emphasizes the similarity factor f2, with limited exploration of alternative mathematical approaches.
• A workshop highlighted the need for current practices and global harmonization in assessing dissolution profile similarity.

Purpose of the Study:

• To discuss mathematical principles of model-independent statistical methods for dissolution profile similarity analysis.
• To provide a decision tree for selecting appropriate statistical methods based on drug product characteristics.
• To offer recommendations for dissolution profile study design, including analytical and statistical considerations for sufficient power.

Main Methods:

• Review and discussion of model-independent statistical methods for dissolution profile similarity.
• Presentation of a decision tree to guide method selection.
• Analysis of dissolution profile data, including multi-batch evaluations.

Main Results:

• In-depth discussion of the mathematical principles underlying various statistical methods.
• A practical decision tree to aid in the selection of appropriate statistical methods.
• Recommendations for optimizing dissolution study design and data evaluation.

Conclusions:

• A deeper understanding of testing objectives and statistical properties is essential for selecting appropriate dissolution similarity methods.
• The provided decision tree and recommendations support informed application of statistical methods in practice.
• The study contributes to harmonizing global practices for in vitro dissolution similarity assessment in drug product quality evaluation.