Risk Assessment of Drug Product Content Uniformity Release Failure: A Bayesian Approach

• Activity Number: 233
• Type: Contributed
• Date/Time: Monday, August 3, 2009 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): David J. LeBlond*+
• Companies: Abbott Laboratories
• Keywords: drug content uniformity ; Bayesian applications ; risk assessment ; compendial tests ; WinBUGS ; nonclinical statistics

Abstract

Drug potency unit dose content uniformity (CU) is a key quality indicator of a drug product batch. Compendial standards based on 2-stage sampling plans (e.g., USP905) strictly regulate CU. Stage 2 testing triples analytical costs. A CU failure wastes resources and valuable active ingredient. Recent FDA and ICH initiatives promote science and risk based decision making and underscore the importance of predicting a drug product's stage 2 testing and CU failure probabilities. A hierarchical Bayesian approach to the prediction of CU failure of future batches is described and illustrated with historical data. The effects of number of historical batches and prior distributions are investigated. The approach represents a straight-forward decision tool that avoids over-conservative frequentist approximations, incorporates prior product knowledge, and is simple to implement using WinBUGS.