Equivalence Test and Margin Determination for Biosimilar Analytical Assessment

• Activity Number: 658
• Type: Contributed
• Date/Time: Thursday, August 13, 2015 : 8:30 AM to 10:20 AM
• Sponsor: Biopharmaceutical Section
• Author(s): Yi Tsong* and Xiaoyu Dong and Meiyu Shen
• Companies: FDA and FDA and FDA/CDER/OB
• Keywords: biosimilar ; analytical equivalence ; Equivalence margin

Abstract

Analytical equivalence and similarity assessments are the foundation for biosimilar application. The quality attributes of the product can be categorized into three tiers. Tier-1 quality attributes of the product are amendable to statistical analysis and most critical. They are subjected to the statistical equivalence testing. The test is a scaled average equivalence assessment in nature. Assuming the reported value from each lot is the mean of multiple within-lot observations, equivalence is established if the mean difference of the two products falls with the equivalence margins. We propose that the margin is the standard deviation of the reference product multiplied by a constant. This constant is derived based on proper power to pass the equivalence test when the true difference is no more than one-eighth of the standard deviation of the reference product when there are 10 biosimilar and 10 reference lots. In practice, the standard deviation of the reference product may need to be estimated from the sample. Pending on the balance and unbalance of lot numbers, difference approaches are also recommended to achieve the best results of type I error rate control and power.