ABSTRACT
This chapter provides an overview of statistical methods including the methods recommended by the United States Food and Drug Administration (FDA) for selection of a non-inferiority/similarity margin. A comparison of these methods indicates that there exist no methods which are uniformly superior to others. The selection of non-inferiority/similarity has an impact on the sample size required for achieving a desired power for the assessment of biosimilar products. In practice, however, it is not uncommon to have a disagreement in non-inferiority/similarity margin selection between the sponsors and the regulatory agency. In this case, a risk/benefit assessment is suggested for determining the non-inferiority/similarity margin in biosimilar product development. This chapter examines the performances of several criteria for risk/benefit assessment for selection of non-inferiority/similarity.
