ABSTRACT
For biosimilarity assessment between a proposed biosimilar (test) product and an innovative biological (reference) product, the United States Food and Drug Administration (FDA) recommended a stepwise approach for obtaining the totality-of-the-evidence in support of regulatory approval of the submission. The stepwise approach starts with analytical similarity assessment followed by pharmacokinetic (PK) and pharmacodynamic (PD) similarity and clinical similarity. When performing analytical similarity assessment on certain critical quality attributes (CQAs) which are relevant to clinical outcome, a potential drift in mean response and/or drift in variability associated with the response of the reference product over time may be observed (Kim et al., 2017). When there is drift in the reference product over time, it is of interest to determine which lots (i.e., reference lots before the drift, after the drift, or combined) should be used for analytical similarity assessment. In this chapter, statistical tests for a possible drift in mean, a drift in variability, and a drift in both mean and variability are derived. If the drifts have been statistically confirmed, it is suggested that regulatory guidance regarding which lots should be used for analytical similarity assessment should be developed.
