Case studies, practical issues and observations on population and individual bioequivalence.

Abstract

The FDA has proposed replacing the 1992 average bioequivalence (ABE) with population and individual bioequivalence (PBE and IBE). This has led to considerable public discussion between regulatory, academic and industry experts. At the heart of the discussion has been the relatively modest amount of available data to examine the behaviour of the PBE and IBE criteria. A retrospective analysis of 22 data sets from 15 replicate cross-over bioequivalence studies has been conducted (n=12-74). AUC and C(max) parameters from these studies were analysed using ABE, PBE and IBE methods. Of the 22 data sets for AUC, 19 pass ABE, all pass PBE and 20 pass IBE. Of the three data sets that failed ABE, all passed PBE and one passed IBE. The results for C(max) are more variable. Of the 16 data sets where ABE is demonstrated, one data set failed both PBE and IBE. Of the six data sets that failed ABE, two passed both PBE and IBE, three passed PBE but not IBE and one failed all three criteria. There were five data sets that passed ABE and PBE but not IBE. Additional practical issues involving the behaviour of the new criteria and its expected impact on sample size for highly variable drug products will be presented. The characterization of key parameters and their interrelationships will also be discussed with particular emphasis on the subject by formulation term in the IBE criteria. It is concluded that more studies and simulations are desirable before full-scale implementation of PBE and IBE criteria.