PII-188 - DEEP DIVE INTO GENERIC DRUG APPLICATIONS TO SEEK DATA-DRIVEN HARMONIZATION OF BIOEQUIVALENCE CRITERIA FOR NARROW THERAPEUTIC INDEX DRUGS.

M. Mina, K. Anim Anno, Z. Zhang, W. Sun, L. Zhang, W. Jiang; U.S. Food and Drug Administration.

Background: U.S. Food and Drug Administration (FDA) recommends a fully replicated, 4-way crossover bioequivalence (BE) study for generic narrow therapeutic index (NTI) drugs where BE is based on reference scaled and unscaled average BE limits, as well as test and reference within-subject variability comparison (sWT vs sWR) of pharmacokinetic (PK) parameters. Most other regulatory agencies have different BE approaches and criteria for NTI drugs. This study is to subject the 4-way crossover BE data received by FDA against various BE criteria from different agencies or modified FDA criteria to compare the passing rate, seeking data-driven harmonization of NTI BE criteria.

Methods: BE criteria among 6 agencies (Table 1) were compiled and modified FDA criteria were explored (e.g., considering BE limits no narrower than 94.87-105.41% or 90.00-111.11%, without variability comparison). The FDA’s internal database was utilized to collect 4-way crossover NTI Abbreviated New Drug Application (ANDA) BE data (83 ANDAs with 169 BE studies). These data were then subjected to different BE criteria to obtain the passing rates.

Results: BE passing rates were summarized in Table 1 when subjecting data against various regulatory agencies’ NTI BE criteria. The lowest passing rate 77.51% was obtained with EMA criteria, where most failed studies (97.31%) had sWR > 0.1. When sWR < 0.05 ( < 1.4% of collected PK parameters in this range), half of the PK parameters failed current FDA criteria when comparing reference vs itself. When modifying current FDA criteria to be no narrower than 90.00-111.11% for AUC and Cmax, the passing rate increased from 88.17% to 91.72%. When this proposed criterion applied to AUC only, the passing rate increased to 96.45%. When removing variability comparison, the increase in passing rate is less than 1.18%.

Conclusion: This study helps clarify the strengths and limitations of different BE criteria and provide data-driven options to aid in global harmonization of NTI BE criteria, e.g., applications of capping BE limits no narrower than 90.00-111.11%, tighter criteria for Cmax only if Cmax is important for safety, efficacy, or drug level monitoring, consideration of partially replicated study with and without variability comparison.