Evaluating Within-Subject Variability for Narrow Therapeutic Index Drugs.

The US Food and Drug Administration (FDA) reference-scaled average bioequivalence approach scales the bioequivalence (BE) limits of narrow therapeutic index drugs (NTIDs) to the intrasubject or within-subject variability (WSV) of the reference-listed drug. A clinical study was conducted to evaluate the WSV of warfarin (Coumadin), 10 mg, administered to 10 healthy volunteers exhibiting similar cytochrome P450 2C9 and vitamin K epoxide reductase alleles on 3 study days. Individual intrasubject coefficients of variation for maximum plasma concentration and area under the curve (0-72 hour) ranged from 3.7-15% and from 4.3-16.2%, respectively (R-warfarin) and from 5.4-19.1% and from 2.5-11.9%, respectively (S-warfarin). Two BE tests were performed on a WSV distribution obtained by bootstrapping 1,000 replicates of the clinical data, yielding passing rates of 95-97% for the mean comparison and 84-87% for the variability comparison. The variability comparison passing rate was lower than expected for an NTID product tested against itself, but it may provide further assurance of BE.