High variability in drug pharmacokinetics complicates determination of bioequivalence: Experience with verapamil

Purpose. For the assessment of bioequivalence it is assumed that drug clearance in each subject on each of the study days is the same and any observed differences in AUC and/or Cmax between a brand and generic formulation are due to differences in bioavailability. We hypothesized that this assumption was invalid for highly variable drugs such as verapamil and tested it by comparing bioavailability for the brand vs itself. Methods. To avoid any contribution from potential formulation differences, we evaluated bioavailability for Isoptin SR 240 mg tablets in 9 healthy volunteers on 2 occasions separated by 1 week as part of a larger study. A validated HPLC assay was used to measure serial blood samples over 36 hours. Results. The AUC(0-t), varied 3.8 fold among subjects and 5/9 subjects had > 30% difference in AUC(0-t), on the 2 days. After log transformation, the mean AUC(0-t) +/- %cv (ng . h/mL) on Occasion 1 (878 +/- 38) was 23% greater (p = 0.031) than on Occasion 2(713 +/- 41). The 90% confidence interval of Occasion 1/Occasion 2 was 106-143%. The Cmax varied > 9 fold (30-278 ng/mL) among subjects. The intra-subject difference between days ranged from -46% to +298%. The 90% confidence interval was 72-152% for Cmax. Since the same lot of Isoptin was used in the same subjects on 2 occasions, the observed differences must be due to biological variability in verapamil pharmacokinetics, not formulation differences. Conclusions. The intra-subject biological variability complicates bioequivalence assessment and can lead to an erroneous assumption of bioinequivalence.