Limitations of S-warfarin truncated area under the concentration-time curve to predict cytochrome P450 2c9 activity.

OBJECTIVE: Phenotyping cytochrome P450 (CYP) 2C9 activity with S-warfarin requires extensive blood sampling to characterize area under the concentration-time curve (AUC). This retrospective data analysis was conducted to determine if truncated S-warfarin AUCs can be used to measure CYP2C9 activity.

METHODS: S-warfarin plasma concentrations were obtained from healthy adults (n = 84) genotyped as CYP2C9 extensive metabolizers (EMs) from 6 published studies. Subjects received a single 10 mg oral warfarin dose during baseline and treatment conditions. AUC zero to infinity (AUCINF) and truncated AUCs at 48 h (AUC(48)), at 72 h (AUC(72)) and at 96 h (AUC(96)) were determined by noncompartmental analysis. Equivalence was determined via least squares geometric mean ratios (LS-GMRs) with 90% confidence intervals (CI) within 0.8-1.25.

RESULTS: A lack of equivalence was observed for AUC(48) and AUC(72) compared to AUC(INF) during baseline conditions in all evaluated studies and during treatment conditions in 5 of 6 studies. Equivalence was observed for AUC(96) compared to AUC(INF) during all baseline and treatment conditions. Results were consistent across all evaluated AUCs between baseline and treatment without a CYP2C9-mediated drug-drug interaction and during induction with an oral contraceptive. During inhibition with fluvastatin, a lack of equivalence was observed with AUC(INF)(LS-GMR [90%CI] = 1.25 [1.16-1.34]) and AUC(96) (1.2 [1.13=1.27]). In contrast, equivalence was observed for AUC(48)(1.15 [1.08-1.22]) and AUC(72) (1.18 [1.11-1.24]).

CONCLUSIONS: S-warfarin truncated AUC(48) and AUC(72) poorly characterize AUC(INF) and are unable to detect weak CYP2C9 inhibition with fluvastatin. S-warfarin phenotyping parameters need to ensure blood sampling of at least 96 h to characterize AUC and thus CYP2C9 activity.