Relative contribution of CYP2C9 and VKORC1 genotypes and early INR response to the prediction of warfarin sensitivity during initiation of therapy

Chun Li, Ute I Schwarz, Marylyn D Ritchie, Dan M Roden, C Michael Stein, Daniel Kurnik, Chun Li, Ute I Schwarz, Marylyn D Ritchie, Dan M Roden, C Michael Stein, Daniel Kurnik

Abstract

Genetic variants in CYP2C9 and VKORC1 strongly affect steady-state warfarin dose. However, these variants also affect early international normalized ratio (INR) values during warfarin initiation. We examined whether CYP2C9/VKORC1 genotypes provide information about warfarin sensitivity additional to that provided by early INR responses. In 214 patients starting warfarin with INR-guided dose adjustments, we determined whether CYP2C9 and VKORC1 genotypes were associated with early measures of warfarin sensitivity (time to INR >/= lower limit of therapeutic range; time to INR > 4; and first stable warfarin dose) after adjusting for early (days 4-6) and week 1 (days 7-9) INR values. Early INRs were associated with all outcomes (all P < .001) and were more informative than genotypes. For time to INR more than or equal to the lower limit of therapeutic range, adding either early INRs or genotypes to a baseline model (clinical variables only) increased the goodness-of-fit (R(2)) from 0.05 to 0.42 and 0.19, respectively (full model, R(2) = 0.46). For first stable warfarin dose, adding either early INRs or genotypes to the baseline model increased the R(2) from 0.08 to 0.32 and 0.27, respectively (full model, R(2) = 0.40). After inclusion of week 1 INRs, CYP2C9 (P = .08) and VKORC1 (P = .30) were not associated with stable warfarin dose. Thus, much of the information provided by CYP2C9 and VKORC1 genotypes during warfarin initiation is captured by the early INR response.

Figures

Figure 1
Figure 1
Comparison of goodness of fit (R2) among 4 models for each of the 3 outcome variables. (A) Time to INR ≥ the lower limit of therapeutic range. (B) Time to INR > 4. (C) Early stable warfarin dose. In each panel, the first bar (□) represents the baseline model (including age, ethnicity, sex, amiodarone use, target INR, and cumulative warfarin dose); second bar (), the baseline model + VKORC1 / CYP2C9 genotypes; the third bar (), the baseline model + early INR values; fourth bar (■), the full model. The difference between the last 2 bars (■ and ) in each panel represents the contribution of genotypes after adjustment for baseline covariates and INRearly.

Source: PubMed

3
Subscribe