Retrospective evidence for clinical validity of expanded genetic model in warfarin dose optimization in a South Indian population.

AIM

To optimize warfarin dose in patients at risk for thrombotic events, we have recently developed a pharmacogenomic algorithm, which explained 44.9% of the variability in warfarin dose requirements using age, gender, BMI, vitamin K intake, CYP2C9 (*2 and *3) and VKORC1 (*3, *4 and -1639 G>A) as predictors. The aim of the current study is to develop an expanded genetic model that can explain greater percentage of warfarin variability and that has clinical validity.

PATIENTS & METHODS: CYP2C9*8, CYP4F2 V433M, GGCX G8016A and thyroid status were added to an expanded genetic model (n = 243).

RESULTS

The expanded genetic model explained 61% of the variability in warfarin dose requirements, has a prediction accuracy of ±11 mg/week and can differentiate warfarin sensitive and warfarin resistant groups efficiently (areas under receiver operating characteristic curves: 0.93 and 0.998, respectively; p < 0.0001). Higher percentage of International Normalized Ratios in therapeutic range (52.68 ± 4.21 vs 43.80 ± 2.27; p = 0.04) and prolonged time in therapeutic range (61.74 ± 3.18 vs 47.75 ± 5.77; p = 0.03) were observed in subjects with a prediction accuracy of <1 mg/day compared with subjects with prediction accuracy >1 mg/day. In the warfarin-resistant group, primary hypothyroidism was found to induce more resistance while in the warfarin-sensitive group, hyperthyroidism was found to increase sensitivity.

CONCLUSION

The expanded genetic model explains greater variability in warfarin dose requirements and it prolongs time in therapeutic range and minimizes out-of-range International Normalized Ratios. Thyroid status also influences warfarin dose adjustments.