Race influences warfarin dose changes associated with genetic factors

Area of Science:

• Pharmacogenomics
• Clinical Pharmacology
• Genetics

Background:

• Current warfarin dosing algorithms often adjust for race, but this approach may attenuate the true differential effects of clinical and genetic factors.
• This attenuation can be more pronounced in underrepresented racial groups, potentially leading to suboptimal dosing.

Purpose of the Study:

• To evaluate whether the effect of clinical (age, body surface area, chronic kidney disease, amiodarone use) and genetic factors (CYP2C9 variants, rs12777823, VKORC1, CYP4F2) on warfarin dose differs by race.
• To compare the predictive ability of race-combined versus race-stratified warfarin dosing models.

Main Methods:

• Regression analyses were performed on data from 1357 patients in a prospective cohort study.
• Differential effects of predictors by race were assessed using predictor-race interactions in race-combined analyses.
• Race-stratified and race-combined models were compared for predictive accuracy.

Main Results:

• Warfarin dose was influenced by age, body surface area, chronic kidney disease, amiodarone use, CYP2C9*3, and VKORC1 in both European Americans and African Americans.
• CYP2C9*2 and CYP4F2 variants affected dose in European Americans, while rs12777823 affected dose in African Americans.
• Race-stratified analysis improved warfarin dose prediction in both groups compared to race-combined analysis, with significant differences in the impact of CYP2C9*2, rs12777823, and VKORC1 between races.

Conclusions:

• The effect of key clinical and genetic predictors on warfarin dose significantly differs by race.
• Race-stratified warfarin dosing algorithms demonstrate superior predictive ability compared to race-adjusted algorithms.
• These findings support the recommendation for race-stratified warfarin dosing algorithms to optimize patient outcomes and explain divergent results in pharmacogenetic trials.