Pharmacogenetic-guided Warfarin Dosing Algorithm in African-Americans

Abstract

We aim to develop warfarin dosing algorithm for African-Americans. We explored demographic, clinical, and genetic data from a previously collected cohort of 163 African-American patients with a stable warfarin dose. We explored 2 approaches to develop the algorithm: multiple linear regression and artificial neural network (ANN). The clinical significance of the 2 dosing algorithms was evaluated by calculating the percentage of patients whose predicted dose of warfarin was within 20% of the actual dose. Linear regression model and ANN model predicted the ideal dose in 52% and 48% of the patients, respectively. The mean absolute error using linear regression model was estimated to be 10.8 mg compared with 10.9 mg using ANN. Linear regression and ANN models identified several predictors of warfarin dose including age, weight, CYP2C9 genotype *1/*1, VKORC1 genotype, rs12777823 genotype, rs2108622 genotype, congestive heart failure, and amiodarone use. In conclusion, we developed a warfarin dosing algorithm for African-Americans. The proposed dosing algorithm has the potential to recommend warfarin doses that are close to the appropriate doses. The use of more sophisticated ANN approach did not result in improved predictive performance of the dosing algorithm except for patients of a dose of ≥49 mg/wk.

Clinical Pharmacogenetics Implementation Consortium (CPIC) Guideline for Pharmacogenetics‐Guided Warfarin Dosing: 2017 Update

“…Regression analysis showed that addition of this SNP improves the dosing algorithm published by the International Warfarin Pharmacogenetics Consortium (IWPC) by 21%. Further studies have demonstrated the importance of this SNP in African Americans (28). Although this variant is common in other ethnic populations, an association with warfarin dose has only been detected among African Americans suggesting it is not the underlying cause but likely inherited with other variant(s) on a haplotype that influences warfarin dose in this population.…”

Warfarin Pharmacogenomics in Diverse Populations

“…Such studies observed improvements in performance (R 2 = 38%) versus traditional genotype-guided algorithms, such as the IWPC (R 2 = 26%) ( Table 2). 35 In a retrospective analysis using electronic health records, clinical information alone explained 24% of variation in stable dose in AAs, the IWPC algorithm improved R 2 to 29%, and an expanded genetic algorithm, including CYP2C9*6, CYP2C9*8, CYP4F2*3, and CALU rs339097, improved the R 2 to 41%. 37 Similar results were observed in other studies, in which AA-specific algorithms explain more of the phenotypic variation (R 2 = 27%) than IWPC (R 2 = 15%) or clinical (R 2 = 16%) algorithms.…”

“…The majority of warfarin pharmacogenetic studies in individuals of African ancestry have been performed in AA populations from the United States . CYP2C9*2 and CYP2C9*3 are not as useful in predicting warfarin dose requirements in AAs, partly because of their lower prevalence compared with individuals of European ancestry .…”

“…Subsequent studies developed AA‐specific algorithms that incorporate variants with warfarin dose effects in AAs, such as CYP2C9*5, *6, *8, and *11 and rs12777823. Such studies observed improvements in performance ( R 2 = 38%) versus traditional genotype‐guided algorithms, such as the IWPC ( R 2 = 26%) (Table ) . In a retrospective analysis using electronic health records, clinical information alone explained 24% of variation in stable dose in AAs, the IWPC algorithm improved R 2 to 29%, and an expanded genetic algorithm, including CYP2C9*6 , CYP2C9*8 , CYP4F2*3 , and CALU rs339097, improved the R 2 to 41% .…”

“…The majority of warfarin pharmacogenetic studies in individuals of African ancestry have been performed in AA populations from the United States. 15,28,[35][36][37] CYP2C9*2 and CYP2C9*3 are not as useful in predicting warfarin dose requirements in AAs, partly because of their lower prevalence compared with individuals of European ancestry. 38 Important warfarin pharmacogenetic variants in AAs include CYP2C9*5,*6,*8,*11, and 18786T, rs12777823 in the CYP2C gene cluster, VKORC1-1639G>A, GGCX (CAA) 16/17, and CALU rs339097 ( Table 1).…”