In this issue of Blood, Daneshjou et al describe significant advances in the pharmacogenetics of warfarin in African Americans. First, they identify a genetic variant associated with extreme warfarin doses through whole-exome sequencing. Second, they test the functional implications of this genetic variant that is specific for African populations.1
Pharmacogenetic studies of warfarin aiming to explain the interindividual variation in dose requirements have attracted great interest over the last decade. It is now known that the dose needed for therapeutic anticoagulation is influenced by genetic variation in the genes that encode the main warfarin metabolizing enzyme (CYP2C9) and the target of warfarin (VKORC1).2,3 Pharmacogenetic dosing algorithms incorporating the 3 alleles CYP2C9 *2 and *3 and VKORC1 rs9923231 account for at least ∼40% of the variability in warfarin dose in whites, whereas they only explain ∼20% in people of African descent, largely due to the lower frequencies of these alleles in the latter population.4-6 This difference was illustrated by the recent randomized clinical trials comparing gene-based warfarin dosing with clinical dosing: the European Pharmacogenetics of Anticoagulant Therapy (EU-PACT) and the American Clarification of Optimal Anticoagulation through Genetics (COAG).7,8 In EU-PACT, with ∼99% white patients, gene-based dosing improved time in therapeutic range and time to therapeutic anticoagulation.7 In contrast, the more genetically diverse COAG trial failed to show a difference in time in therapeutic range between gene-based and clinical dosing, and gene-based dosing actually performed worse than clinical dosing in the 27% that were of African ancestry.8
In light of this, it is clear that warfarin-dosing algorithms incorporating only the CYP2C9*2 and *3 and VKORC1 rs9923231 alleles do not adequately explain the variability in dose in African Americans. Perera and colleagues therefore conducted a genome-wide association study in African Americans to find additional variants contributing to warfarin dose.9 In this cohort, a novel association with dose was detected on chromosome 10 near CYP2C18 (rs12777823). This common variant was not associated with warfarin dose requirements in whites or Asians, suggesting that it is not causal but in linkage disequilibrium with 1 or more rare causal variants in people of African descent. Daneshjou et al set out to find such variants by exome sequencing 103 African Americans on extreme warfarin doses; however, no rare causal variants were found.1 Instead they discovered an association with a common polymorphism (rs7856096) on chromosome 9, which was replicated in 372 African Americans. It is located in the folylpolyglutamate synthase gene (FPGS) that encodes a cytosolic and mitochondrial enzyme involved in the folate homeostasis. The rs7856096 variant was associated with lower FPGS gene expression in 3 African lymphoblastoid cell lines. According to the Encyclopedia of DNA Elements data from several cell types, rs7856096 is located in a regulatory region upstream of exome 2, and it might be an alternative promoter in the liver cell line HepG2.
The study by Daneshjou et al adds complexity to the pharmacogenetics of warfarin by identifying an association with the folate pathway.1 FPGS catalyses the polyglutamation of folic acid and of certain antifolate drugs,10 but its connection with warfarin is unknown. The rs7856096 variant is common in African Americans with an allele frequency of ∼20%, which is why it has the potential to be important in this population. However, this variant and the previously found rs12777823 need to be studied in larger cohorts to prove that they are generalizable to other African populations.
Conflict-of-interest disclosure: The author declares no competing financial interests.