Ethnic Diversity and Warfarin Pharmacogenomics

Abstract

Warfarin has remained the most commonly prescribed vitamin K oral anticoagulant worldwide since its approval in 1954. Dosing challenges including having a narrow therapeutic window and a wide interpatient variability in dosing requirements have contributed to making it the most studied drug in terms of genotype-phenotype relationships. However, most of these studies have been conducted in Whites or Asians which means the current pharmacogenomics evidence-base does not reflect ethnic diversity. Due to differences in minor allele frequencies of key genetic variants, studies conducted in Whites/Asians may not be applicable to underrepresented populations such as Blacks, Hispanics/Latinos, American Indians/Alaska Natives and Native Hawaiians/other Pacific Islanders. This may exacerbate health inequalities when Whites/Asians have better anticoagulation profiles due to the existence of validated pharmacogenomic dosing algorithms which fail to perform similarly in the underrepresented populations. To examine the extent to which individual races/ethnicities are represented in the existing body of pharmacogenomic evidence, we review evidence pertaining to published pharmacogenomic dosing algorithms, including clinical utility studies, cost-effectiveness studies and clinical implementation guidelines that have been published in the warfarin field.

Keywords: ancestry; ethnic diversity; pharmacogenetics; pharmacogenomics; underrepresented populations.

FIGURE 1

An overview of the genes involved in warfarin’s mechanism of action (Wadelius et al., 2007; Wadelius and Pirmohamed, 2007). ABCB1, ATP Binding Cassette Subfamily B Member one; APOE, Apolipoprotein E; CALU, Calumenin; CYP, Cytochrome P450; EPHX1, Epoxide hydrolase one; GGCX, Gamma-glutamyl carboxylase; F2, Coagulation factor II gene or prothrombin; F5, Coagulation factor V gene; F7, Coagulation factor VII gene; F9, Coagulation factor IX gene; F10, Coagulation factor X; NQ O 1, NAD (P) H dehydrogenase, quinone one; NR1I2/3, Nuclear Receptor Subfamily 1 Group I Member 2/3; ORM, Orosomucoid; PROC, Protein C; PROS1, Protein S; PROZ, Protein Z; SERPINC1, Serpin Family C Member one; VKORC1, Vitamin K epoxide reductase complex subunit 1. Gene names are italicized and shown in green.

FIGURE 2

Number of pharmacogenomic studies that developed dosing algorithms as of 20 May 2020. A shows the number of pharmacogenomic studies that included at least 5% or 100% of a population category. On the other hand, B, C show the country-breakdown per population category with B corresponding to the 5% cut-off and C, the 100% cut-off. As recruitment was not stratified by location/country, studies that recruited from multiple countries are excluded in B, C.

FIGURE 3

Genetic variables included in the pharmacogenomic algorithms. A, B, respectively, show the number and percentage of pharmacogenomic algorithms including a specific genetic variant. APOE, Apolipoprotein E; CYP, Cytochrome P450; VKORC1, Vitamin K epoxide reductase complex subunit 1.