Warfarin pharmacogenetics: development of a dosing algorithm for Omani patients

• DOI: 10.1038/jhg.2012.94
• Publication Date: 2012-08-02T06:20:59Z
• Authors: Anil Pathare; Murtadha Al Khabori; Salam Alkindi; Shoaib Al Zadjali; Rhea Misquith; Hammad Khan; Claudine Lapoumer...; Andras Paldi; Rajagopal Krishna...
• Keywords: Adult; Male; 0301 basic medicine; Genotype; Oman/ethnology; Cytochrome P-450 Enzyme System/genetics; Ethnic Groups/genetics; Population/methods; dosage; Linkage Disequilibrium; 03 medical and health sciences; Pharmacogenetics/*methods; Genetic; Cytochrome P-450 Enzyme System; Atrial Fibrillation; Genetics; Ethnicity; Humans; Drug Dosage Calculations; Prospective Studies; Cytochrome P450 Family 4; Polymorphism; Warfarin/*administration & Genetic Association Studies; Aged; Cytochrome P-450 CYP2C9; 0303 health sciences; Atrial Fibrillation/drug therapy; Middle Aged; 3. Good health; [SDV] Life Sciences [q-bio]; Genetics, Population; Aryl Hydrocarbon Hydroxylases/genetics; Genetic Loci; *Algorithms; Linear Models; Mixed Function Oxygenases/genetics; Female; Aryl Hydrocarbon Hydroxylases; Venous Thrombosis/drug therapy; Algorithms

The objective of our present study was to develop a warfarin dosing algorithm for the Omani patients, as performances of warfarin dosing algorithms vary across populations with impact on the daily maintenance dose. We studied the functional polymorphisms of CYP2C9, CYP4F2 and VKORC1 genes to evaluate their impact on the warfarin maintenance dose in an admixed Omani patient cohort with Caucasian, African and Asian ancestries. We observed a 64-fold inter-patient variability for warfarin to achieve stable international normalized ratio in these patients. Univariate analysis revealed that age, gender, weight, atrial fibrillation, deep vein thrombosis/pulmonary embolism and variant genotypes of CYP2C9 and VKORC1 loci were significantly associated with warfarin dose in the studied patient population. However, multiple regression model showed that only the atrial fibrillation, and homozygous CYP2C9 variant genotypes (*2/*3 and *3/*3) and VKORC1 GA and AA genotypes remained significant. A multivariate model, which included demographic, clinical and pharmacogenetic variables together explained 63% of the overall inter-patient variability in warfarin dose requirement in this microgeographically defined, ethnically admixed Omani patient cohort on warfarin. This locally developed model performed much better than the International Warfarin Pharmacogenetics Consortium (IWPC) model as the latter could only explain 34% of the inter-patient variability in Omani patients. VKORC1 3673G>A polymorphism emerged as the single most important predictor of warfarin dose variability, even in this admixed population (partial R(2)=0.45).