722 VOLUME 84 NUMBER 6 | DECEMBER 2008 | www.nature.com/cpt ARTICLES nature publishing group Warfarin is the most commonly used oral anticoagulant in the world. Te maintenance dose is usually determined by moni- toring prothrombin time using an international normalized ratio (INR). Management of warfarin therapy requires fre- quent INR monitoring because of its narrow therapeutic index and large inter- and intra-individual variability. Incorrect dos- age, especially during the initial phase of treatment, carries a high risk of bleeding or failure to prevent thromboembolism. 1 Polymorphisms in two genes—CYP2C9 and VKORC1—have been repeatedly found to associate with the clinical response to warfarin. 2–13 CYP2C9 encodes the cytochrome P-450 (CYP) enzyme, CYP2C9, which is primarily responsible for the dispo- sition of S-warfarin, the most active isomer of clinically used racemic warfarin. 14 VKORC1 codifes the vitamin K epoxide reductase subunit I (VKORC1), a key enzyme of the vitamin K cycle and the molecular target of coumarin anticoagulants. 15,16 CYP2C9 and/or VKORC1 genotypes have been included as cov- ariates in several algorithms developed for the estimation of indi- vidual warfarin dosing requirements. 2–7,10–12 As the frequency distribution of polymorphisms in these two genes varies across populations (http://pgeni.unc.edu/purpose.asp and pharmgkb. org/views/project.jsp?pId=56), the relative contribution of the CYP2C9 and VKORC1 genotypes to estimates of warfarin dosing is not constant worldwide. For example, the VKORC1 genotype was reported to be not useful or of limited value for predicting warfarin dose requirements in African Americans, 6,10 whereas it explained 18 and 37% of the variance in warfarin doses among Chinese and Malays, respectively. 4 Here we report a study of the impact of CYP2C9 and VKORC1 polymorphisms, as well as nongenetic covariates, in predicting the stable warfarin thera- peutic dose in Brazilian patients under chronic treatment with this anticoagulant for prevention of thromboembolism. Because of the heterogeneity and extensive admixture of the present-day Brazilian population, extrapolation of data derived from well- defned ethnic groups is clearly not applicable to the majority of Brazilians. 17–19 RESULTS Table 1 summarizes the demographics and clinical character- istics of the 390 study subjects. Te stable warfarin dose var- ied over a 15-fold range (5–75 mg/week) in the overall study population, with mean (SD) values of 28.9 (±12.3), 32.9 (±12.4), and 35.3 (±14.6) in the white, intermediate, and black groups, respectively (P = 0.0004). Pair-wise comparison revealed that the weekly warfarin dose was signifcantly lower in white than in either intermediate (P = 0.006) or black (P = 0.0003) patients. Tere was no diference in the weekly warfarin dose between intermediate and black patients (P = 0.22). 1 Divisão de Farmacologia, Instituto Nacional de Câncer, Rio de Janeiro, Brazil; 2 Programa de Computação Científica, Fundação Oswaldo Cruz, Rio de Janeiro, Brazil; 3 Ambulatório de Anticoagulação, Instituto Nacional de Cardiologia de Laranjeiras, Rio de Janeiro, Brazil; 4 Laboratorio de Neurociências (LIM-27), Instituto de Psiquiatria, Faculdade de Medicina, Universidade de São Paulo, São Paulo, Brazil. Correspondence: G Suarez-Kurtz (kurtz@inca.gov.br) Received 29 May 2008; accepted 16 July 2008; advance online publication 27 August 2008. doi:10.1038/clpt.2008.166 Pharmacogenetics of Warfarin: Development of a Dosing Algorithm for Brazilian Patients JA Perini 1 , CJ Struchiner 2 , E Silva-Assunção 3 , ISC Santana 1 , F Rangel 3 , EB Ojopi 4 , E Dias-Neto 4 and G Suarez-Kurtz 1 A dosing algorithm including genetic (VKORC1 and CYP2C9 genotypes) and nongenetic factors (age, weight, therapeutic indication, and cotreatment with amiodarone or simvastatin) explained 51% of the variance in stable weekly warfarin doses in 390 patients attending an anticoagulant clinic in a Brazilian public hospital. The VKORC1 3673G>A genotype was the most important predictor of warfarin dose, with a partial R 2 value of 23.9%. Replacing the VKORC1 3673G>A genotype with VKORC1 diplotype did not increase the algorithm’s predictive power. We suggest that three other single-nucleotide polymorphisms (SNPs) (5808T>G, 6853G>C, and 9041G>A) that are in strong linkage disequilibrium (LD) with 3673G>A would be equally good predictors of the warfarin dose requirement. The algorithm’s predictive power was similar across the self-identified “race/color” subsets. “Race/color” was not associated with stable warfarin dose in the multiple regression model, although the required warfarin dose was significantly lower (P = 0.006) in white (29 ± 13 mg/week, n = 196) than in black patients (35 ± 15 mg/week, n = 76).