RESUMEN
OBJECTIVE: To evaluate the impact of the two most common CYP2C9 variant alleles (*2 and *3) on the maintenance dose of warfarin and on the quality of anticoagulation control in Brazilians. METHODS: Patients (n = 103) initiated warfarin therapy with 5 mg/day (or 2.5 mg/day when over 80 years old). The international normalized ratio (INR) was targeted between 2 and 3, monitored every week until four consecutive adequate measures had been obtained, and then monthly. Serious hemorrhagic events were defined by the need for inpatient hospitalization. CYP2C9 genotyping was obtained by PCR-RFLP. RESULTS: The frequencies of CYP2C9*2 and CYP2C9*3 were 0.097 and 0.073, respectively, with genotypic distribution fitting Hardy-Weinberg equilibrium. CYP2C9 genotype was the only clinical feature associated with the risk of severe bleeding (one-sided P = 0.019, Fisher exact method), with an odds ratio of 4.8 (95% confidence interval of 1.4-16.6) for any variant genotype as compared to CYP2C9*1*1. Patients with either CYP2C9*2 or CYP2C9*3 were equally difficult to maintain in the INR target range, showing significantly (one-sided P = 0.038, Mann-Whitney U-test) reduced ratio of adequate INR measures (0.54 +/- 0.2), when compared to CYP2C9*1*1 patients (0.63 +/- 0.2). Patients with CYP2C9*3, but not CYP2C9*2, required significantly (one-sided P = 0.001, Mann-Whitney U-test) lower warfarin maintenance doses (3.1 +/- 1.8 mg) than CYP2C9*1*1 patients (5.3 +/- 2.1 mg). CONCLUSION: Patients with either CYP2C9*2 or CYP2C9*3 show higher risk of over-anticoagulation compared to CYP2C9*1*1 subjects and could benefit from a reduction in the initial warfarin standard dose (e.g., to 2.5 mg/day).