Non-genetic factors and polymorphisms in genes CYP2C9 and VKORC1: predictive algorithms for TTR in Brazilian patients on warfarin

Purpose This study was designed to evaluate the association of non-genetic factors and polymorphisms CYP2C9*2 (rs1799853), CYP2C9*3 (rs1075910), and VKORC1-G1639A (rs9923231) with time in therapeutic range (TTR), and to build a regression model to predict the quality of oral anticoagulation control in a sample of Brazilian patients. Methods This is a retrospective cohort study developed at an anticoagulation clinic of a university hospital. Overall, 312 patients were included. The quality of oral anticoagulation control was evaluated by TTR. TTR was dichotomized for analysis, using two cutoff points for classification as inadequate (TTR <= 60.0%) and optimal (TTR >= 75.0%) control. Results The average age was 60.4 +/- 13.5 years, with a predominance of women (187; 59.9%). The -G1639A polymorphism of the VKORC1 gene, when evaluated, based on the recessive inheritance pattern [AA x (GA + GG)], patients with AA genotype exhibited a higher TTR (68.2% versus 62.8%, p = 0.017). TTR <= 60.0% was associated with number of drugs in chronic use, assistance for warfarin administration, reports of not taking warfarin, absenteeism, sex (female), and target INR (International Normalized Ratio; 2.00-3.00). TTR >= 75.0% was associated with sex (male), target INR (2.00-3.00), assistance for warfarin administration, reports of not taking warfarin, and absenteeism. The two algorithms proposed showed adequate ability to predict TTR presenting good sensitivity and specificity. Conclusions Our findings provided useful information for risk stratification depending on TTR level and for future investigations on the quality of oral anticoagulation control in Brazilian anticoagulation clinics.