Pharmacogenomics of coumarin anticoagulants: are we underestimating the role of CYP2C9?

Thromb Haemost 2008; 100: 175–176 More than 60 years after its successful introduction for the prevention of thromboembolic events, the coumarin anticoagulant warfarin is still one of the most frequently prescribed cardiovascular drugs. The recently conducted Active W study once again established its therapeutic superiority over even a combination of antiplatelet drugs for the prevention of stroke in patients suffering from atrial fibrillation, the most common indication for coumarins (1). However, despite its undisputable therapeutic benefits, coumarins are notorious for the difficulties which are found in establishing and maintaining an adequate dose, reflecting the inter-individual and intra-individual variability in dose requirements and necessitating frequent monitoring of the International Normalized Ratio (INR). This variability is a consequence of age, of genetic factors and of environmental factors such as compliance, variations in vitamin K intake, several disease states and drug interactions. The main genetic research targets are the CYP2C9 gene, encoding the main metabolizing enzyme of coumarins, and theVKORC1 gene, encoding the target protein for coumarins on Vitamin K epoxide. Since the pioneering study by Aital et al. (2) who reported a higher bleeding risk and lower dose requirements in warfarin users carrying a CYP2C9*2 or *3 allele in a British population, many studies have examined the effect of CYP2C9 variant alleles on the anticoagulant status of the different coumarins (3). Since 2004 the scope of pharmacogenetic studies in users of coumarins has been broadened by the identification of the VKORC1 gene (4), and by the finding in several studies that polymorphisms of the VKORC1 gene contributed considerably more to the variability in warfarin dose requirements than polymorphisms of the CYP2C9 gene (5, 6).This finding raised new expectations about enhancing the safety of warfarin therapy by genotype-guided dosing. In this issue of Thrombosis and Haemostasis Meckley et al. add a valuable contribution to the pharmacogenetic research area of coumarin anticoagulants (7). They evaluated the relative effects ofVKORC1 and CYP2C9 genotypes on several relevant patient outcomes, such as time to therapeutic range, bleeding events and overanticoagulation, in a European-American cohort of warfarin patients. Their finding that the VKORC1 genotype explains a larger part of the inter-individual variability in warfarin dose requirements than the CYP2C9 genotype (27% and 12%, respectively) is in agreement with most other studies (3). However, for the clinically more relevant outcomes bleeding, overanticoagulation and time to stable anticoagulation, Meckley et al. found a strong association with the CYP2C9 variant status, whereas the VKORC1 variant status was only significantly associated with overanticoagulation in the first month of warfarin therapy. In two studies among users of the coumarins acenocoumarol and phenprocoumon similar results have been reported (8, 9). For both coumarins, the VKORC1 genotype explained a considerably larger part of the variability in dose requirements than the CYP2C9 genotype, whereas the time to stable anticoagulation was associated only with the CYP2C9 genotype: with the CYP2C9*3 allele in users of acenocoumarol and, perhaps somewhat surprisingly, with the CYP2C9*2 allele in users of phenprocoumon. In these studies the risk of overanticoagulation was strongly increased in patients with variant alleles of both CYP2C9 and VKORC1 compared to patients with no variant alleles or a variant allele of either CYP2C9 or VKORC1. More importantly, in a prospective follow up study among AfricanAmerican and European-American warfarin users, Limdi et al. (10) found an increased risk of major bleeding in patients with a variant CYP2C9 genotype in both populations compared to patients with the CYP2C9 wild type, whereas no such association was found with the VKORC1 1173 C/T genotype. This bleeding risk was more strongly increased before stabilisation on warfarin, but a significant association of major bleeding risk with the CYP2C9 genotype persisted thereafter. Although some studies failed to find an association between the CYP2C9 genotype and time to stable anticoagulation in users of warfarin (11, 12), these studies also failed to report such an association with the VKORC1 genotype. So, the majority of the hitherto performed studies and the study of Meckley et al. (7) in this issue, strongly suggest an association of the CYP2C9 variant status with the clinically relevant outcomes bleeding and time to © 2008 Schattauer GmbH, Stuttgart