Use of Pharmacogenetics in Guiding Treatment with Warfarin

Warfarin is the most widely used oral anticoagulant for the treatment of thromboembolic disorders and for stroke prophylaxis. Warfarin is a problematic drug because it exhibits large interindividual variation in the required therapeutic dose, has a narrow therapeutic range, and shows multiple food and drug interactions. Its anticoagulant effect is monitored by measuring the international normalized ratio (INR), which is a function of the time required for a patient’s blood to coagulate relative to the time it takes for a reference blood sample. Although warfarin has been used in humans for more than 50 years, its main side effect— bleeding—is a leading cause of hospital admission and drugrelated death (1, 2 ). This problem has made patients and clinicians yearn for a new efficient and safe oral anticoagulant drug that does not require frequent monitoring. In Europe, a new oral anticoagulant drug (dabigatran) claimed to have these qualities has been licensed for short-term primary prevention of venous thromboembolic events, but its effectiveness in longterm secondary thromboprophylaxis remains to be shown. Furthermore, the daily cost of dabigatran is 5 times that of warfarin therapy including INR tests. To switch all warfarin patients (currently 1% of the population in many Western countries) to dabigatran would boost national costs in countries with subsidized drug programs; therefore, national authorities will probably encourage the continued use of warfarin, even when oral thrombin inhibitors become available for longterm thromboprophylaxis. Given that warfarin is likely to maintain its position as the most widely used oral anticoagulant for the forseeable future, it is crucial to improve the safety of this drug. The risk of over-anticoagulation and bleeding is especially high before stable anticoagulation has been established. One way to minimize this risk would be to shorten the time to stable anticoagulation by tailoring the initial dose for each patient. The required warfarin dose, which can vary 20-fold among individuals, can be roughly estimated from clinical and demographic factors, such as age, body weight, concurrent disease, and drug and food interactions (3 ). A number of dosage algorithms that use clinical and demographic factors have been tested and are able to reduce the time to therapeutic anticoagulation (4 ). More recent discoveries have shown that variation in the genes that encode the main enzyme responsible for S-warfarin metabolism (CYP2C9, cytochrome P450, family 2, subfamily C, polypeptide 9) and the target of warfarin (VKORC1, vitamin K epoxide reductase complex, subunit 1) influence dose requirements by affecting pharmacokinetics and pharmacodynamics (5 ). Polymorphisms in these genes are also associated with the risk of over-anticoagulation during initiation of warfarin therapy (6 ). A large prospective study on warfarin pharmacogenetics provided probabilities of overanticoagulation (INR 4) in patients with different CYP2C9 and VKORC1 alleles (Fig. 1) (7 ). During the first month of treatment, CYP2C9*3/*3 individuals had a 22-fold increased risk of an INR 4 and a tendency for more episodes of serious bleeding compared with individuals with CYP2C9*1/*1. Patients homozygous for VKORC1 variants had a 4.5-fold increased risk of an INR 4 within 5 weeks (7 ). Genotyping the CYP2C9 and VKORC1 genes could avert overdosing in patients who are warfarin sensitive because of these polymorphisms. Several pharmacogenetic algorithms that predict warfarin maintenance doses have been developed by combining genetic, clinical, and demographic factors with warfarin-dosing data and INR measurements (3 ). If genetic testing is integrated into routine warfarin therapy, it is estimated that American warfarin users would annually avoid 4500 to 22 000 serious bleeding events (8 ). The American regulatory agency, the Food and Drug Administration, decided to update the label of warfarin in 2007 to encourage lower initiation doses in patients with CYP2C9 and VKORC1 variant alleles. 1 Department of Medical Sciences, Clinical Pharmacology, Uppsala University Hospital, Uppsala, Sweden. * Address correspondence to the author at: Clinical Pharmacology, Uppsala University Hospital, Entrance 61, 3rd Floor, Uppsala, NA, Sweden SE-75185. Fax 46-18-6113703; e-mail [email protected]. Received August 18, 2008; accepted December 17, 2008. Previously published online at DOI: 10.1373/clinchem.2008.115964 2 Human genes: CYP2C9, cytochrome P450, family 2, subfamily C, polypeptide 9; VKORC1, vitamin K epoxide reductase complex, subunit 1. Clinical Chemistry 55:4 709–711 (2009) Point/Counterpoint