The purpose of the investigators' study is to determine the clinical utility of a warfarin-dosing algorithm that incorporates genetic information (VKORC1 and CYP2C9 alleles) for adult patients initiating warfarin therapy.

Almost 20 million prescriptions are written for warfarin each year in the US and yet it is one of the most problematic drugs in the modern medical formulary. Warfarin has a narrow therapeutic window and the hemorrhagic or thrombotic implications of modest over- or under-dosing can be devastating. Warfarin is one of the leading causes of emergency department visits and hospitalizations due to adverse drug events worldwide. Adverse events from warfarin are more common during the initial months of treatment before the optimal dose is determined. Moreover, there is substantial individual variation in response to warfarin necessitating frequent monitoring and dosage adjustments. The monitoring and dosing of warfarin is so problematic that many primary care physicians have abdicated this role to specialized "warfarin clinics" which are devoted solely to following patients on this agent. Unfortunately, no good alternatives to warfarin exist for the common indications requiring chronic anticoagulation such as atrial fibrillation, deep vein thrombosis, pulmonary embolism, and artificial heart valves.

Pharmacogenomics offers substantial hope for improved care of patients taking warfarin. One group estimated that formally integrating genetic testing into routine warfarin therapy in the US could result in the avoidance of 85,000 serious bleeding events and 17,000 strokes annually with a cost savings of over $1 billion annually. Common single nucleotide polymorphisms (SNPs) in the gene encoding Vitamin K Epoxide Reductase (VKOR) substantially affect one's sensitivity to warfarin, mediating a doubling or halving of the dose required for optimal anticoagulation. Warfarin inhibits clotting by inhibiting the enzyme VKOR, and thus inhibiting vitamin K dependent clotting factors. A number of recent retrospective studies have shown that polymorphisms in the VKOR gene may account for 20-30% of the variance in warfarin dose seen in patients on stable, long-term warfarin therapy.

Another genetic determinant of variance in warfarin dose is the cytochrome p450 2C9 enzyme CYP2C9. It is almost wholly responsible for metabolism of the more active S-enantiomer of warfarin. The 2C9*2 and 2C9*3 polymorphisms in the CYP2C9 gene are associated with reduced warfarin metabolism, and a number of retrospective studies have shown that these polymorphisms may account for 10-15% of the variance in warfarin dosage in patients on stable, long-term warfarin therapy. In addition, the variant CYP2C9 alleles have been associated with longer times to stabilization of INR and a higher risk for bleeding events. These polymorphisms are seen in ~20-30% of the Caucasian population, but are rare in African Americans and Asians. Together, known VKOR and CYP2C9 variants may account for 40% of the variability in warfarin dosing.

By combining clinical information such as weight, height, and concomitant medications with VKOR and CYP2C9 genotypic information, several algorithms have been devised that calculate warfarin doses. These algorithms have been shown to accurately predict warfarin doses in retrospective studies of patients already on long-term stable warfarin doses. Some small, pilot studies in orthopedic patients suggest that prospective genetic-based dosing is feasible and may result in achieving stable doses sooner in patients with certain genetic variants. However, the prospective studies are small, pilot studies limited to orthopedic patients that did not include medical patients with common indications requiring chronic oral anticoagulation. They are also limited by study designs that include only historical controls. No RCTs have been reported in the literature and further evaluation is needed to determine the utility and cost-effectiveness of genetic-based algorithms. The NHLBI is planning a double-blind, randomized three-arm trial, but the trial will not begin enrolling subjects until 2008 at the earliest and data analysis and dissemination is planned to begin beyond 2011.

Inclusion Criteria:

• Adults ≥18 years old
• Patients who are beginning warfarin for a variety of diseases or conditions that require long-term oral anticoagulation with target INR > 2.0 for at least 3 months
• Subjects that will be following up in UNC anticoagulation clinics at the Ambulatory Care Center or the Family Medicine Center

Exclusion Criteria:

• Patients who are unable to complete the study materials (questionnaires) with or without assistance (for example, those with dementia)
• Non-English speaking patients
• Patients who are being started on anticoagulation intended to last < 3 months or whose target INR is < 2.0
• Patients who have a history of treatment with warfarin and a known dose requirement will be excluded (as they should be restarted on the previous dose)
• Pregnant women will be excluded because warfarin is a teratogen and pregnant women should not take the medication
• Patients will also be excluded if their treating physician does not agree to use the recommended INR dose or feels that the patient should not be enrolled in the study