Warfarin Dosing and VKORC1/CYP2C9

Updated: Nov 10, 2015
  • Author: Kyong Chong, MD, PharmD; Chief Editor: Keith K Vaux, MD  more...
  • Print


Warfarin (Coumadin) is the most commonly used vitamin K antagonist. It has demonstrated effectiveness for the primary and secondary prevention of venous thromboembolism, for the prevention of systemic embolism in patients with prosthetic heart valves or atrial fibrillation, as an adjunct in the prophylaxis of systemic embolism after myocardial infarction, and for reducing the risk of recurrent myocardial infarction.

Anticoagulant therapy with warfarin is characterized by a wide inter-individual variation in dose requirements and a narrow therapeutic index. Therefore, accurate dosing is critical for safely managing patients on this drug. Because nongenetic influences such as body size and age are poor predictors of an individual's dose requirement, there has been considerable investigation into the genetic influences on warfarin dose requirements.

To see more information about warfarin, please go to the drug monograph by clicking here.

Warfarin is metabolized primarily via oxidation in the liver by CYP2C9, and exerts its anticoagulant effect by inhibiting the protein vitamin K epoxide reductase complex, subunit 1 (VKORC1). Three single nucleotide polymorphisms (SNPs), two in the CYP2C9 gene and one in the VKORC1 gene, have been found to play key roles in determining the effect of warfarin therapy on coagulation.

The nomenclature for the CYP2C9 SNPs is unique: the normal, or wild-type, variant is referred to as *1 ("star 1"), the two polymorphic versions are *2 ("star 2") and *3 ("star 3"), and each person can carry any two versions of the SNP. For example, a person with two normal copies would be *1/*1, a person with only one polymorphism could be *1/*2, and a person with both polymorphisms could be *2/*3. The prevalence of each variant varies by race; 10% and 6% of Caucasians carry the *2 and *3 variants, respectively, but both variants are rare (< 2%) in those of African or Asian descent. [1]

CYP2C9*1 metabolizes warfarin normally, CYP2C9*2 reduces warfarin metabolism by 30%, and CYP2C9*3 reduces warfarin metabolism by 90%. Because warfarin given to patients with *2 or *3 variants will be metabolized less efficiently, the drug will remain in circulation longer, so lower warfarin doses will be needed to achieve anticoagulation.

In the VKORC1 1639 (or 3673) SNP, the common G allele is replaced by the A allele. Because people with an A allele (or the "A haplotype") produce less VKORC1 than do those with the G allele (or the "non-A haplotype"), lower warfarin doses are needed to inhibit VKORC1 and to produce an anticoagulant effect in carriers of the A allele. The prevalence of these variants also varies by race, with 37% of Caucasians and 14% of Africans carrying the A allele. [2]

Recent genome wide association studies have not only confirmed these observations but also identified a novel association between rs2108622 in CYP4F2 and reduced hepatic CYP4F2, higher levels of hepatic vitamin K, and higher warfarin dose requirements. [3, 4, 5, 6, 7]


Clinical Implications of the Genetic Mutation

These three SNPs play key roles in determining (1) the dose of warfarin required to produce a therapeutic INR (typically 2.0 to 3.0); (2) the risk of bleeding or of producing supratherapeutic INR (>4.0); and (3) the time required to achieve a stable therapeutic dose. [8, 9, 10]

Carriers of CYP2C9*2 and CYP2C9*3 require, on average, a 19% and 33% reduction, respectively, per allele in warfarin dose vs those who carry the *1 allele. Carriers of the VKORC1 A allele require, on average, a 28% reduction per allele in their warfarin dose compared to those who carry none. [11, 12]

As expected, using standard dosing algorithms in patients with these variants leads to adverse clinical and laboratory outcomes because of their genetically mediated sensitivity to the drug. In particular, standard dosing algorithms lead, on average, to a 2- to 3-fold increased risk of serious or life threatening bleeding or an out-of-range INR (>4.0) in carriers of the *2 or *3 alleles of CYP2C9. [11] Similarly, carriers of the VKORC1 A allele are also at a 2- to 3-fold higher risk of an INR >4.0 during initiation of warfarin therapy when standard dosing algorithms are used. [12]

As the result of this patient population sensitive to warfarin and variability, the time required to achieve a "stable" INR between 2.0 and 3.0 is significantly delayed in carriers of all three SNPs. [11, 12] Overall, using a combination of genetic and clinical factors to predict the maintenance warfarin dose appears to be more accurate than using clinical factors alone. [13]


Based on the influence of these SNPs and the observations that carriers of certain alleles are at higher risk for adverse clinical and laboratory outcomes with standard warfarin dosing algorithms, the FDA updated the prescribing information for warfarin in January 2010 with specific recommendations for dosage range initiation in carriers of the CYP2C9 and VKORC1 variants. These recommendations are listed below.

Table 1. FDA Warfarin Dosage Recommendations (ie Expected Therapeutic Dosage Ranges) for Carriers of the CYP2C9 and VKORC1 Variants [14] (Open Table in a new window)



CYP2C9 Variant
  *1/*1 *1/*2 *1/*3 *2/*2 *2/*3 *3/*3
GG 5-7 mg 5-7 mg 3-4 mg 3-4 mg 3-4 mg 0.5-2 mg
AG 5-7 mg 3-4 mg 3-4 mg 3-4 mg 0.5-2 mg 0.5-2 mg
AA 3-4 mg 3-4 mg 0.5-2 mg 0.5-2 mg 0.5-2 mg 0.5-2 mg


Two attempts have been made to improve laboratory outcomes by initiating warfarin therapy using a pharmacogenetics-guided approach. The first study only used the CYP2C9 SNPs and showed that the 95 patients who were randomized to pharmacogenetics-based therapy achieved a stable INR significantly sooner than did the 96 patients given standard warfarin therapy. The second study tailored the dose to all three SNPs, but failed to show any significant advantage of a pharmacogenetic-guided approach with respect to their primary endpoint of percent out-of-range INRs. Nevertheless, they did show that the pharmacogenetic approach more accurately approximated stable doses with smaller and fewer dosing changes and INRs. [15, 16, 17]

In patients where genetic information is not available but are at increased risk of bleeding with standard dosing algorithms, guidelines suggest starting with a reduced (2-5 mg) initial dose and basing the frequency of monitoring on the INR response. [18]