Mobile Site ›
Communique Print the PDF of this entire issue

Warfarin: Genotyping and Improving Dosing


Receive notification when new Communiqué articles are published:

August 2008


Warfarin, an anticoagulant developed over 50 years ago, is now one of the most commonly prescribed drugs in the United States. New prescriptions are given to an estimated 2 million patients annually to treat or prevent venous thrombosis, pulmonary embolism, and thromboembolic complications from atrial fibrillation and cardiac valve replacement.1 Warfarin is indicated to reduce the risk of death from recurrent heart attacks or thromboembolic events such as stroke2 and also is prescribed postsurgically for open heart procedures or orthopedic surgeries.3 As atrial fibrillation and its associated risk of stroke increases with an aging population, more and more patients will likely require anticoagulant treatment.4

Warfarin-associated adverse drug events (ADEs) are the leading cause for emergency room visits, ranking second only to insulin-associated ADEs. The most common ADEs associated with warfarin are bleeding events or formation of blood clots, as a result of overdosing or underdosing, respectively. Each year, warfarin-associated ADEs cost the United States health care system between $100 million and $2 billion.1

Although warfarin is very effective, achieving a stable, steady-state dose is difficult due to varied individual responses to the drug, requiring significant patient management to avoid serious ADEs.3 It is now recognized that patients who are sensitive to warfarin can be identified using recently developed genetic tests to identify patients who may be at risk for warfarin-related ADEs. This new genetic information can assist physicians to determine the most appropriate dosing strategy for the patient, improving patient outcome and reducing episode of care costs.

Standard Warfarin Dosing

While warfarin is effective, regulating warfarin dosing is challenging because of the drug’s narrow therapeutic range. Often, the difference between a therapeutic and a toxic dose is small. Due to this narrow therapeutic range, patients must undergo frequent coagulation testing (ie, prothrombin time, or PT) to monitor their response to the drug. Because PT assays are subject to significant variation across laboratories and testing systems due to differences in reagent sensitivities, the INR (international normalized ratio) was developed and is now the standard method for monitoring warfarin response.5 The INR is determined using the following calculation, providing more comparable results across different assay systems5:

INR = (patient PT/mean normal PT)ISI

ISI = international sensitivity index

To avoid bleeding or clotting incidents, it is critical to obtain a stable INR within the therapeutic range (in most situations 2.0–3.0, although it can be higher in high-risk patients).5 Anything below the target INR puts patients at risk for developing blood clots; anything higher puts them at risk for bleeding.3 When warfarin therapy is first initiated, patients must undergo frequent blood testing to achieve an INR in the correct range. Once stabilized, testing can be scheduled less frequently; the most common practice is monthly INR monitoring of stabilized patients.

While physicians have routinely determined dosages based on age, gender, and weight, genetic differences have now been shown to account for a significant amount of variability in individual dosing requirements. For example, because of differences in genetic makeup, one patient may require a higher than standard dose, while another patient may require a lower one. Additionally, other external factors such as drug-drug interactions, nutritional status, health habits (eg, diet, smoking, and alcohol use), and the presence of other illnesses affect how much warfarin each individual needs to maintain his or her anticoagulation levels.6 Because these external factors can change, on-going monitoring is necessary, even after a stable PT/INR has been achieved.

Risks of Warfarin Use

The most serious risks are bleeding, strokes, and death. Warfarin-associated intracerebral hemorrhage (ICH), which has a 50% mortality rate, is one of the most severe ADEs. A 2007 Mayo Clinic study estimated the number of warfarin-related ICH ADEs at 8,000 to 10,000 each year.8 From January 1993 to June 2006, of the warfarin-associated bleeding events reported to FDA’s Adverse Events Reporting System, 10% were recorded as fatal.7 According to the data collected, as the number of warfarin prescriptions has increased, so has the number of reported ADEs.7

Warfarin Metabolism

Warfarin consists of S- and R-isomers; S-warfarin is the more potent form.3 When given orally, warfarin is absorbed into the gastrointestinal system. As it circulates through the bloodstream, warfarin is bound to albumin. The drug is metabolized by enzymes in the liver, with the S- and R-isomers altered via different paths. Warfarin acts by interfering with the metabolism of vitamin K, which is necessary for production of key coagulation factors. Warfarin inhibits vitamin K recycling by blocking its metabolism at the vitamin K-epoxide intermediate; this process decreases the amount of available vitamin K, creating an anticoagulant effect.5

The process of warfarin metabolism is extremely complex and involves both pharmacokinetics and pharmacodynamics.5 Pharmacokinetics involves the rate at which the drug is absorbed, distributed, eliminated, and transformed in the body. In other words, it’s what happens to the drug in the body. Pharmacodynamics involves how the drug response changes during those processes, as well as the systemic effects of the drug.5

The CYP2C9 enzyme is responsible for metabolizing S-warfarin to inactive products (pharmacokinetics), and is coded for by the CYP2C9 gene. A second gene, VKORC1, encodes the enzyme vitamin K epoxide reductase complex subunit 1 (VKORC1), which is part of the vitamin K cycle and the target of warfarin therapy (pharmacodynamics).6 Variants (polymorphisms) within these 2 genes affect warfarin metabolism and the subsequent dose needed to maintain a stable INR (Figure 1). As much as 40% of dosing variability among patients may be due to genetic differences (for both warfarin resistance and warfarin sensitivity).6

Pharmacokinetic Pathway
Pharmacokinetic metabolism of warfarin involves the CYP2C9 gene,3 which is one of the cytochrome (CYP) P450 genes. This group of genes codes for enzymes responsible for metabolizing more than half of the drugs in use today. The CYP2C9 gene encodes a liver enzyme (CYP2C9) that metabolizes the more active isomer of warfarin (S-warfarin) to inactive products. Several polymorphisms of the CYP2C9 gene have been identified that decrease the activity of the enzyme. This decreased enzyme activity, in turn, may increase serum warfarin levels, resulting in overmedicating the patient and increasing the INR above the therapeutic target level, and accounting for increased bleeding incidents in some patients.3 In these patients, lower doses may need to be prescribed.5

Pharmacodynamic Pathway
Warfarin’s pharmacodynamic pathway involves VKORC1. In the normal coagulation process, factors II, VII, IX, and X undergo gamma glutamyl carboxylation to become active. Gamma glutamyl carboxylation requires the presence of vitamin K; hence, factors II, VII, IX, and X are referred to as the vitamin K-dependent coagulation factors. Vitamin K epoxide (oxidized vitamin K) is a by-product of gamma carboxylation. In turn, VKORC1 replenishes vitamin K. Anticoagulation occurs as warfarin inhibits VKORC1, decreasing the amount of vitamin K, and resulting in decreased levels of functional vitamin K-dependent coagulation factors.6 (Figure 1)

The VKORC1 gene encodes the VKORC1 enzyme, a small transmembrane protein of the endoplasmic reticulum. VKORC1 is primarily transcribed in the liver, although it is present in smaller amounts in the heart and pancreas. A polymorphism within the promoter of VKORC1, specifically, a guanine to adenine substitution (GA) at position -1639, decreases expression of the gene. Lower levels of enzyme are synthesized, decreasing the availability of vitamin K. Thus, a reduced warfarin dose is needed to compensate for the effects of this polymorphism in order to maintain the target INR.5 Research has shown that the VKORC1-1639A genotype occurs in African Americans (4% to 5%), Caucasians (14% to 17%), and Asians (72% to 78%). For Asians, the research findings explain the higher incidence of warfarin sensitivity.3

Vitamin K cycle and warfarin metabolism
Figure 1. Vitamin K cycle and warfarin metabolism

Drug Interactions
Drug-drug and drug-metabolite interactions are a concern for physicians who are monitoring patients on multiple medications. Because many drugs are metabolized by CYP P450 enzymes, the way these drugs interact can affect how much warfarin an individual patient will require to reach a stable INR. Drugs that are metabolized by CYP2C9 compete with warfarin for available enzyme, while other drugs increase or decrease CYP2C9 activity.

For example, when taken with warfarin, the medication lovastatin may decrease warfarin’s rate of metabolism, possibly increasing warfarin toxicity and necessitating a lower dose. Conversely, someone taking rifampin may require higher doses of warfarin. These effects can occur even without the CYP2C9 polymorphisms associated with altered warfarin metabolism.

Table 1 shows some of the drugs that undergo metabolism by CYP2C9, or decrease or increase CYP2C9 activity. Combining these drugs with warfarin may alter the rate of warfarin elimination, perhaps increasing the risk of overdosing.

Herbal Supplements and Food Interactions
Studies indicate that dietary intake and herbal use affect metabolic and pharmacologic interactions with warfarin. The warfarin label lists several botanical medications (or herbals) that have been reported to affect the level of anticoagulation, including:2

  • Increase effect: garlic, ginkgo biloba, ginseng, and cranberry
  • Decrease effect: St. John’s wort and coenzyme Q10
  • Interfere with warfarin’s effectiveness: vitamin K-rich foods, including green leafy vegetables such as spinach and broccoli
  • Independent anticoagulant effect: aniseed, capsicum, cassia, celery, cinnamon, licorice, parsley, wild carrot, and wild lettuce

The drug labeling does not indicate that patients should avoid eating these foods and supplements, but it cautions against changing the diet and increasing the amounts without notifying the physician. Also, patients need to inform their physician about any herbal supplements they are taking.2

Genetic Testing

Mayo Medical Laboratories recently introduced #89033 Warfarin Sensitivity Genotyping, which is used to identify variants in both the CYP2C9 and VKORC1 genes. Physicians may elect to utilize this test in several situations:

  • For those patients who already take warfarin, but who require numerous modifications in their prescribed dosing to obtain a stable INR, such testing may identify the source of variability in response (genetic vs environmental causes)
  • For those patients who have had bleeding or clotting incidents while receiving warfarin therapy
  • For patients who are prescribed warfarin therapy for the first time, either before or shortly after initiating therapy (eg, preoperative genotyping), especially if there is a family history of difficulties with warfarin use

Results from the test can help minimize risks for patients who take warfarin. When the INR is adequately controlled, patient safety improves because the risks for bleeding incidents or stroke are reduced.

Table 2 indicates the relationship between polymorphisms detected by #89033 Warfarin Sensitivity Genotyping and the effect on the activity of the enzyme encoded for by that allele.

A person who is CYP2C9 homozygous wild type is considered an extensive metabolizer, meaning a normal metabolizer of warfarin. Patients whose CYP2C9 results indicate reduced enzyme activity require lower doses of warfarin.9 The -1639GA polymorphism reduces VKORC1 expression; a GA or AA (vs GG) genotype leads to a decrease in mRNA expression in the liver (and a decrease in enzyme production). Patients with this variation also usually require reduced doses of warfarin. Individuals who have polymorphisms on both VKORC1 and CYP2C9 usually require even lower doses of warfarin to maintain the INR in the target range.

Mayo’s report includes a discussion of the test findings and indicates the impact on warfarin metabolism for the patient.

Type of Drug
Generic Drug Name
Drugs metabolized by CYP2C9 that may affect warfarin metabolisma
Fluconazole, isoniazid, metronidazole,
sulfamethoxazole, voriconazole
Warfarin (more active S-isomer)
Fluoxetine, paroxetine, sertraline
Other drugs
Fenofibrate, fluvastatin, lovastatin,
phenylbutazone, probenecid, phenytoin,
tamoxifen, torsemide, zafirlukast
Drugs that may significantly increase warfarin metabolismb
Rifampin, secobarbital
Drugs that may significantly decrease warfarin metabolismc
Amiodarone, atorvastatin, fenofibrate, fluconazole, fluvastatin, isoniazid, lovastatin, simvastatin, ticlopidine, voriconazole
Table 1. CYP2C9-related drug interactions
a Coadministration of these drugs may decrease the rate of elimination of other drugs metabolized by CYP2C9
b Coadministration of these drugs induces the synthesis of CYP2C9, resulting in increased CYP2C9 activity and metabolism of warfarin. A dose increase may be needed to maintain the INR in the target range.
c Coadministration of these drugs may decrease the rate of metabolism of CYP2C9-metabolized drugs, including warfarin, increasing the possibility of toxicity. The drug labeling should be consulted for additional information.

FDA Warfarin-Labeling Requirement Changes
In August 2007, the Food and Drug Administration (FDA) introduced new labeling to reflect the information available about genotypes and warfarin. The ADEs are well-known for this drug, and the updated labeling cites results from clinical studies where ADEs were recorded. In addition, the labeling in the Precaution section has been updated to inform physicians that gene variations for their patients may influence how patients respond to warfarin therapy.2

Estimates of Cost Effectiveness

In 2006, McWilliam et al from the AEI Brookings Joint Center for Regulatory Studies published a working paper that summarizes research from available adverse event data for warfarin and the estimated cost of those ADEs.1 The paper estimates the health benefits and reduction in health care costs if patients undergo genetic testing. The authors believe that the results of the FDA Critical Path Initiative will show that implementing genetic testing can reduce health care costs for warfarin-related ADEs by $1.1 billion annually.1

Estimated hospitalization costs for either a bleeding or a clotting ADE are $18,000 to $25,000 for each patient.9 For those patients who suffer strokes while taking warfarin, estimated costs for hospitalization and medical care are ~ $40,000 per patient. If 2 million patients start warfarin each year, and an estimated 22% are hospitalized for ADEs, the overall health care costs can be astounding. McWilliam et al surmise that with genetic testing followed by appropriate dosing, the risk of stroke could be reduced 50% and costs could be reduced significantly.1 While the AEI Brookings report predicted significant financial benefits for utilizing warfarin genotyping, further studies are needed to determine the actual benefit of genotype-based dosing regimens.

FDA Critical Path Initiative
The FDA is very interested in the safe use of warfarin because of the numerous ADEs reported in the last several years,1 and has included warfarin safety as one of its Critical Path Initiatives. These initiatives are partnerships designed to incorporate the latest scientific knowledge and techniques into developing products that are safe and effective.10 For warfarin, the Critical Path Initiative includes funding for research studies to develop dosing guidelines and algorithms based on genetic information.

CYP2C9 Allele Nucleotide Change Effect on Enzyme Metabolism
*1 None (wild-type) Extensive metabolizer (normal)
*2 430CT Reduced activity
*3 1075AC Minimal activity
*4 1076TC Reduced activity
*5 1080CG Reduced activity
*6 8181delA No activity
VKORC1 Nucleotide Change Effect on Enzyme Metabolism
  1639GA Reduced synthesis
Table 2. CYP2C9 and VKORC1—Polymorphisms and effect on enzyme activity

Medco-Mayo Study
Pharmacy benefits management businesses also are interested in safe dosing for warfarin. Mayo Clinic has partnered with one large benefits management organization, Medco, in a prospective study designed to improve the care and safety of patients who are prescribed warfarin.9 This trial will also provide guidance concerning the best way to use the data that is available and whether the proposed dosing suggestions ( are appropriate. The study objective is to determine the economic benefits of genetic testing where the results may help physicians determine an effective starting dose for patients.9 Some argue that until clinical trials are completed, genetic testing should not be used. Others argue that such testing may be advantageous prior to the overall trial results, especially for patients with a history of difficulties in titrating warfarin or if bleeding would be a particular concern. Other candidates for genotyping are individuals on other drugs known to interfere with warfarin pharmacokinetics.


Warfarin therapy is a double-edged sword; its benefits are clear, but the cost of adverse events is high. Studies are currently under way to examine the potential to increase patient safety and reduce medical costs through the use of genetic tests for warfarin sensitivity. While large-scale studies are ongoing, developing familiarity by using the testing in high-risk individuals is likely to be of benefit.

Authored by Shindley L, Dale JC, Masoner DE, Moyer TP, Jaffe AS, O’Kane DJ

Note: A recently published study (Caraco et al) further substantiates the benefit of warfarin genotyping; following a prescribed warfarin dosing algorithm, study subjects dosed based on genotype achieved therapeutic INR faster and experienced fewer bleeding events than study subjects dosed using a standardized (non-genomic) warfarin dosing protocol.

  • Caraco Y, Blotnick S, Muszkat M: CYP2C9 genotype-guided warfarin prescribing enhances the efficacy and safety of anticoagulation: A prospective randomized controlled study. Clin Pharmacol Ther 2008;83:460-470
  • Caraco Y, Bejarano-Achache I, Shaoul H, et al: Combined CYP2C9-VKORC1 genotype guided warfarin loading enhances the efficacy and safety of anticoagulation. Preliminary findings from a prospective, double blind, randomized, controlled study. Clin Pharmacol Ther 2008;83:S8-S8


  1. McWilliam A, Lutter R, Nardinelli C: Health care savings from personalizing medicine using genetic testing: The case for warfarin. Working Paper 06-23. AEI Brookings Joint Center for Regulatory Studies 2006 November
  2. US Food and Drug Administration: Warfarin product label; Bristol-Myers Squibb Company; Accessed March 3, 2008
  3. Baudhuin LM: Warfarin pharmacogenetics: The challenge of laboratory testing. Clin Lab News 2008 February
  4. Miyasaka Y, Barnes ME, Cha SS, et al: Risk and trends of recurrent ischemic stroke following incident atrial fibrillation: data from 2 decades (1980-2000) (Abstract 1068-111). J Am Coll Cardiol 2005 Feb;45(3 Suppl A):387A
  5. Hirsh J, Fuster V, Ansell J, Halperin JL: American Heart Association / American College of Cardiology Foundation Guide to Warfarin Therapy. Circulation 2003;107:1692-1711
  6. Flockhart D, O’Kane D, Williams M, Watson M: Pharmacogenetic testing for CYP 2C9 and VKORC1 alleles for warfarin use: An ACMG position paper. Genet Med 2008;10(2):139-150
  7. Wysowski DK, Nourjah P, Swartz L: Bleeding complications associated with warfarin use: a prevalent adverse effect resulting in regulatory action. Arch Intern Med 2007;167(13):1414-1419
  8. Aguilar MI, Hart RG, Kase CS, et al: Treatment of Warfarin-Associated Intracerebral Hemorrhage: Literature Review and Expert Opinion. Mayo Clin Proc 2007;82(1):82-92
  9. Ratner ML: Medco: Market multiplier for personalized medicine. In Vivo 2007;25(11):4
  10. US Food and Drug Administration: Critical path initiative: warfarin dosing. Accessed March 4, 2008. Available at: