The Pharmacogenetics of Warfarin: Insights from COAG and EU-PACT

Editor's Note: Based on

  1. Kimmel SE, French B, Kasner SE, et al. Pharmacogenetic versus a clinical algorithm for warfarin dosing. N Engl J Med 2013;369:2283-93.
  2. Pirmohamed M, Burnside G, Eriksson N, et al. A Randomized Trial of Genotype-Guided Dosing of Warfarin. N Engl J Med 2013;369(24):2294-303.

Background

A goal of therapy with Vitamin K antagonists (VKAs) is to keep the international normalized ratio (INR) in the therapeutic range to reduce thrombotic and bleeding risk. Polymorphisms in two genes CYP2C9 (involved in the metabolism of warfarin S-enantiomer) and VKORC1 (involved in the vitamin K cycle) together with age and body- surface area, account for about 50% of the variability in VKAs dose requirements.1

To date, genotyping before prescription of warfarin has not been recommended in clinical practice guidelines2 because of the lack of data from randomized trials. Thus, two studies, the EU-PACT (European Pharmacogenetics of Anticoagulant Therapy) 3 and the COAG (Clarification of Optimal Anticoagulation through Genetics) 4 trials, were undertaken, to evaluate whether genotyping in patients initiating VKAs would improve anticoagulation control.

Methods

The EU-PACT study from the United Kingdom and Sweden, and the COAG trial, from the United States, are two multicenter, randomized, controlled studies involving patients with an indication for oral anticoagulation therapy.

The single-blinded EU-PACT trial randomized 455 patients naive to VKAs therapy. In the genotype-guided group, warfarin doses were prescribed according to a pharmacogenetic-based algorithm for the first five days. The control group received a standard loading dose. After the initiation period, the treatment was managed according to routine clinical practice during a follow-up period of three months.

The double-blind COAG trial enrolled 1015 patients, randomly assigned to receive doses of warfarin during the first 5 days that were determined according to an algorithm, which included both clinical variables and genotype data, or to an algorithm that included clinical variables only. The doses after day five were determined according to usual local clinical practice, with a follow up period of 28 days.

In both trials the primary outcome was the percentage of time in the therapeutic range (TTR) for INR during 12 weeks for the EUPACT trial and within four weeks for the COAG trial.

Results

In the EU-PACT trial, the TTR during the 12 weeks was significantly higher for the genotype-guided group compared to the control group (Table 1).

Table 1. EU-PACT: Primary Endpoint
Endpoint
(TTR at 12 weeks)
Genotype-Guided Group (%) Control Group (%)

Adjusted Difference (Percentage Points; 95% CI)

P Value

All patients

67.4

60.3

7.0 (3.3 to 10.6)

<0.001

In the COAG trial during the 4 weeks, there was no significant difference in the mean percentage of TTR between the two groups, however there was a significant interaction between race and dosing strategy. Among African American patients the mean percentage of TTR was lower in the genotype than in the clinically guided group, while there were no differences for non-African Americans between groups (Table 2).

Table 2. COAG: Primary Endpoint

Endpoint
(TTR at 4 weeks)
Genotype-Guided Group (%) Clinically Guided Group (%)

Adjusted Mean Difference (Percentage Points 95% CI)

P Value

All patients

45.2

45.4

-0.2 (-3.4 to 3.1)

0.91

Race

     

0.03

   African Americans

35.2

43.5

-8.3 (-15 to -2.0)

0.01

   Non-African Americans

48.8

46.1

2.8 (-1.0 to 6.6)

0.15

Conclusions

The two trials have come to divergent conclusions; in the EU-PACT trial, pharmacogenetic-based dosing was associated with a higher percentage of TTR than was standard dosing; in the COAG trial, genotype-guided dosing of warfarin did not improve anticoagulation control.

Commentary/Perspective

Although these two trials vary in organization and structure (duration of study, vitamin K antagonist used, double- versus single-blind design, racial characteristics, method for dosing in the control group and timing of primary endpoint), they also share similar features (large, multicenter, randomized studies; genotyping of CYP2C9 and VKORC1; TTR as primary end point during the initial therapy).

Multiple issues can be considered to explain the different results between the two trials:

  1. While in the EU-PACT trial the genotype group was compared with standard dosing methods used in clinical practice, the COAG trial used clinical variables incorporated in an algorithm for the clinically-based dosing strategy, leading to a possible more precise control in the INR values.
  2. In the EU-PACT trial, all Day one dosing was determinate based on genotype, given that results were available within two hours. In the COAG trial, 99% of subjects had genetic information by Day three, thus the different timing of genotyping results availability could have altered the outcome.
  3. The different prevalence of the genetic variants could also have influence the divergent outcomes: in the EU-PACT trial, homozygous VKORC1 genotype was more represented than in the COAG trial (17% versus 11%), as well as homozygous carriers for the CYP2C9 (3.4% versus 1%).This is relevant because homozygous carriers are more sensitive to warfarin and require a lower daily dose 5.
  4. In the COAG trial, one-third of the patients were African-Americans, and the percentage of TTR within this population was lower in the genotype arm. This could be explained considering that the frequency of the VKORC1 and CYP2C9 alleles varies between different ethnic groups and, in African Americans, other genetic variants not assessed in this study contribute to the variability in patient response to warfarin 6.
  5. The TTR was determined within 12 weeks In the EU-PACT trial and during four weeks in the COAG trial. In the EU-PACT trial, at four weeks there was no significant difference between the genotype- and the clinically-guided group, showing that the different time of the primary endpoint assessment could have contributed to the outcomes difference.

Despite the introduction of new oral anticoagulants that do not depend on CYP2C9 or VKORC1 genetic variations, improved safety of VKAs is still useful. VKAs will continue to be used frequently because they are inexpensive and remain the treatment of choice for patients with mechanic valves and severe renal insufficiency. Although in both trials there were no differences in the rates of bleeding or thromboembolic events, these studies were not powered to assess these clinical outcomes.Therefore the most important question, whether pharmacogenetic testing affects clinical outcomes, has not been answered and further studies are needed to determine it.

References

  1. Holbrook A, Schulman S, Witt DM, et al. Evidenced-based management of anticoagulation therapy: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidenced-Based Clinical Practice Guidelines. Chest 2012;141:Suppl:e152S-e184S.
  2. Anderson JL, Horne BD, Stevens SM, et al. Randomized trial of genotype-guided versus standard warfarin dosing in patients initiating oral anticoagulation.Circulation 2007;116:2563-70.
  3. Pirmohamed M, Burnside G, Eriksson N, et al. A Randomized Trial of Genotype-Guided Dosing of Warfarin. N Engl J Med 2013;369(24):2294-303.
  4. Kimmel SE, French B, Kasner SE, et al. A Pharmacogenetic versus a Clinical Algorithm for Warfarin Dosing. N Engl J Med. 2013;369(24):2283-93
  5. Lindh J.D. Holm L. Andersson M.L. Rane A. Influence of CYP2C9 genotype on warfarin dose requirements—a systematic review and meta-analysis. EJCP2009;65(4):365-75.
  6. Parera MA, Cavallari LH, Limdi NA, et al. Genetic variants associated with warfarin dose in African-American individuals: a genome-wide association study.Lancet 2013;382(9894):790-6.

Clinical Topics: Anticoagulation Management, Arrhythmias and Clinical EP, EP Basic Science

Keywords: Anticoagulants, Blood Coagulation, Control Groups, Double-Blind Method, Genotype, Hemorrhage, International Normalized Ratio, Pharmacogenetics, Vitamin K, Vitamin K 1, Warfarin


< Back to Listings