New Lancet analysis supports benefit of pharmacogenomic testing for warfarin

Published: May 15th, 2015

Category: Stories

CYP2C9 and VKORC1 genotypes identified individuals who required lower doses of warfarin and were more likely to be over-anticoagulated, according to a recent analysis of the ENGAGE AF–TIMI 48 trial by Mega and colleagues.1

ENGAGE AF–TIMI 48 was a prospective, randomized study that compared edoxaban 30 mg, edoxaban 60 mg, and warfarin titrated to a target INR of 2.0 to 3.0 (1:1:1 randomization) for the prevention of stroke and systemic embolism in 21,105 individuals with atrial fibrillation.2 Initial warfarin doses were assigned by the local investigator. In this analysis of ENGAGE AF–TIMI 48, investigators compared clinical events according to genotype in a subgroup of study patients (n = 14,348).

Patients were genotyped for CYP2C9 (*2 and *3) and VKORC1 (-1639G to A). Patients assigned to warfarin (n=4,833) were classified based on genotype results into normal warfarin responders (n = 2,982; 61.7%), sensitive responders (n = 1,711; 35.4%) and highly-sensitive responders (n = 140; 2.9%). Researchers found that patients classified as sensitive or highly-sensitive responders were more likely to be over-anticoagulated, with a greater portion of time with INR values >4.0 in the first 90 days (P < 0.0001). Within the 334 warfarin patients who had an overt bleeding event in the first 90 days, the rate of an overt bleeding event in the first 90 days was significantly higher in sensitive and highly sensitive responders as compared with normal responders. Genotype was not significantly associated with increased bleeding risk in patients randomized to edoxaban. However, both doses of edoxaban were associated with reduced bleeding in the first 90 days in highly sensitive responders to warfarin but not in normal.

Clinical Implications

Findings of this large analysis are important given mixed results of previous studies on the clinical benefit of genetic testing with warfarin. Investigators noted that study results demonstrated “clear and significant associations” between CYP2C9 and VKORC1 genotypes and warfarin bleeding outcomes and also supported the role of genetic data in complementing traditional clinical predictors of adverse effects with warfarin.

In an accompanying editorial3, Alan HB Wu, PhD, highlighted the importance of reimbursement for pharmacogenomic testing for warfarin outside of a clinical trial. “I hope that publication of Mega and colleagues’ findings will lead the Center for Medicare and Medicaid Services to reconsider the position on reimbursement so that individuals with multiple gene variants in CYP2C9 and VKORC1 can be offered alternative anticoagulant medications,” Wu wrote.

Find more information about warfarin pharmacogenomics and links to clinical dosing algorithms in the PharmGKB.


  1. Mega JL et al. Genetics and the clinical response to warfarin and edoxaban: findings from the randomised, double-blind ENGAGE AF-TIMI 48 trial. Lancet 2015;385:2280–7.
  2. Giugliano RP et al. Edoxaban versus warfarin in patients with atrial fibrillation. NEJM 2013;369:2093–104.
  3. Wu AHB. Pharmacogenomic testing and response to warfarin [Comment]. Lancet 2015;385:2231-2.