Genetics and the clinical response to warfarin and edoxaban: findings from the randomised, double-blind ENGAGE AF-TIMI 48 trial

Lancet. 2015 Jun 6;385(9984):2280-7. doi: 10.1016/S0140-6736(14)61994-2. Epub 2015 Mar 11.


Background: Warfarin is the most widely used oral anticoagulant worldwide, but serious bleeding complications are common. We tested whether genetic variants can identify patients who are at increased risk of bleeding with warfarin and, consequently, those who would derive a greater safety benefit with a direct oral anticoagulant rather than warfarin.

Methods: ENGAGE AF-TIMI 48 was a randomised, double-blind trial in which patients with atrial fibrillation were assigned to warfarin to achieve a target international normalised ratio of 2·0-3·0, or to higher-dose (60 mg) or lower-dose (30 mg) edoxaban once daily. A subgroup of patients was included in a prespecified genetic analysis and genotyped for variants in CYP2C9 and VKORC1. The results were used to create three genotype functional bins (normal, sensitive, and highly sensitive responders to warfarin). This trial is registered with, number NCT00781391.

Findings: 14,348 patients were included in the genetic analysis. Of 4833 taking warfarin, 2982 (61·7%) were classified as normal responders, 1711 (35·4%) as sensitive responders, and 140 (2·9%) as highly sensitive responders. Compared with normal responders, sensitive and highly sensitive responders spent greater proportions of time over-anticoagulated in the first 90 days of treatment (median 2·2%, IQR 0-20·2; 8·4%, 0-25·8; and 18·3%, 0-32·6; ptrend<0·0001) and had increased risks of bleeding with warfarin (sensitive responders hazard ratio 1·31, 95% CI 1·05-1·64, p=0·0179; highly sensitive responders 2·66, 1·69-4·19, p<0·0001). Genotype added independent information beyond clinical risk scoring. During the first 90 days, when compared with warfarin, treatment with edoxaban reduced bleeding more so in sensitive and highly sensitive responders than in normal responders (higher-dose edoxaban pinteraction=0·0066; lower-dose edoxaban pinteraction=0·0036). After 90 days, the reduction in bleeding risk with edoxaban versus warfarin was similarly beneficial across genotypes.

Interpretation: CYP2C9 and VKORC1 genotypes identify patients who are more likely to experience early bleeding with warfarin and who derive a greater early safety benefit from edoxaban compared with warfarin.

Funding: Daiichi Sankyo.

Publication types

  • Randomized Controlled Trial
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Anticoagulants / administration & dosage
  • Anticoagulants / adverse effects
  • Anticoagulants / therapeutic use*
  • Atrial Fibrillation / complications
  • Atrial Fibrillation / drug therapy*
  • Atrial Fibrillation / genetics*
  • Cytochrome P-450 CYP2C9 / genetics
  • Double-Blind Method
  • Factor Xa Inhibitors / administration & dosage
  • Factor Xa Inhibitors / therapeutic use*
  • Genetic Variation
  • Genotype
  • Hemorrhage / chemically induced
  • Humans
  • International Normalized Ratio
  • Pharmacogenetics
  • Pyridines / administration & dosage
  • Pyridines / therapeutic use*
  • Risk Assessment
  • Stroke / prevention & control
  • Thiazoles / administration & dosage
  • Thiazoles / therapeutic use*
  • Vitamin K Epoxide Reductases / genetics
  • Warfarin / administration & dosage
  • Warfarin / adverse effects
  • Warfarin / therapeutic use*


  • Anticoagulants
  • Factor Xa Inhibitors
  • Pyridines
  • Thiazoles
  • Warfarin
  • Cytochrome P-450 CYP2C9
  • VKORC1 protein, human
  • Vitamin K Epoxide Reductases
  • edoxaban

Associated data