The relationship of renal function to outcome: A post hoc analysis from the EdoxabaN versus warfarin in subjectS UndeRgoing cardiovErsion of Atrial Fibrillation (ENSURE-AF) study

Research output: Contribution to journalArticlepeer-review


  • Gregory Lip
  • Naab Al-Saady
  • Michael D. Ezekowitz
  • Maciej Banach
  • Andreas Goette

Colleges, School and Institutes

External organisations

  • Cardiovascular/Metabolic, Covance Inc, Osprey House, Maidenhead Park Office
  • Department of Cardiovascular Medicine, Sidney Kimmel Jefferson Medical College at Thomas Jefferson University and Lankenau Medical Center
  • Department of Hypertension, MUL
  • St. Vincenz-Hospital
  • Working Group: Molecular Electrophysiology, University Hospital Magdeburg


Background: The ENSURE-AF study (NCT 02072434) of anticoagulation for electrical cardioversion in non-valvular atrial fibrillation (NVAF) showed comparable low rates of bleeding and thromboembolism between the edoxaban and the enoxaparin–warfarin treatment arms. This post hoc analysis investigated the relationship between renal function and clinical outcomes. Methods: ENSURE-AF was a multicenter, PROBE evaluation trial of edoxaban 60 mg, or dose reduced to 30 mg/day for weight ≤60 kg, CrCl (creatinine clearance; Cockroft-Gault) ≤50 ml/min, or concomitant P-glycoprotein [P-gp] inhibitors) compared with therapeutically monitored enoxaparin–warfarin in 2,199 NVAF patients undergoing electrical cardioversion. Efficacy and safety outcomes and time in therapeutic range (TiTR) in the warfarin arm were analyzed in relation to CrCl in prespecified ranges ≥15 and ≤30; >30 and ≤50; >50 and <80 and ≥80 ml/min, and an exploratory ≥95 ml/min analysis. Results: 1,095 subjects were randomized to edoxaban and 1,104 to enoxaparin– warfarin. Mean age was 64.3±10 and 64.2±11 years. Mean TiTR was progressively lower with reducing CrCl strata, being 66.8% in those with CrCl >30 to ≤50, compared to 71.8% in those with CrCl ≥80. The Odds Ratios for the primary efficacy and safety endpoints were comparable for the different predefined renal function strata; given the small numbers the 95% CI included 1.0. In the subset of those with CrCl ≥95, the Odds Ratios showed consistency with the other CrCl strata. When CrCl was assessed as a continuous variable, there was a non-significant trend towards higher major or clinically relevant non-major bleeding with reducing CrCl levels, with no significant differences between the 2 treatment arms. When we assessed CrCl at baseline compared to end of treatment, there were no significant differences in CrCl change between the edoxaban and enoxaparin–warfarin arms. The proportions with worsening of renal function (defined as a decrease of >20% from baseline) were similar in the 2 treatment arms. Conclusion: Given the small number of events in ENSURE-AF, no effect of renal (dys)function was demonstrated in comparing edoxaban to enoxaparin–warfarin for cardioversion; efficacy and safety of edoxaban remained consistent even in patients with normal or supranormal renal function.


Original languageEnglish
Pages (from-to)16-22
JournalAmerican Heart Journal
Early online date20 Jul 2017
Publication statusPublished - 1 Nov 2017


  • cardioversion , oral anticoagulant , renal function , stroke , thromboembolism