WARCEF - Warfarin vs ASA in HFrEF in sinus rhythm

in

Homma S, et al. Warfarin and aspirin in patients with heart failure and sinus rhythm. N Engl J Med 2012;366:1859-69.

Bottom-line: In patients with HFrEF in sinus rhythm and without any additional risk factors for cardiac embolism, warfarin reduced the risk of ischemic stroke, and increased the risk of major and minor hemorrhages and HF hospitalization (possibly from anemia or cessation/reduction of HF meds during bleeding events).

For every 1000 patients with HFrEF in sinus rhythm, treatment warfarin instead of ASA would result in 7 fewer ischemic strokes, 11 extra HF hospitalizations, 9 extra major and 43 extra minor bleeds per year. Thus for most patients represented here, the risks far outweigh the benefits.

 

Patients (n=2305)

  • Inclusion
    • Age 18+ y
    • HFrEF
      • NYHA class I-IV
      • LVEF 35% or less assessed in the past 3 months
    • Normal sinus rhythm
    • Modified Ranking scale (mRS) <5 (i.e. no more than moderately severe disability)
    • Planned treatment with a beta-blocker plus ACEI/ARB/hydralazine+nitrate
  • Exclusion:
    • Clear indication for either warfarin or ASA
    • High risk of cardiac embolism
      • AF
      • Mechanical heart valve
      • Endocarditis
      • Intracardiac thrombus (mobile or pedunculated)
  • Screened ? -> randomized & analyzed 2305
  • Typical trial patient
    • Age 61 y
    • Male 80%
    • North American ~50%
    • NYHA class I (14%), II (55%), III (30%), IV (~1%)
    • mRS 0 (41%), 1 (31%), 2 (23%), 3-4 (5%)
    • Ischemic HF etiology 43%
    • PMHx
      • Stroke/TIA 13%
      • MI 48%
      • Current smoker 17%
      • ETOH consumption >2 onces/day - 25%
      • HTN 61%
      • Diabetes 32%
      • AF 4%
    • BMI 29
    • BP 124/74 mm Hg
    • HR 72 bpm
    • LVEF 25%
    • Meds
      • ACEI/ARB 98%
      • Beta-blocker 90%
      • Mineralocorticoid antagonist 60%
      • Diuretic 80%
      • Nitrate 25%
      • Statin 83%
      • Prior to randomization: ASA (59%), other antiplatelet (7%), oral anticoagulant (8%)
    • ICD 18%

Interventions

  • Warfarin with target INR 2.75 (range 2.0 to 3.5)
    • Mean INR during trial 2.5 +/- 1, time in target range=63%
    • Spent 66% of follow-up on study treatment
  • ASA 325 mg PO once daily
    • Spent 68% of follow-up on study treatment

Results @ mean 3.5 y

  • No statistically significant difference in
    • Primary outcome (death, ischemic stroke, intracerebral hemorrhage): Warfarin 26.4% vs ASA 27.5%, hazard ratio (HR) 0.93 (95% confidence interval 0.79-1.10)
    • Secondary outcome (primary outcome, MI, HF hospitalization): 39.1% vs 37.4%, HR 1.07 (0.93-1.23)
    • Death: 23.5% vs 22.6%
  • Statistically significant
    • Reduction with warfarin in
      • Ischemic stroke: 2.5% vs 4.7%, HR 0.52 (0.33-0.82, NNT 46)
        • Per year: 0.7% vs 1.4% (NNT 143/year)
    • Increase with warfarin in
      • HF hospitalization: 20.9% vs 17.5%, HR 1.21 (1.00-1.47, NNH 30)
        • Per year: 6.8% vs 5.7% (NNH 91/year)
      • Major hemorrhage: 1.8%/year vs 0.9%/year, HR 2.05 (1.36-3.12, NNH 112/year)
      • Minor hemorrhage: 11.6%/year vs 7.3%/year, HR 1.56 (1.34-1.81, NNH 24/year)

Generalizability

  • The study population represents a typical HFrEF population of both ischemic & non-ischemic etiology with good to excellent use of background therapy for systolic dysfunction
    • 43% of patients had ischemic cardiomyopathy, therefore qualifying for ASA for secondary prevention
    • 57% had non-ischemic cardiomyopathy, for which antiplatelet therapy would not be indicated routinely. Since this trial has no placebo group, it's not able to answer whether these patients need any antithrombotic therapy at all in the absence of another indication.
  • The target INR in this trial (2.75, range 2.0-3.5) was higher than typically used in practice for AF (2.5, range 2.0-3.0), but similar to that used for mechanical mitral or high-risk aortic valves (3.0, range 2.5-3.5), however the mean INR during the trial was 2.5. Similarly, the ASA dose (325 mg/d) was higher than that typically used for AF/secondary prevention (75-100 mg/d). Thus, bleeding in both groups is likely to be higher than with regimens used in practice, particularly in the ASA group.

Issues with internal validity?

  • No: Centrally randomized, allocation-concealed, double-dummy double-blind trial with blinded outcome adjudication, and low loss-to-follow-up (1.5%) analyzed using intention-to-treat population.

Other studies

  • 3 other RCTs (WASH 2004, HELAS 2006, WATCH 2009) with a total of 1358 patients compared warfarin to ASA in patients with HF. Meta-analysis of all 4 trials demonstrates results consistent with WARCEF: Warfarin cut the risk of ischemic stroke by ~half, & ~doubled the risk of major & minor bleeds compared to ASA.