ACTIVE W
(2008)Objective
To compare the efficacy of oral anticoagulation (OAC) versus dual antiplatelet therapy (clopidogrel + aspirin) in preventing vascular events in patients with atrial fibrillation.
Study Summary
• Warfarin was superior to clopidogrel plus aspirin in preventing vascular events in atrial fibrillation without increased major bleeding.
Intervention
Warfarin (dose-adjusted to INR 2–3) vs. clopidogrel 75 mg daily + aspirin 75–100 mg daily. Median follow-up: 1.28 years.
Inclusion Criteria
Patients with atrial fibrillation (permanent, persistent, or ≥2 episodes of paroxysmal) and at least one risk factor for stroke, including age >75, hypertension, diabetes, heart failure, previous stroke/TIA, or vascular disease.
Study Design
Arms: Warfarin vs. Clopidogrel + Aspirin
Patients per Arm: Warfarin: 3371, Clopidogrel + Aspirin: 3335
Outcome
• Primary outcome (stroke, systemic embolism, MI, or vascular death): 3.93%/yr (warfarin) vs. 5.60%/yr (DAPT), RR 1.44, 95% CI 1.18–1.76, P=0.0003.• Stroke: 1.36%/yr vs. 2.44%/yr.• Major bleeding: 2.21%/yr (warfarin) vs. 2.42%/yr (DAPT), not significantly different.• In prior stroke/TIA subgroup: 6.22%/yr (DAPT) vs. 2.99%/yr (warfarin), RR 2.13, P=0.007.• Trial was stopped early for clear superiority of warfarin.
Bottom Line
Oral anticoagulation was significantly more effective than clopidogrel plus aspirin in preventing vascular events among patients with atrial fibrillation.
Major Points
- ACTIVE W was part of the 3-trial ACTIVE program (ACTIVE A, ACTIVE W, ACTIVE I) — the 'W' arm specifically compared warfarin to clopidogrel+aspirin in AF patients who WERE eligible for anticoagulation, definitively establishing OAC superiority.
- Stopped early by the data safety monitoring board after clear superiority of warfarin over DAPT: primary composite (stroke, SE, MI, vascular death) 3.93%/yr DAPT vs 3.37%/yr OAC (RR 1.44, 95% CI 1.18–1.76, p=0.0003). The 44% excess risk with DAPT was driven primarily by stroke.
- Stroke was dramatically higher with DAPT: 2.09%/yr vs 1.25%/yr with warfarin (RR 1.60, p<0.01) — a 60% relative increase in stroke risk. This was predominantly ischemic stroke, proving that antiplatelets cannot adequately prevent cardioembolic events.
- Major bleeding was NOT significantly different: 2.42%/yr DAPT vs 2.21%/yr OAC (RR 1.10, p=0.77) — refuting the common clinical assumption that DAPT is 'safer' than warfarin. The bleeding trade-off did not compensate for the efficacy loss.
- 6,706 patients across 522 centers in 33 countries. Open-label design with blinded outcome adjudication (PROBE design). Mean follow-up only 1.28 years due to early termination.
- Companion trial ACTIVE A (published 2009) tested clopidogrel+aspirin vs aspirin alone in AF patients INELIGIBLE for warfarin — showed modest benefit of DAPT over aspirin alone (RR 0.89) but with more bleeding, filling the therapeutic niche for warfarin-ineligible patients.
- ACTIVE W results were pivotal in establishing that clopidogrel+aspirin is NOT an acceptable substitute for anticoagulation in AF — a question that was common in clinical practice, particularly for patients who refused or were perceived to be 'too fragile' for warfarin.
- Subgroup analysis showed consistent OAC superiority: benefit was particularly pronounced in patients with prior stroke/TIA (secondary prevention) and those with CHADS2 ≥2, but even low-risk patients showed OAC advantage.
- Historical context: ACTIVE W was published before any DOAC trial. It set the stage for the DOAC era by firmly establishing that antiplatelets are inferior to anticoagulation in AF, making the subsequent DOAC vs warfarin comparison the critical next question.
- Warfarin TTR in ACTIVE W was 63.5% — better than ROCKET AF (55%) but similar to RE-LY (64%), suggesting reasonably well-managed anticoagulation in the control arm.
Study Design
- Study Type
- Randomized, open-label, non-inferiority trial
- Randomization
- Yes
- Blinding
- Open-label with blinded outcome adjudication
- Sample Size
- 6706
- Follow-up
- Mean 1.28 years
- Centers
- 522
- Countries
- Canada, Germany, Italy, Poland, Russia, UK, USA, Others
Primary Outcome
Definition: Composite of stroke, systemic embolism, MI, or vascular death
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| 3.37% per year | 3.93% per year | 1.44 (1.18–1.76) | 0.001 |
Limitations & Criticisms
- Open-label design — patients and physicians knew the assignment, potentially influencing event reporting and clinical management. Though outcomes were adjudicated blindly, awareness of treatment may have altered background therapy and threshold for hospitalization.
- Very short follow-up (mean 1.28 years) due to early termination — long-term bleeding consequences of chronic DAPT vs OAC remain uncertain, though the efficacy gap clearly favored OAC even in this short period.
- 78% of patients had prior VKA experience — making the warfarin arm perform better than it might in a VKA-naive population. This selection bias may underestimate the real-world challenges of warfarin management.
- DAPT arm used clopidogrel (not more potent P2Y12 inhibitors like ticagrelor or prasugrel) — though there is no mechanistic reason to expect P2Y12 inhibition to prevent cardioembolic events regardless of potency.
- Did not include any DOAC comparator — predates the DOAC era. Cannot inform the DOAC vs DAPT comparison directly, though subsequent trials (ACTIVE A + AVERROES) addressed the aspirin-only niche.
- Composite endpoint included MI and vascular death in addition to stroke/SE — the composite was driven primarily by stroke, but MI results were less clear (no significant difference).
- Relatively moderate-risk population (mean CHADS2 2.0) — OAC superiority may be even more pronounced at higher CHADS2 scores where stroke rates are highest.
- Industry-sponsored (Sanofi-Aventis, manufacturer of clopidogrel/Plavix, and BMS) — paradoxically, the trial disproved the hypothesis that the sponsor's own drug (clopidogrel) could substitute for warfarin.
- No analysis of hemorrhagic stroke subtypes — OAC's benefit on ischemic stroke prevention was clear, but the ICH rate comparison between groups was not robustly reported.
Citation
Connolly SJ et al. Lancet. 2006;367(9526):1903–1912.