In anticoagulation-eligible patients with nonvalvular atrial fibrillation, left atrial appendage closure with Watchman FLX was noninferior to NOAC therapy for cardiovascular death, stroke, or systemic embolism at 3 years and reduced non-procedure-related bleeding, but ischemic stroke was numerically higher with device therapy.
Study design
- Prospective, international, randomized trial
- Open-label design with blinded outcome adjudication
- 141 sites
- 1:1 randomization
- N = 3,000
- Follow-up: 3 years for primary endpoints
Population
- Nonvalvular atrial fibrillation
- Suitable candidates for long-term oral anticoagulation
- Mean age 71.7 years
- 31.9% women
- Mean CHA₂DS₂-VASc 3.5
- Mean HAS-BLED 1.3
- Excluded some higher-risk groups, including advanced HF and LVEF <30%
Interventions
- LAA closure with Watchman FLX: 1,499
- NOAC therapy: 1,501
- NOAC selected at clinician discretion
- Device arm used protocol-directed post-implant antithrombotic therapy
Primary outcome
- Primary efficacy: cardiovascular death, stroke, or systemic embolism
- 5.7% with LAA closure vs 4.8% with NOAC
- Difference 0.9 percentage points; 95% CI -0.8 to 2.6
- P < 0.001 for noninferiority
- Primary safety: non-procedure-related major or clinically relevant non-major bleeding
- 10.9% with LAA closure vs 19.0% with NOAC
- HR 0.55; 95% CI 0.45-0.67
- P < 0.001 for superiority
LAA closure vs NOAC
Primary efficacy endpoint at 3 years
5.7% vs 4.8% | Noninferior
Primary efficacy: cardiovascular death, stroke, or systemic embolism at 3 years.
LAA closure vs NOAC
Primary safety endpoint at 3 years
10.9% vs 19.0% | HR 0.55
Primary safety: non-procedure-related major or clinically relevant non-major bleeding at 3 years.
Secondary outcomes
- Secondary net clinical benefit favored LAA closure: 15.1% vs 21.8%
- Procedural + non-procedural major or clinically relevant non-major bleeding: 12.8% vs 19.0%
- Major bleeding: 5.9% vs 6.4%
- No difference in cardiovascular death, systemic embolism, or hemorrhagic stroke
- Ischemic stroke numerically higher with LAA closure: 3.2% vs 2.0%
- Hemorrhagic stroke: 0.4% vs 0.4%
Safety
- Non-procedure-related bleeding was lower with LAA closure
- Periprocedural complication rates were low
- Procedural success reported at 98.8%
- Periprocedural stroke: 0.1%
- Pericardial effusion: 0.6%
- Procedure-related major bleeding: 0.4%
- No procedure-related mortality reported
- Device-related thrombus detected at 4 months in 4.8%
Interpretation
- CHAMPION-AF supports LAA closure as a potential alternative to NOACs in selected anticoagulation-eligible patients with nonvalvular AF
- The main advantage was less long-term bleeding
- The efficacy endpoint met noninferiority, but it was not clearly better than NOAC therapy
- The ischemic stroke signal with LAA closure is the major caveat
- Best framed as a shared decision-making option, not a blanket replacement for NOACs
- Longer-term 5-year follow-up will be important
Doshi, S. K., Kar, S., Nair, D. G., Waggoner, T., Agarwal, H., Moussavian, M., … CHAMPION-AF Investigators. (2026). Left atrial appendage closure or anticoagulation for atrial fibrillation. The New England Journal of Medicine. https://doi.org/10.1056/NEJMoa2517213