Outcomes in patients with atrial fibrillation randomized to receive LAA closure or oral anticoagulation: Primary results of the CHAMPION-AF clinical trial
Reported from ACC.26
During the first Late-Breaking Trial session of the 2026 ACC Congress held in New Orleans, Saibal Kar, principal investigator of the CHAMPION-AF trial, presented the 3-year results of this groundbreaking study, which were simultaneously published in the NEJM (DOI: 10.1056/NEJMoa2517213).
Left atrial appendage (LAA) occlusion has so far been considered an alternative embolic protection strategy for patients with an indication for oral anticoagulation who are at high bleeding risk. This trial compares a device-based strategy with medical therapy for stroke prevention in patients with non-valvular atrial fibrillation (AF) without contraindications to chronic anticoagulation.
The underlying hypothesis is that LAA occlusion may provide embolic protection comparable to direct oral anticoagulant (DOAC) therapy while significantly reducing bleeding risk by avoiding long-term anticoagulation.
What was the design of the trial?
- Hypothesis
The CHAMPION-AF study was designed to evaluate whether the Boston Scientific Watchman FLX device could serve as a reasonable first-line alternative to DOACs in patients with non-valvular AF who are eligible for oral anticoagulation.
- Study population
CHAMPION-AF randomised 3,000 patients with documented non-valvular AF and a CHA₂DS₂-VASc score ≥2 (men) or ≥3 (women). Patients were enrolled across investigational sites in the United States, Europe, Japan, and Brazil between October 2020 and November 2022.
Key exclusion criteria included myocardial infarction, stroke, or transient ischemic attack within 30 days before enrollment, and an International Society on Thrombosis and Haemostasis (ISTH)-defined major bleeding event within 30 days prior to randomization and LVEF<30%.
- Treatment arms
Patients were randomized 1:1 to receive treatment with either Watchman FLX (device arm) or DOAC (control arm).
In the device arm, post-implant antithrombotic regimens included DOAC plus aspirin, DOAC alone, or dual antiplatelet therapy (clopidogrel and aspirin). Apixaban and rivaroxaban were the recommended DOACs at standard AF dosing. Anticoagulation was discontinued at the 3-month follow-up if technical success (defined as a leak ≤5 mm) was achieved, with continuation of single antiplatelet therapy (aspirin or a P2Y12 inhibitor, if tolerated)
- Endpoints
Three co-primary endpoints were chosen:
#1 primary efficacy endpoint: composite of stroke (ischemic /hemorrhagic), cardiovascular death, and systemic embolism at 3 years, powered for non-inferiority.
#2 primary safety endpoint: non–procedure-related bleeding at 3 years based on the ISTH definitions for major bleeding and clinically relevant non-major bleeding, powered for superiority.
#3 primary endpoint: composite of ischemic stroke and systemic embolism at 5 years, powered for non-inferiority.
What is the main result of the trial?
Of the 3,000 enrolled patients, 1,499 were assigned to the device arm and 1,501 to the anticoagulation arm. The mean age was 71 years, and approximately 32% were women. The mean CHA₂DS₂-VASc score was 3.5. Paroxysmal AF was present in ~70% of patients, and 47.8% had undergone prior AF ablation.
In the device arm, procedural success was achieved in 98.8% of cases. Effective closure (residual leak ≤3 mm at 4 months) was observed in 98.6%.
At 3 years, follow-up was completed by 1,390 patients in the device arm and 1,306 in the control arm. Crossover from the control to the device arm occurred in 206 patients.
The primary efficacy endpoint occurred in 81 patients (5.7%) in the device group and 65 patients (4.8%) in the control group (difference 0.9%; 95% CI −0.8 to 2.6; P<0.001 for non-inferiority).
The primary safety endpoint occurred in 154 patients (10.9%) in the device group and 260 patients (19.0%) in the control group (difference −8.1 percentage points; 95% CI −10.8 to −5.5; P<0.001 for superiority).
Notably, similar rates were observed for the secondary endpoint of ISTH major bleeding (procedure- and non–procedure-related): 5.9% in the device arm vs 6.4% in the anticoagulation arm (hazard ratio 0.92; 95% CI 0.68–1.24).
Critical appraisal
These results raise the important question of whether LAA occlusion should be considered a first-line alternative to DOAC therapy in patients with non-valvular AF.
Although the trial demonstrated non-inferiority for efficacy and superiority for safety, these findings must be interpreted in the context of existing evidence and limitations of the current study. The attrition rate and crossover rate were acceptable, and procedural success and effective LAA closure exceeded 98%, supporting the feasibility of an interventional approach to embolic risk management.
However, this study should be regarded primarily as a device trial, as it exclusively evaluated the Watchman FLX system. Therefore, the generalizability of the results to other LAA occlusion devices remains uncertain.
The recently reported CLOSURE-AF trial (U. Landmesser et al., DOI: 10.1056/NEJMoa2513310), which evaluated multiple commercially available devices in high-risk patients, failed to demonstrate non-inferiority compared with contemporary medical therapy, reinforcing the need for caution.
Additionally, CHAMPION-AF excluded patients with an ejection fraction <30%, limiting applicability to a substantial subset of real-world patients.
Finally, high procedural costs and the persistent need for long-term single antiplatelet therapy may preclude demonstration of cost-effectiveness at 3 years in patients with a mean age of 71 years and an expected life span of 8–12 years.
As observed in PFO occlusion studies, both technological advancements and cumulative trial evidence were necessary to identify ideal tools and define patient population most likely to benefit from an interventional approach. A similar trajectory appears to be unfolding for LAA occlusion.
In this context of conflicting evidence regarding the role of LAA occlusion, while CHAMPION-AF trial prove superiority for safety and non-inferiority for efficacy of LAA occlusion with the Watchman FLX system as compared to DOAC, further data and longer follow-up are needed before broadly adopting this strategy as a first-line alternative to anticoagulation.
No comments yet!