B-PO03-155 OUTCOMES OF LEFT ATRIAL APPENDAGE OCCLUSION WITH WATCHMAN IN CONTEMPORARY CLINICAL PRACTICE COMPARED WITH CLINICAL TRIALS: A POOLED ANALYSIS OF PROTECT-AF, PREVAIL, AND THE NCDR LAAO REGISTRY

Abstract

The PROTECT-AF and PREVAIL trials established Watchman left atrial appendage occlusion (LAAO) as an alternative to warfarin for stroke protection in atrial fibrillation. In clinical practice, LAAO is often performed in patients who cannot tolerate long-term anticoagulation and have more cardiac and non-cardiac comorbidities. To compare outcomes of LAAO and control patients from PROTECT-AF and PREVAIL with LAAO patients from the NCDR LAAO Registry. We performed propensity matched analyses, matching up to 5 registry patients with each trial patient. Follow-up was censored at 425 days. The Kaplan Meier method was used to estimate event rates. Cox proportional hazards models, using registry patients as the reference group, were used to compare risk for ischemic stroke, hemorrhagic stroke, systemic embolism, and mortality. Of 38,285 patients enrolled in the LAAO Registry between 1/2016 and 12/2018; 4,874 were matched to 1,059 trial patients (701 LAAO, 362 warfarin). In adjusted analyses, the risk of ischemic stroke in trial patients randomized to LAAO was significantly higher than registry patients (2.7% vs. 0.9%; HR 2.69, p=0.0004); trial patients randomized to warfarin had a similar rate of ischemic stroke compared to registry patients (1.1%; HR 0.91, p=0.85). Risk of hemorrhagic stroke in trial warfarin patients were significantly higher compared with registry patients (1.4% vs. 0.3% HR 3.58, p=0.0156); trial LAAO patients had a similar risk compared with the registry patients (0.3%, HR 0.74, p=0.69). Systemic embolism was rare in all groups (registry patients 0.4%, trial warfarin patients 0.0%, trial LAAO patients 0.3%, p>0.05). Mortality was lower in trial LAAO patients than registry patients (3.0% vs. 4.7%; HR 0.57, p=0.016) but there were no differences compared to trial warfarin patients (4.1%; HR 0.66, p=0.12). In clinical practice, patients undergoing LAAO experience a lower risk of ischemic stroke, a similar risk of hemorrhagic stroke, and a higher risk of death in the 425 days after implant compared with LAAO patients in randomized trials. These results highlight the value of a national registry in understanding the adoption of technology and assessing procedural outcomes in contemporary practice.