Real world outcomes of minimally invasive epicardial surgical left atrial appendage exclusion in atrial fibrillation patients with high risk of stroke and bleeding

Abstract

Abstract Funding Acknowledgements Type of funding sources: Private company. Main funding source(s): AtriCure, Inc Introduction The left atrial appendage (LAA) is the site of thrombus formation in patients with atrial fibrillation (AF) and first line therapy for reducing the risk of stroke is oral anticoagulation (OACs). A subset of patients with AF carry a very high risk of stroke and bleeding and are ineligible for OAC and percutaneous LAA occlusion; these patients may be suitable for minimally invasive thoracoscopic epicardial surgical LAA exclusion (LAAE) which can be used with a wide variety of LAA morphologies with a ~98% rate of complete exclusion. While use of LAAE is well studied as a concomitant procedure, there are relatively few data on its use as a stand-alone procedure. Purpose To compare real-world outcomes of LAAE versus no-LAAE in AF patients at high risk of stroke and bleeding who are not on OACs. Methods We performed an analysis of Medicare beneficiaries with AF who were not on OACs and received LAAE (ICD10-PCS code 02L74CK, Oct 2015 – Dec 31, 2017) in the absence of a concomitant surgical or catheter AF ablation procedure. Patients were propensity-matched (1:4) to otherwise similar patients who did not receive LAAE (control), using age, gender, race, region, AF type, CHA2DS2-VASc, HAS-BLED, and co-morbidities. The primary endpoint was hospital readmission for a thromboembolic event (ischemic stroke, systemic embolism, or transient ischemic attack). Kaplan Meier event rates were calculated and 95% CI were generated via bootstrapping. Results A total of 243 patients who underwent LAAE and 1.7M control patients met study criteria. In the matched cohort, the mean age was 75 years, 61% were men, 77% had non-paroxysmal AF, and the mean CHADS2-VASC and HAS-BLED scores were 5.5 and 4.2, respectively. The 1-year rate of thromboembolism in the LAAE and control arms were 7.3% (95% CI 4.3% -10.9%) and 13.6 (95% CI 11.0% -16.4%) (Figure), respectively, with an absolute and relative risk reduction of 6.2% (95% CI 2.0-10.3, p-value=0.001, NNT 17) and 0.54 (95% CI 0.29-0.68, p=0.001). Conclusion Among high-risk AF patients who were not on OAC, LAAE was associated with a clinically meaningful and significantly lower risk of TE. Additional studies are needed to confirm these finding.