Efficacy and safety of left atrial appendage closure compared with oral anticoagulation in atrial fibrillation: a meta-analysis of randomized controlled trials and propensity-matched studies

BACKGROUNDS: Two recent randomized controlled trials (RCTs), the PROTECT-AF and the PREVAIL, showed that in atrial fibrillation (AF) patients, left atrial appendage closure (LAAC) is comparable to oral anticoagulants (OAC) in the prevention of stroke and could also possibly reduce mortality. Nevertheless, this net clinical benefit was not confirmed in the most recent RCT comparing LAAC vs. OAC, the PRAGUE-17 trial. AIM: aim of the present study was to evaluate the efficacy and safety of LAAC compared with OAC among available high-quality studies. METHODS: A systematic search of electronic databases (Medline, Scopus, Embase and the Cochrane Library) was performed to identify eligible RCTs and observational studies with propensity score matching (PSM) analysis. PRISMA guidelines were used for abstracting data and assessing data quality and validity. Outcomes of interest were the occurrence of cardiovascular death (CVD), all-cause death, all-type stroke, and major bleedings. RESULTS: A total of 3 RCTs and 7 PMS studies involving 25,700 patients were identified. 12,961 patients received LAAC while 12,739 received OAC therapy. After a median follow-up of 2.6 years (IQR 2–4.4), patients who received LAAC had lower risk of CVD (RR = 0.62; 95%CI, 0.51–0.74, I(2 )= 0%), all-cause death (RR = 0.67; 95% CI, 0.57–0.78, I(2) 68%) and major bleedings (RR = 0.68; 95%CI, 0.48–0.95 I(2 )= 87%) compared with patients on OAC. No difference was found between the two groups regarding strokes incidence (RR = 0.94; 95% CI, 0.77–1.15, I(2 )= 0%). CONCLUSIONS: According to this meta-analysis, LAAC has comparable efficacy in the prevention of stroke compared with OAC and a reduced risk of major bleedings, all-cause death and CVD that may be even larger with longer follow-up. SYSTEMATIC REVIEW REGISTRATION: https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=269768, identifier CRD42021269768.