P999Major adverse events with percutaneous left atrial appendage closure in patients with atrial fibrillation in real life setting

Abstract

Abstract Transcatheter left atrial appendage (LAA) closure is an alternative strategy for stroke prevention in atrial fibrillation (AF) patients with an inacceptable risk of bleeding with oral anticoagulation (OAC). A better characterization of major adverse clinical events after LAA closure in daily practice is still needed. Methods We analysed data from all AF patients treated with Watchman or Amplatzer LAA closure according to European guidelines in 8 French cardiology departments. Antithrombotic management was decided for each patient on an individual basis. A Cox regression model was used for multivariable analysis of major adverse events. Yearly rate of ischemic stroke during follow-up was calculated and compared to that expected for a same risk score population. Yearly rate of bleeding was extrapolated from that reported with the HASBLED score. Results A total of 469 consecutive AF patients (299 males, 74.9±8.9 years old, mean CHA2DS2-VASc score 4.5±1.4, HASBLED score 3.7±1.0) received LAA closure from March 2012 to January 2017. There were 272 Watchman devices (58%) and 197 ACP devices (42%) implanted. At discharge, 36% received a single anti platelet therapy (APT), 23% received dual APT, 29% received OAC and no APT, 5% received OAC plus APT and 8% received no antithrombotic therapy. Mean follow up was 11.4 months (median 7, interquartile 3–22 months) during which 70 major adverse events (19 ischemic strokes, 18 major haemorrhages and 33 deaths) were recorded in 69 patients. The annual rate of ischemic stroke was 3.96%, which translates into a 13% relative risk reduction (95% CI −59 to 52%) as compared with the calculated stroke rate for similar CHA2DS2-VASc score after adjustment for exposure to APT and OAC. The annual rate of major bleeding in the study was 3.75%, which corresponds to a 48% relative risk reduction (95% CI 9 to 70%) as compared with the rate that would have been expected based on a comparable HAS-BLED score. Yearly rate of mortality was 7.4% (2.5 to 3 fold higher than in previous randomized trials) and the rate of non-cardiovascular death was 82%. None of the baseline characteristics was predictive of major adverse events, neither in univariate nor in multivariable analysis, which highlights the difficulty in identifying a risk of unfavourable outcome with simple tools. Conclusions AF patients treated with LAA closure may have a lower risk of stroke and bleeding events compared to their theoretical risk. However, our findings indicate that a high rate of major adverse events is observed in these patients during follow-up. This questions the suggested cost-effectiveness of the procedure (with models based on previous trials) for a real-life perspective. A better identification of patients with a relevant benefit of LAA closure is needed among those with long-term anticoagulation contraindication, both for an optimal management of each patient on an individual basis and for a global perspective with limited healthcare resources.