The Left Atrial Appendage in Humans: Structure, Physiology, and Pathogenesis

Nabeela Karim; Siew Yen Ho; Edward Nicol; Wei Li; Filip Zemrak; Vias Markides; Vivek Reddy; Tom Wong


Europace. 2020;22(1):5-18. 

In This Article

Left Atrial Appendage Occlusion

As the LAA is responsible for much of the intra-atrial thrombus formed in NVAF, LAA occlusion could be a way of reducing this risk.

Surgical Left Atrial Appendage Occlusion

Techniques for surgical occlusion include stapling, ligation, amputation, and the LAA Clip System using the Atriclip. Success rates are often inconsistent and range from 17% to 95%,[75,76] additionally, there is no randomized control trial data that shows surgical closure confers a reduction in stroke.[77,78] The LAA occlusion study (LAAOS),[77] a single-centre study of 77 patients, was the first randomized control trial conducted in this field. Left atrial appendage occlusion was achieved in 45% of cases using sutures vs. 72% using staples. Thromboembolic events occurred in 2.6% of patients over a mean follow-up of 13 months. A second randomized control trial: LAAOS II[78] investigated 51 patients undergoing coronary artery bypass grafting and showed no significant difference in thromboembolic events and primary endpoints after 1 year. It should be noted, however, that the trial was designed for bigger numbers and not designed to look at implications of LAA occlusion on stroke outcomes.[78] It is possible that the small study numbers in the individual surgical studies may have impacted on the inconsistent outcomes, and lack of clarity with regards to the impact of surgical closure on stroke events as a meta-analysis of seven studies including two randomized control trials[79] with a total of 3653 patients, demonstrated that stroke and all-cause mortality was significantly reduced in surgical LAA occlusion at 30 day follow-up and latest follow-up. The Atriclip (Table 1) has also undergone evaluation with one study showing >95% success of exclusion[76] and another showing 100% effectiveness.[80]

Percutaneous Occlusion

Percutaneous occlusion involves trans-septal access into the LA to deliver the LAA occlusion device. Five devices are currently marketed in Europe. Amongst the four typical LAA morphologies, there are no differences in success or complication rates, with only atypical morphologies demonstrating lower implantation success rates.[81]

The Watchman device [Conformité Européenne (CE) mark 2005] (Table 1), is the only percutaneous LAA occlusion device to undergo randomized control trial studies. PROTECT-AF (Watchman LAA System for Embolic Protection in Patients with AF)[82] was a multicentre prospective randomized control trial of 707 NVAF patients with a mean CHADS2 score of 2.2. It demonstrated non-inferiority to warfarin for stroke, systemic embolism, or death.[82] After a mean follow-up time of 3.8 years, it displayed both superiority compared with warfarin.[83] The PREVAIL (Evaluation of the Watchman LAA Closure Device in Patients With Atrial Fibrillation vs. Long Term Warfarin Therapy) trial,[84] a prospective randomized control trial, investigated high-risk NVAF patients (mean CHADS2 score 2.6). It demonstrated non-inferiority to warfarin for ischaemic stroke prevention >7 days post-procedure. However, unlike PROTECT-AF, non-inferiority was not achieved for overall efficacy, possibly because the warfarin arm in PREVAIL, had an unusual low stroke rate of 0.73%.[85] Procedural safety was noted to have significantly improved from PROTECT-AF. PREVAIL and PROTECT-AF have been further analysed with 5-year outcomes individually and as a meta-analysis, showing comparability to warfarin for stroke prevention, with additional reductions in major bleeding, particularly haemorrhagic stroke and mortality.[85] ASAP (ASA Plavix Feasibility Study with Watchman LAA Closure Technology) was a prospective non-randomized trial that recruited NVAF patients with a contraindication to warfarin as is indicated on its labelling.[86] Unlike PROTECT-AF, no oral anticoagulation (OAC) was given post-implant, demonstrating the safety of LAA occlusion without warfarin transition.[86] The EWOLUTION prospective registry[87] provided real-world data on the Watchman, consisting of 1021 patients with a mean CHADS2 score of 2.8.[87] The majority of the population (73%), were deemed unsuitable for OAC by their physician, thus were not taking any.[88] A procedural success of 98.5%, was achieved, with an annual stroke rate of 1.1% resulting in an 84% reduction in ischaemic stroke at 1 year, compared with no therapy.[88] Real-world data[87,88] has demonstrated stroke and major bleeding rates to be lower than that reported by randomized control trials, with acute procedural success and complications rates also lower, despite inexperienced operators.[89] This may be due to the majority of patients form the real wold registry being ineligible for OAC,[88,90] whereas both randomized control trials included patients that were eligible for OAC.[82,84] Cost analyses shows LAA occlusion achieves parity with current medical therapies in a relatively short period of time[90] and becomes economically favourable over longer periods.[91,92]

The Amplatzer cardiac plug (ACP) (CE mark 2008) (Table 1) has been analysed in a retrospective multicentre registry of 1047 patients, making it the first large 'real world' report on LAA occlusion in patients unsuitable for OAC due to major bleeding, high risk for bleeding and coronary stenting mandating triple therapy. The patients were at high risk for stroke, with a mean CHA2DS2-VASc of 4. The procedural success rate was 97%, with a periprocedural major adverse event rate of 5% and annual stroke rate of 2.3%.[93] A prospective, single-centre study of 59 patients compared the ACP with its second-generation counterpart, the Amplatzer Amulet (CE mark 2013), demonstrating similar procedural success and short-term clinical outcomes but, with significant reduction in residual leaks for the Amplatzer Amulet.[94] The Amplatzer Amulet (Table 1) has subsequently been studied in a multicentre prospective real-world registry of 1088 NVAF, of which 82.2% had a contraindication to OAC.[95] Successful implantation was achieved in 99% with low periprocedural complications (3.2%). Transoesophageal echo follow-up (mean of 67 days) demonstrated 98.2% of patients had adequate closure and low rates of device-related thrombus (1.5%). The majority (54%) of patients post-implant remained on dual antiplatelets and the rest on single antiplatelet (23%) or OAC (18.9%).[95]

In the future, data on comparisons between the Amplatzer Amulet and Watchman device will become available with the Amplatzer Amulet LAA occluder randomized control trial currently underway.[96] Data comparing LAA occlusion vs. novel OAC will also be available, as several randomized control trials comparing the two are underway (PRAGUE 17,[97] CLOSURE-AF,[98] OCCLUSION-AF,[99] and STROKECLOSE[100]). However, the relatively modest size of these randomized control trials raises doubts that they will be adequately powered to satisfactorily reach relevant and definitive clinical conclusions. In addition, data from the Assessment of the Watchman Device in Patients Unsuitable for OAC (ASAP-TOO) randomized control trial is expected soon.[101]

The Wavecrest LAA occlusion device (CE mark 2008) (Table 1) has undergone a study in 73 patients with a mean CHADS2 score of 2.5, prior stroke in 49%, plus contraindication to warfarin. Left atrial appendage occlusion was achieved in 96%, with 2.7% developing periprocedural complications.[102] Randomized control trial data will be available in the future comparing it to the Watchman (WAVECREST-2).[103]

In addition to percutaneous endocardial devices is suture ligation with the Lariat device (CE mark 2009) (Table 1). It employs a pre-tied suture delivered over the LAA using an epicardial approach.[41] This results not only in occlusion, but essentially exclusion of the LAA due to tissue necrosis.[41] In a retrospective multicentre study of 154 patients with a median CHADS2 score of 3, procedural success was achieved in 87% of patients and major complications occurred in 9.7%, which consisted of death, stroke, myocardial infarction, major bleeding, or cardiac surgery.[104] In addition, 20% had a residual leak at TOE follow-up.[104] Another study reported 2.3% of patients needing urgent cardiac surgery.[105] Due to this, concerns have been raised over its procedural safety.[105] The LARIAT device is currently being studied in a large randomized control trial (aMAZE),[106] for its anti-arrhythmic potential in persistent AF (persAF) patients undergoing catheter ablation.