Atrial fibrillation: Left atrial appendage occlusion

INTRODUCTION — The roles of percutaneous and surgical left atrial appendage occlusion (LAAO) in patients with atrial fibrillation (AF) are discussed here.

AF is a common cause of embolic stroke and other thromboembolic complications. The LAA is a primary source for thromboembolism in AF. Whereas anticoagulation is recommended for most patients with AF, some patients have contraindications to long-term anticoagulation. Anticoagulation in patients with AF is discussed separately. (See "Atrial fibrillation in adults: Selection of candidates for anticoagulation" and "Atrial fibrillation in adults: Use of oral anticoagulants".)

RATIONALE AND LIMITATIONS — Among patients with AF, thrombus in the LAA is the primary source for thromboemboli. A review of studies examining the site of left atrial thrombus (by transesophageal echocardiography [TEE], cardiac surgery, or autopsy) found that in patients with AF without rheumatic heart disease, 90 percent of left atrial thrombi were located in the LAA [1]. The importance of the LAA as a source for thromboemboli among patients with AF provides the rationale for ligation, amputation, or occlusion of the LAA, particularly for patients who have an indication for anticoagulation but cannot take long-term oral anticoagulation. However, the following caveats should be kept in mind:

●Ten percent of left atrial thrombi are not located in the LAA [1], so occluding the LAA may not completely eliminate the risk of embolism from the LA.

●Thrombogenesis in patients with AF may not be limited to the LA. The pathophysiology of thrombogenesis in AF, including left atrial structural and functional abnormalities and hypercoagulable state, is discussed further separately. (See "Mechanisms of thrombogenesis in atrial fibrillation" and "Hemodynamic consequences of atrial fibrillation and cardioversion to sinus rhythm", section on 'Atrial stunning'.)

●If the LAAO has an incomplete closure, this can result in a residual communication between the LAA and LA, also referred to as a residual jet. An incomplete jet is one that is >5 mm; these are believed to predispose to thrombus formation. The importance of smaller jets (ie, ≤5 mm) is uncertain, but one study suggested a low risk of thromboembolic events [2]. In this retrospective review of 6- and 12-month follow-up echocardiograms in over 400 patients who received the WATCHMAN device, there was no significant increase in the rate of thromboembolism in the 32 percent of patients with LAA residual jet width >5 mm compared with those with no residual jet or jet width ≤5 mm. It is likely that most residual holes are small and not associated with embolization of large clots.

INDICATIONS — The indications for surgical and nonsurgical closure are similar.

Patients not undergoing cardiac surgery — The following approaches apply to patients with AF with an indication for anticoagulation (based upon their estimated risk of stroke and other systemic thromboembolism (table 1)) who have a contraindication to long-term anticoagulation. These approaches do not apply to patients with AF who have other indications for anticoagulation (eg, an implanted mechanical valve). (See "Atrial fibrillation in adults: Selection of candidates for anticoagulation".)

Patients with one or more of the following conditions may have a contraindication to long-term anticoagulation and may therefore have an indication for LAA occlusion:

●Thrombocytopenia or known coagulation defect associated with bleeding.

●Recurrent bleeding, including gastrointestinal, genitourinary, or respiratory sites.

●Prior severe bleeding, including intracranial hemorrhage.

●High risk of the patient falling (which may be evidenced by history of prior falls resulting in injury) despite safety measures.

●Strong indication for combined use of dual antiplatelet and anticoagulant therapy.

●Poor compliance with or intolerance of anticoagulant therapy.

For patients not undergoing cardiac surgery who have AF with an indication for anticoagulation (based upon risk of stroke and other systemic thromboembolism (table 1)) but cannot be on long-term oral anticoagulation, we suggest placement of a percutaneous LAAO device. (See "Atrial fibrillation in adults: Selection of candidates for anticoagulation", section on 'Approach to deciding whether to anticoagulate'.)

However, it should be noted that short-term antithrombotic therapy (anticoagulant or dual antiplatelet therapy) is generally recommended after placement of the device as discussed below. (See 'After percutaneous closure' below.)

Therefore, patients who cannot have even short-term blood-thinning medicines (anticoagulation therapy, P2Y12 inhibitors, and aspirin) are generally not candidates for LAAO. Most patients can tolerate at least one antiplatelet agent for a short duration.

The efficacy and safety of percutaneous LAAO as an adjunct to long-term oral anticoagulation in patients with AF has not been established. A rationale for this approach would be to attempt to reduce the risk of stroke and other systemic embolic events beyond the risk reduction provided by anticoagulation, but this approach would entail procedural risks that might offset potential benefits.

Both efficacy and safety of LAAO devices compared with long-term anticoagulation were demonstrated in a meta-analysis that included three randomized, open-label, controlled trials (PROTECT AF, PREVAIL, and PRAGUE-17) [3]. A total of 933 patients were randomly assigned to percutaneous LAA closure, and 583 were randomly assigned to oral anticoagulation (65 percent were on warfarin, and 35 percent on a direct oral anticoagulant [DOAC]) and followed for an average of 39 months. The mean age was 73.3 years, and the mean CHA₂DS₂-VASc score was 4.1. The study made the following observations. The implanted device was WATCHMAN, Amulet, and WATCHMAN FLX in 86, 13.5, and 0.5 percent, respectively.

●The mean age was 73.3 years, and the mean CHA₂DS₂-VASc score was 4.1.

●Successful implantation occurred in 93.3 percent of patients

●There was no significant difference in the all-stroke or systemic embolism rate between the groups (6.1 versus 5.8 percent; risk ratio [RR] 0.98, 95% CI 0.65-1.48).

●LAA closure was associated with a lower risk of hemorrhagic stroke (0.5 versus 2.4 percent; RR 0.22, 95% CI 0.08-0.58), cardiovascular death (5.4 versus 8.2 percent; RR 0.65, 95% CI 0.44-0.95), and all-cause death (14.3 versus 17.8 percent; RR 0.78, 95% CI 0.62-0.99).

●Major bleeding was similar in the two groups (RR 0.89, 95% CI 0.66-1.20), but non-procedure-related major bleeding was lower with LAA closure (RR 0.53, 95% CI 0.38-0.74).

●Procedure-related complications occurred in 6.8 percent.

The efficacy of specific percutaneous LAA occlusion devices is described below. (See 'Percutaneous devices' below.)

Patients undergoing surgery — Surgical ligation or amputation of the LAA can be performed without significant morbidity or mortality in patients with AF who are undergoing cardiac surgery for other indications. Surgical ligation or amputation of the LAA can be performed without significant morbidity or mortality in patients with AF who are undergoing cardiac surgery for other indications. Surgical LAAO is most commonly performed in patients undergoing mitral valve or Maze surgery. (See "Atrial fibrillation: Surgical ablation", section on 'Maze procedure' and 'Surgical LAA closure' below.)

For patients with AF who are undergoing surgery for another indication and who also have a CHA₂DS₂-VASc score of at least 2 (table 1), we consider LAAO.

●Patients with no contraindication to anticoagulation – In such patients, we recommend concomitant surgical LAAO as an adjunct to oral anticoagulation, as LAAO may help further reduce future risk of thromboembolic events. A randomized trial and observational studies support the efficacy of LAAO for selected patients with AF undergoing cardiac surgery:

•Randomized trial – The role of surgical LAAO as an adjunct to anticoagulation is supported by the results of the LAAOS III trial in individuals with AF (mean age 71) undergoing cardiac surgery for another indication with a CHA₂DS₂-VASc score of at least 2 (mean score of 4.2) [4]. In this multicenter trial, 2379 participants were randomly assigned to the occlusion group, and 2391 to the no-occlusion group, with a mean of 3.8 years of follow-up. A total of 92.1 percent of the participants received the assigned procedure. The majority of participants continued oral anticoagulation after the procedure (76.8 percent at three years). Stroke or systemic embolism occurred in 114 participants (4.8 percent) in the occlusion group and in 168 (7.0 percent) in the no-occlusion group (hazard ratio [HR] 0.67, 95% CI 0.53-0.85). Incidence rates of perioperative major bleeding, hospitalization for heart failure, and death were similar in the two groups.

Surgical LAAO is performed using a variety of techniques. In the LAAOS III trial, LAAO was performed by amputation and closure, stapler closure, double-layer linear closure from within the atrium (in patients undergoing minithoracotomy; this approach required transesophageal confirmation of occlusion), or closure with an approved surgical occlusion device [4]. Intraoperative TEE confirmation of LAA closure is recommended.

•Observational studies – Two large retrospective cohort studies suggested a benefit from surgical LAAO in patients with AF undergoing cardiac surgery [5,6], whereas one smaller study did not [7]:

In a study of 75,782 patients undergoing cardiac surgery, 8590 propensity-matched pairs (with and without surgical LAAO) were identified and studied; 74.9 percent had a history of AF, and 30.5 percent were taking an oral anticoagulant at baseline [6]. During follow-up (mean 2.1 years), LAAO was associated with a reduced risk of ischemic stroke or systemic embolism (1.14 versus 1.59 events per 100 patient-years; HR 0.73, 95% CI 0.56-0.96) and mortality (3.01 versus 4.30 events per 100 patient-years; HR 0.71, 95% CI 0.60-0.84). Most of the benefit was seen in patients with prior AF. LAAO was also associated with modestly higher rates of AF-related outpatient visits (11.96 versus 10.26 events per patient-year; rate ratio 1.17, 95% CI 1.10-1.24) and hospitalizations (0.36 versus 0.32 events per patient-year; rate ratio 1.13, 95% CI 1.05-1.21).

In contrast, a smaller study (with 461 propensity-matched pairs) found similar risks of stroke and mortality in patients with and without surgical LAAO [7].

●Patients with a contraindication to long-term anticoagulation – In such patients, we suggest concomitant surgical LAAO, as the procedure is generally safe and may provide some protection against future stroke. However, there are limited data to support surgical LAAO without long-term anticoagulation. Long-term anticoagulation is usually used after surgical LAAO. Patients with an absolute contraindication to anticoagulation must be able to take dual antiplatelet therapy for up to six weeks postoperatively. (See 'After surgical occlusion' below.)

In a retrospective cohort study of 10,524 Medicare recipients with AF undergoing cardiac surgery, surgical LAAO (n = 3892) compared with no occlusion (n = 6632) was significantly associated with lower risk of readmission for thromboembolism at three years (4.2 versus 6.2 percent; HR 0.67; 95% CI 0.56-0.81) [5]. Among patients discharged without anticoagulation, LAAO was associated with a lower risk of thromboembolism (4.2 versus 6.0 percent; HR 0.26; 95% CI 0.17-0.40), but not among patients discharged with anticoagulation (4.1 versus 6.3 percent; HR 0.88; 95% CI 0.56-1.39). In the cohort of patients discharged with oral anticoagulation, surgical LAAO was not associated with thromboembolism but was associated with a lower risk for hemorrhagic stroke, presumably related to eventual discontinuation of oral anticoagulation among the LAAO patients.

In patients without AF, there is no established role for LAAO.

PREPROCEDURE PLANNING — Among persons with AF, the anatomy of the LAA is heterogenous including varied shapes, neck and long axis diameters, and number of lobes [8]. Specific LAA measurements (eg, LAA depth and mouth diameter) can be useful for LAA closure planning; these can be obtained by transesophageal echocardiogram (TEE), coronary computed tomography angiography, or cardiac magnetic resonance imaging. No specific LAA preprocedural imaging is required for surgical LAAO.

If upon cardiac imaging an LA/LAA thrombus is seen, the LAAO procedure is contraindicated. If an LAA thrombus is visualized, the we initiate or intensify anticoagulation (as appropriate) and continue it for six to eight weeks. At that time, we repeat LAA imaging to reassess the thrombus and if resolved, we perform LAAO. (See "Atrial fibrillation in adults: Use of oral anticoagulants".)

Patients should be asked about sensitivity to Nitinol (an alloy made of nickel and titanium) or any other material that is part of the specific LAAO device; in the case of an allergy, the device would be contraindicated. For instance, both the Watchman and Amplatzer devices contain Nitinol. Nickel allergy is very common. (See "Nickel hypersensitivity and coronary artery stents".)

Ideally, patients receiving a percutaneous device are able to tolerate oral anticoagulation (vitamin K antagonist or direct oral anticoagulant [DOAC]) or dual antiplatelet therapy for at least six weeks after the device is implanted and cannot have LAA clot present at the time of implantation. Some experts would perform LAAO in patients who cannot tolerate anticoagulants or antiplatelets, whereas others require at least two to four weeks of aspirin postprocedure.

LAAO device implantation can be performed with transesophageal echocardiography (TEE) guidance under general anesthesia or intracardiac echocardiographic guidance under local anesthesia, but the practice varies [9,10].

PROCEDURE — In most centers, the procedure is performed using transesophageal echocardiographic (TEE) guidance under deep sedation or general endotracheal anesthesia. Some centers perform the procedure under sedation and only fluoroscopic guidance.

After vascular access is obtained in the femoral vein, a comprehensive TEE is performed (if it was not performed previously) to assess the LAA anatomy and specifically to rule out presence of any clot/thrombus, as this is a contraindication to proceed. After this, a transseptal puncture is performed. Obtaining the correct location of the puncture in the intraatrial septum is important and is guided by TEE. Once the septum is punctured, we administer heparin with a goal activated clotting time >250 seconds; this is done to avoid thrombosis of the percutaneous catheters and occlusion device. Angiography is performed in two orthogonal views for sizing. Device sizing and selection are based on the anatomy. Once the device is deployed, stability is assessed and then the device is released. Patients usually spend one night in the hospital and are discharged the following day on antiplatelets therapy.

PERCUTANEOUS DEVICES — Among the available percutaneous LAA closure devices, the WATCHMAN is the only device for which randomized trials have shown comparable efficacy and safety with long-term warfarin for the prevention of stroke and systemic embolization. The Amulet device was shown to be noninferior to the WATCHMAN device in the Amulet IDE study. Selection between the WATCHMAN FLX and AMULET device is largely based on anatomy and operator comfort/experience.

The WATCHMAN (largely replaced by the second-generation WATCHMAN FLX), the Amplatzer Cardiac Plug, and the Amplatzer Amulet are the leading implanted LAA endovascular devices worldwide and have CE (The European Union device approval). The WATCHMAN, WATCHMAN FLX, and Amulet devices are approved by the United States Food and Drug Administration (FDA) (table 2).

Endovascular devices — Percutaneous devices require a transseptal puncture from a femoral venous access to enable device implantation into the LAA from a purely endovascular approach (figure 1). These procedures are done under transesophageal echocardiography (TEE) or intracardiac echocardiography guidance. Efficacy and some device-specific safety outcomes are discussed here. Procedural complications are discussed in greater detail below. (See 'Device safety and complications' below.)

WATCHMAN device — The WATCHMAN device is the most commonly implanted percutaneous LAAO device and has the most robust data to support its use. It is a self-expandable cage deployed in the LAA, using a transseptal approach (figure 1). The device is covered by a layer of permeable polyethylene terephthalate (PTFE) membrane, which endothelializes within 45 days in dog models.

The WATCHMAN device comes in five sizes (table 3) and has 10 fixation anchors to aid device stability. The second-generation WATCHMAN FLX device also comes in five sizes (table 3), with a slightly broader range of dimensions, and has 12 fixation anchors.

●Efficacy randomized trials – Newer-generation WATCHMAN devices were shown to be both efficacious and safe in a nonrandomized study of 400 patients (PINNACLE FLX study) [11]. The first-generation WATCHMAN device was comprehensively evaluated in two randomized trials (PROTECT AF and PREVAIL) in patients with nonvalvular AF eligible for oral anticoagulation [12,13]. The five-year outcomes of the PREVAIL trial, combined with the five-year outcomes of the PROTECT AF trial, were reported in 2017 [14]. The results demonstrated that LAA closure with WATCHMAN device provided stroke prevention in nonvalvular AF that was comparable to warfarin, with additional reductions in major bleeding, particularly hemorrhagic stroke, and mortality.

•In the PROTECT AF noninferiority trial, 707 patients with AF and a CHADS₂ score ≥1 were randomly assigned in a 2:1 ratio to either the device or long-term warfarin (international normalized ratio 2 to 3) and followed for 18 months (table 4) [15]. Device implantation was successful in 91 percent of patients in whom it was attempted. Postimplantation, patients were continued on warfarin for 45 days, followed by switching warfarin to clopidogrel plus aspirin to six months, followed by aspirin alone indefinitely.

The risk of composite outcome (stroke, systemic embolism, or cardiovascular death) was similar in the intervention and control groups (3 versus 4.9 events per 100 patient-years, respectively; rate ratio 0.62; 95% Bayesian credible interval 0.35-1.25).

The primary safety endpoint (major bleeding, pericardial effusion, procedure-related stroke, or device embolization) occurred more often in the device group (7.4 versus 4.4 events per 100 patient-years, respectively; rate ratio 1.69; 95% Bayesian credible interval 1.01-3.19). Most of the events in the device group occurred early. Of these, approximately 50 percent were pericardial effusions requiring drainage. The device embolization rate was 0.6 percent.

In a subsequent study from PROTECT AF, at 12 months, patients treated with the closure device had improvement in health-related quality-of-life measures, while these measures declined among patients treated with warfarin [12,13].

•The PREVAIL study was mandated by the FDA to further evaluate the safety profile and confirm the efficacy of WATCHMAN for regulatory approval. This study randomly assigned 407 patients to WATCHMAN or warfarin. Participants had CHADS₂ score ≥2, or 1, if ≥1 of the following was present: female ≥75 years old, left ventricular ejection fraction 30 to 34.9 percent, age 65 to 74 with diabetes or coronary artery disease, or age ≥65 with documented congestive heart failure [13]. At 18-month follow-up, the first coprimary efficacy endpoint (composite of stroke, systemic embolism, and cardiovascular/unexplained death) with WATCHMAN versus warfarin did not achieve noninferiority criteria (0.64 versus 0.063; rate ratio 1.07; 95% CI 0.57-1.89). However, the second coprimary efficacy endpoint (stroke or systemic embolism >7 days postrandomization) did achieve noninferiority (0.025 versus 0.020; risk difference 0.0053; 95% CI -0.0190 to 0.0273). Early safety events occurred in 2.2 percent of the WATCHMAN arm, which was significantly lower than in PROTECT AF, satisfying the prespecified safety performance goal.

●Evidence in patients with absolute contraindications to oral anticoagulation – LAA closure with WATCHMAN has been evaluated in 150 patients with absolute contraindications to oral anticoagulation in nonrandomized studies. In the prospective multicenter, nonrandomized ASAP study, patients with a CHADS₂ score ≥1 were treated with WATCHMAN and six months of a thienopyridine (clopidogrel or ticlopidine), as well as life-long aspirin [16]. During a mean follow-up of 14.4 months, stroke or systemic embolism occurred at a rate of 2.3 percent per year. This rate was lower than predicted rates for CHADS₂-matched cohorts of individuals taking either aspirin (7.3 percent) or clopidogrel (5 percent). (See "Atrial fibrillation in adults: Selection of candidates for anticoagulation", section on 'Use'.)

●Real-world observational studies – In the EWOLUTION study of real-world experience with WATCHMAN, procedural success was achieved in 98.5 percent, and serious procedural complications within seven days occurred in only 2.8 percent [17,18]. At one-year follow-up, the ischemic stroke rate was 1.1 percent, which represented 84 percent reduction compared with the historical estimate based upon CHA₂DS₂-VASc score [18].

Amplatzer devices — Globally, the Amplatzer devices (Cardiac Plug and Amulet) are the second most commonly implanted endovascular LAA closure devices after WATCHMAN.

●Description and procedural success rates – The Amplatzer Cardiac Plug (figure 2) [19] device is constructed of Nitinol mesh, consisting of a proximal LA disk and a distal LAA lobe connected by a short waist. The lobe contains six pairs of stabilizing wires designed to increase stability within the appendage. The second-generation Amplatzer Cardiac Plug device (Amulet) received European device approval in 2013 and FDA approval in 2021, and incorporates significant advances, including wider lobe and more stabilizing wires (up to 10 pairs) for improved device stability, and larger lobe size to occlude larger appendages. A multicenter Amulet postmarketing registry report, including 1088 patients [20], showed procedural success was 99 percent. Procedural and in-hospital major adverse events occurred in 3.2 percent of patients.

●Amplatzer versus WATCHMAN – The Amplatzer device is shorter than the WATCHMAN device and may be more advantageous in individuals with short appendages. In a randomized trial comparing the two devices, the Amulet device was noninferior to the WATCHMAN in terms of primary safety and efficacy endpoints, but was associated with more procedure-related complications. The Amulet IDE trial randomly assigned 1878 patients with AF (mean CHA₂DS₂-VASc score of 4.6) to receive either the Amulet or WATCHMAN percutaneous LAAO device [21]. The rate of ischemic stroke or systemic embolism at 18 months was similar with the Amulet and WATCHMAN devices (2.8 percent for both). Major bleeding and all-cause death were similar in the two groups (10.6 versus 10 percent and 3.9 versus 5.1 percent, respectively). LAAO with residual jet ≤5 mm (ie, adequate LAA exclusion) was more frequent with the Amulet occluder than the WATCHMAN device (98.9 versus 96.8 percent, ie, Amulet was superior to WATCHMAN in reducing peridevice leak). However, the following rates of procedure-related complications were higher for Amulet versus WATCHMAN:

•Early pericardial effusion/tamponade (<2 days postprocedure; 1.3 versus 1.1 percent)

•Late pericardial effusion/tamponade (>2 days postprocedure; 1.1 versus 0.1 percent)

•Device embolization (0.7 versus 0.2 percent)

WaveCrest device — The WaveCrest is another endovascular device designed specifically for closure of the LAA. It is currently used in the European Union but is not approved for use in the United States. This device consists of a single lobe that occludes the LAA (figure 3). Unlike WATCHMAN, the WaveCrest is covered by a foam layer on the LAA side and PTFE on the side facing the LA. It has several anchors along the LAA side. This device is meant to be deployed quite proximally in the LAA, rather than deep within the structure. Thus, this device is another alternative to the WATCHMAN device if the LAA is too small to accommodate deeper devices.

LARIAT system — The LARIAT system is a nonsurgical (percutaneous) device (picture 1) approved by the FDA for soft tissue closure ("approximation") only [22-26], but not specifically for prevention of thromboembolism with LAAO. Its use for LAA closure is "off-label" in the United States, and it is rarely used. Unlike the endovascular devices, the LARIAT system requires access to both the endocardial and the epicardial space (the latter via a subxiphoid percutaneous approach). A magnetic guide is placed within the LAA to allow the epicardially placed lasso to tie off the LAA. Patients who have had prior cardiac surgery or unusual LAA anatomy are not candidates for this procedure (table 2).

The aMAZE trial randomly assigned 610 patients with persistent AF to LAA ligation with LARIAT plus pulmonary vein antral isolation (a type of catheter ablation of AF) versus pulmonary vein antral isolation alone and followed patients for efficacy at rhythm control (ie, freedom from antiarrhythmic medications) and safety [27]. The primary effectiveness endpoint (freedom from antiarrhythmic drug therapy at 12 months) was 64.3 percent in the LAA ligation/pulmonary vein antral isolation group versus 59.9 percent in the pulmonary vein antral isolation alone group (criterion for superiority not met). The primary safety endpoint occurred in 3.4 percent of the LAA ligation/pulmonary vein antral isolation group. Residual communication between the LA and LAA ≤1 mm at 12 months postprocedure occurred in the majority of patients in the LARIAT plus catheter ablation group (85 percent).

SURGICAL LAA CLOSURE — Types of surgical closure include stapler occlusion, amputation and suture, and epicardial device closure (table 2). These are usually done in patients who are undergoing cardiac surgery for another indication such as coronary artery bypass grafting or cardiac valve replacement surgery. A detailed discussion of the risks and benefits of various types of surgical closure is beyond the scope of this topic.

LAAO or excision may be incomplete, as has been detected by transesophageal echocardiography (TEE) studies. Patients with incomplete LAAO continue to be at risk for LAA thrombus and thromboembolic events. In three separate cases series, incomplete surgical LAA ligation was reported in 10 to 40 percent of patients undergoing the procedure [28-30]. Among 50 patients undergoing mitral valve surgery and LAAO, incomplete appendage closure occurred in 36 percent of patients, and stroke occurred in 22 percent of patients with incomplete appendage ligations [29].

A minimally invasive thoracotomy (versus median sternotomy) is a less invasive approach that provides transpericardial (epicardial) access for LAA closure with an endovascular device; examples of devices used with this approach are listed in a table (table 5) [31]

The efficacy and safety of stand-alone thoracoscopic LAA appendectomy have not been established, and it is infrequently performed, given the widespread availability of percutaneous LAAO. The feasibility of stand-alone thoracoscopic LAA appendectomy was addressed in a study of 30 patients with AF and prior thromboembolism who had contraindications to oral anticoagulation [32]. Thoracoscopic LA appendectomy was not associated with increased mortality or major complications. At a mean follow-up of 16 months, no patients had experienced a recurrence of thromboembolism.

POSTPROCEDURE MANAGEMENT

Endocarditis prophylaxis — We generally recommend bacterial endocarditis prophylaxis for up to six months after device placement. Antimicrobial regimens for prophylaxis and relevant procedures are discussed separately. (See "Prevention of endocarditis: Antibiotic prophylaxis and other measures".)

After percutaneous closure — We prescribe short-term anticoagulation and obtain follow-up transesophageal echocardiograms (TEEs) in patients who have had percutaneous LAAO.

●Primary anticoagulation strategies – For most patients, we prescribe an oral anticoagulant (warfarin with target international normalized ratio 2 to 3 or direct oral anticoagulant [DOAC]) plus aspirin (81 to 325 mg daily) for 45 days, followed by once-daily clopidogrel (75 mg) plus aspirin (81 to 325 mg) for six months; thereafter, we continue once-daily aspirin (81 to 325 mg) alone indefinitely.

This protocol is based on that used in the PROTECT AF trial [12]. However, in patients with absolute contraindications to oral anticoagulation, dual antiplatelet therapy (aspirin plus clopidogrel) is used for six months postprocedure, as in the ASAP study, or for one to three months for patients at high risk of bleeding.

As of September 2022, FDA labeling for patients with nonvalvular AF who undergo WATCHMAN FLX LAAO placement was expanded to include 45 days of oral anticoagulant therapy plus dual antiplatelet therapy. Data supportive of this approach were from an analysis of more than 17,000 patients from the NCDR-LAAO Registry, which demonstrated no significant difference in rates of major adverse events at 45 days postimplantation, whether patients were discharged from the hospital on dual antiplatelet therapy, a DOAC and aspirin, or warfarin and aspirin.

●Alternative anticoagulation strategies

•Alternative imaging-based strategy – An alternative strategy was used in the nonrandomized, newer-generation WATCHMAN study (PINNACLE FLX), which confirmed device safety and efficacy [11]. This strategy included treatment with a DOAC through at least a 45-day follow-up, with apixaban or rivaroxaban strongly recommended. Patients were also prescribed concomitant low-dose (81 to 100 mg) aspirin. On evidence of adequate LAA seal (leak ≤5 mm) at the 45-day TEE evaluation, patients were directed to discontinue DOAC therapy and begin a dual antiplatelet therapy regimen of clopidogrel (75 mg) plus low-dose aspirin until six months postimplantation, followed by low-dose aspirin indefinitely. If a leak >5 mm was measured at the 45-day follow-up, patients continued DOAC plus aspirin and were reevaluated at six months postimplant. If there were no leaks >5 mm at the subsequent follow-up visit, patients could forego dual antiplatelet therapy and proceed straight to life-long low-dose aspirin. Postimplantation follow-up visits were required at 45 days and 6, 12, 18, and 24 months.

•Long-term, half-dose DOAC – This regimen versus standard anticoagulation treatment was tested in a nonrandomized study of 555 patients post-WATCHMAN device implantation [33]. The primary endpoint was a composite of device-related thrombosis, thromboembolic events, and bleeding. After 13 months, device-related thrombosis occurred in 12 patients, all in the full-dose DOAC group (3 versus 0 percent). The risk of nonprocedural major bleeding was more favorable in the half-dose DOAC group (0.5 percent versus 4 percent). The rate of the primary composite endpoint was higher in the standard treatment group (10 versus 1 percent). Further randomized studies are needed before we can make recommendations about the long-term, half-DOAC anticoagulation strategy.

●Postprocedural imaging – We obtain a transthoracic echocardiogram one day postprocedure to ensure there is no pericardial effusion. We also obtain a TEE between one and six months postprocedure to screen for or device-related thrombus (DRT), pericardial effusion, or a peridevice leak.

•DRT is present in approximately 2 to 5 percent of patients, depending on time of screening and device type, and is associated with risk of stroke or transient ischemic attack [16,20,34-36]. For patients with documented DRT, the optimal antithrombotic regimen is not known. Options for treatment of DRT include one of the following anticoagulants that may be combined with aspirin: DOAC (such as apixaban or rivaroxaban; avoid dabigatran) for 8 to 12 weeks, warfarin (target international normalized ratio 2 to 3) for 8 to 12 weeks, or low molecular weight heparin for two to four weeks [36].

•It is important to check for any residual communication between the LA and LAA after device closure. In some cases, the presence of a leak will prompt shared decision-making discussions with patients regarding continued oral anticoagulation or closure of the leak with other devices such as a coil.

It is uncertain if the size of the peridevice leak should guide further management, and prior data are mixed [37]. Previously, a peridevice leak of <5 mm was thought to be of little clinical significance (this amount of leak was arbitrarily used to define successful occlusion in device trials), whereas a leak ≥5 mm was an indication for continuing oral anticoagulation and/or additional endovascular procedure to close the residual leak (eg, with another percutaneous closure device or coils). Recent real-world, registry-based data suggest that small leaks may be associated with higher bleeding and thromboembolic risk [38]. (See 'Rationale and limitations' above.)

A study of over 51,000 people in the National Cardiovascular Data LAAO Registry showed that small (>0 to 5 mm) leaks after LAAO were associated with a modestly higher incidence of thromboembolic and bleeding events, and large leaks (>5 mm) were not associated with adverse events, although higher proportions of these patients were maintained on anticoagulation. In this study, small leaks were common; 25.8 percent had small leaks and 0.7 percent had large leaks. A higher proportion of patients with large leaks were on warfarin at 45 days versus small or no leak (44.9 versus 34.4 and 32.4 percent, respectively). At 6 and 12 months, anticoagulant use decreased but remained more frequent in patients with large leaks. Thromboembolic and bleeding events were uncommon in all groups. However, compared with patients with no leak, those with small leaks had slightly higher odds of stroke/transient ischemic attack/systemic embolization (hazard ratio [HR] 1.15; 95% CI 1.03-1.29), major bleeding (HR 1.11; 95% CI, 1.03-1.12), and any major adverse events (HR 1.12; 95% CI 1.05-1.16). There were no significant differences in adverse events between patients with large leaks and patients with small or no leaks.

After surgical occlusion

●Patients with no contraindication to anticoagulation – These patients are treated with anticoagulation indefinitely. Long-term anticoagulation in patients with AF is discussed separately. (See 'Patients undergoing surgery' above and "Atrial fibrillation in adults: Use of oral anticoagulants".)

●Patients with a contraindication to long-term anticoagulation – These patients are treated with anticoagulation (typically a DOAC) plus aspirin (81 to 325 mg daily) for six weeks to three months after the procedure, followed by TEE assessment of the completeness of occlusion.

Patients with an absolute contraindication to oral anticoagulants instead receive dual antiplatelet therapy (once-daily aspirin 81 to 325 mg plus clopidogrel 75 mg) during the pre-TEE period of six weeks to three months. If the TEE demonstrates that the LAA is completely occluded and long-term anticoagulation is contraindicated, anticoagulation is discontinued, and aspirin is continued indefinitely. If any leak is present after surgical LAAO, oral anticoagulation should optimally be continued.

This approach is supported by the results of an observational study of 72 patients undergoing surgical LAAO [39]. The study found that the risk of stroke or systolic embolism was higher in patients with incomplete LAA closure compared with those with complete closure (24 versus 2 percent during 44-month mean follow-up), despite similar rates of anticoagulation and recurrent AF. Among patients with incomplete closure, those with stroke or systemic embolism had smaller LAA neck diameters (2.8 versus 7.1 mm) than those without these adverse outcomes.

DEVICE SAFETY AND COMPLICATIONS — Percutaneous LAAO has become safer over time, possibly due to increased operator experience with the procedure and improvements in device iterations. Outcomes were compared between the 542 patients in the randomized study and 460 patients in the nonrandomized Continued Access Protocol registry who underwent WATCHMAN implantation after the randomized trial was completed [34]. A study done 10 years later showed a significant decline in the rate of procedure- or device-related safety events within seven days compared with those in the earlier randomized trial (3.7 versus 7.7 percent, respectively). In a large registry of over 38,000 WATCHMAN implantations, the procedural success rate was very high at 98.3 percent, which was even higher than reported in pivotal clinical trials such as PROTECT AF [40].

The complications of the LARIAT are described above. (See 'LARIAT system' above.)

There are several potential complications that can arise from the procedure itself or that are due to the presence of the device in the LAA. Some of these complications can occur in the acute periprocedural period, whereas others occur over a longer time frame. Management of these complications is crucial and underscores the importance of aggressive patient follow-up using appropriate imaging to look for such complications. Furthermore, understanding factors that lead to increased risk of complications is also important in terms of patient counseling and shared decision-making.

Complications of cardiac surgery are described in detail separately. (See "Postoperative complications among patients undergoing cardiac surgery".)

In patients undergoing percutaneous LAAO, these complications have been described:

●Death and cardiac arrest – Death and cardiac arrest are infrequent complications of WATCHMAN procedures. Among 38,158 patients who underwent LAAO in the National Cardiovascular Data LAAO Registry, in-hospital, procedural-related death occurred in 0.19 percent and cardiac arrest in 0.24 percent of patients [40].

●Vascular – The majority of periprocedural complications are vascular (ie, resulting femoral vein puncture). The incidence of such complications is infrequent with adoption of newer techniques to access the femoral vein and to stop bleeding after the procedure. Complications that may arise due to the transseptal puncture itself can also be minimized with ultrasound-guided puncture. Among 38,158 people who underwent LAAO in the National Cardiovascular Data LAAO Registry, the prevalence of a major vascular complication was 0.15 percent [40].

●Pericardial effusion – The LA and LAA are thin walled; thus, manipulation of transseptal equipment, large-bore sheaths, and devices can traumatize the wall, resulting in pericardial effusion. The incidence of pericardial effusion has declined over time, with improved operator skills; in the earlier PROTECT-AF study of WATCHMAN devices, the incidence of pericardial effusion requiring intervention was 4.3 percent [34], whereas in the more recent WATCHMAN clinical trials and registries, the pooled incidence of cardiac tamponade was about 1.3 percent [41,42]. In a report from the National Cardiovascular Data LAAO Registry, among 65,355 patients, 881 (1.3 percent) developed procedural pericardial effusions, percutaneous drainage was performed in 1.18 percent, and cardiac surgery was performed in 0.27 percent of the study population [42]. Pericardial effusion not requiring intervention was reported in 669 patients (1.02 percent). Signs and symptoms of a pericardial effusion and tamponade are discussed separately. (See "Pericardial effusion: Approach to diagnosis", section on 'Diagnostic approach'.)

●Periprocedure stroke – Peridevice leaks (also called device-related leaks), air embolism, and device-related thrombus can lead to near-term stroke.

Anticoagulation and antiplatelet therapy after the procedure are discussed above. (See 'Postprocedure management' above.)

In the National Cardiovascular Data LAAO registry, in-hospital ischemic stroke was documented in 0.12 and hemorrhagic stroke in 0.01 percent of patients, respectively [40]. Most events were related to air embolism from inadequate device prepping. Fewer complications are due to thromboembolism from preexisting or de-novo equipment-related LAA thrombus. Air embolism cerebral circulation can cause a transient ischemic attack or stroke. Once a thrombus is detected, additional heparin should be administered, and thrombus aspiration via a large sheath may be attempted. The procedure should be completed expeditiously with either deployment of the device or removal of the sheath/device from the LA. Full neurological assessment should be performed when the patient awakes from general anesthesia, and emergent stroke team consultation is required if the patient has neurological deficits.

●Device embolization – Device embolization is a rare but serious complication of LAAO. With the legacy WATCHMAN device, the reported incidence was 0.24 percent in clinical trials [41], with an even lower incidence reported in the real-world National Cardiovascular Data LAAO registry at 0.07 percent [40]. Most embolizations do not cause symptoms; however, palpitations, heart failure, hypotension, and cardiac arrest had been reported. The first-line strategy for management is percutaneous retrieval of the embolized device.

●Device erosion – This complication is very rare (occurring in <1 percent or patients) but can be devastating. Long-term follow-up with appropriate imaging (ie, a transesophageal echocardiogram [TEE] between one and six months postprocedure) to look for new pericardial effusion is important to screen for this complication [43].

The following issues regarding LAAO adverse events and hospital readmission are also important to recognize:

●Greater adverse events in females – Females have a higher risk of in-hospital adverse events after LAAO [44]. In the National Cardiovascular Data Registry, 49,000 patients (41 percent female) underwent LAAO between 2016 and 2017. Females were more likely than males to experience any adverse event (6 versus 4 percent), including a pericardial effusion requiring drainage (1.2 versus 0.5 percent) or major bleeding (1.7 versus 0.8 percent). Also, female patients were more likely to have a hospital stay >1 day (16 versus 12 percent) or experience death (adjusted odds ratio 2.01; 95% CI 1.31-3.09), although death was rare and absolute differences minimal (0.3 versus 0.1 percent). Female patients in this study were older than male patients but less likely to have diabetes and heart failure.

●High readmission – Among 14,000 patients with LAAO in the Nationwide Inpatient Sample between 2016 and 2017, the 30-day readmission rate was 9.4 percent [45]. Of these, 12 percent of readmissions were attributable to gastrointestinal bleeding. Risk factors for readmission included drug abuse ( odds ratio 4.1; 95% CI 1.34-12.54) and anemia (OR 1.88; 95% CI 1.12-3.18).

RECOMMENDATIONS OF OTHER GROUPS — The LAAOS III trial on surgical LAAO was completed after major society guidelines were published, so this evidence was not incorporated in those guidelines. (See 'Patients undergoing surgery' above.)

A consensus statement from the European Heart Rhythm Association and the European Association of Percutaneous Cardiovascular Interventions published in 2020 states that LAAO is acceptable in patients with a contraindication to oral anticoagulation or those who are unwilling or unable to take oral anticoagulation [46].

The United States Food and Drug Administration (FDA) approval is somewhat more liberal than the European Society of Cardiology guideline for patients eligible for oral anticoagulation and has other indications for nonpharmacologic therapy (eg, lifestyle indications, labile international normalized ratio).

The 2019 focused update of the 2014 American Heart Association/American College of Cardiology/Heart Rhythm Society guideline for the management of patients with AF makes a weak recommendation for surgical excision of LAA at the time of cardiac surgery [47,48]. In addition, it states that percutaneous LAAO may be considered in patients with AF at increased risk of stroke who have contraindications to long-term anticoagulation.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Atrial fibrillation" and "Society guideline links: Arrhythmias in adults".)

SUMMARY AND RECOMMENDATIONS

●Rationale and limitations – Among patients with atrial fibrillation (AF), thrombus in the left atrial appendage (LAA) is the primary source for thromboemboli. This is the rationale for excluding the appendage in selected patients. (See 'Rationale and limitations' above.)

However, thrombogenesis in patients with AF may not be limited to the LA, and the procedure does not always result in complete occlusion of the LAA.

●Indications

•Patients not undergoing cardiac surgery – For patients with AF with an indication for anticoagulation (based upon risk of stroke and other systemic thromboembolism (table 1)) but who have a contraindication to long-term anticoagulation, we suggest a percutaneous LAA occlusion (LAAO) procedure (Grade 2B). Among these devices, we prefer the WATCHMAN. (See 'Patients not undergoing cardiac surgery' above and 'WATCHMAN device' above.)

This approach does not apply to patients with an indication for anticoagulation other than AF (eg, an implanted mechanical valve). (See "Atrial fibrillation in adults: Selection of candidates for anticoagulation".)

•Patients with AF and a CHA₂DS₂-VASc score of at least 2 (table 1) who are undergoing cardiac surgery for another indication

For such patients, if there is no contraindication to long-term anticoagulation, we recommend concomitant surgical LAAO. (Grade 1B) (See 'Patients undergoing surgery' above.)

For such patients, if there is a contraindication to long-term anticoagulation, we suggest concomitant surgical LAAO (Grade 2C). (See 'Patients undergoing surgery' above.)

●Preprocedure planning – Specific LAA measurements can be useful for LAA closure planning; these can be obtained by transesophageal echocardiogram (TEE), cardiac computed tomography angiography, or cardiac magnetic resonance imaging. If cardiac imaging shows an LA/LAA thrombus, the LAAO procedure is contraindicated. (See 'Preprocedure planning' above.)

●Percutaneous devices – The WATCHMAN device is the most commonly implanted percutaneous LAAO device and has the most robust data to support its use (table 3). (See 'Percutaneous devices' above.)

●Surgical devices – Types of surgical closure include stapler occlusion, amputation and suture, and epicardial device closure (table 2). (See 'Surgical LAA closure' above.)

●Postprocedure management

•After percutaneous closure

We treat with short-term antithrombotic therapy (anticoagulant and antiplatelet therapy) after placement of the device. Patients with an absolute contraindication to oral anticoagulants instead receive dual antiplatelet therapy postprocedure.

We obtain a TEE between one and six months postprocedure. Device-related thrombus and/or peridevice leak ≥5 mm may be indications for continued anticoagulation. (See 'After percutaneous closure' above.)

•After surgical occlusion

Patients with a contraindication to long-term anticoagulation – These patients are treated with short-term anticoagulation (typically a DOAC) plus aspirin for six weeks to three months after the surgery, followed by TEE assessment of the completeness of occlusion. Patients with an absolute contraindication to oral anticoagulants instead receive dual antiplatelet therapy during the pre-TEE period of six weeks to three months.

If the TEE demonstrates that the LAA is completely occluded and long-term anticoagulation is contraindicated, only aspirin is continued indefinitely. If any leak is present postsurgical LAAO, oral anticoagulation should be continued. (See 'After surgical occlusion' above.)

Patients with no contraindication to anticoagulation – These patients are treated with anticoagulation indefinitely regardless of whether a rhythm control intervention is performed. (See 'Patients undergoing surgery' above.)

●Complications – These are rare but include pericardial effusion/tamponade, vascular injury, device erosion or embolization, stroke, death, and cardiac arrest. (See 'Device safety and complications' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Alan Cheng, MD, who contributed to earlier versions of this topic review.