This article is more than 5 years old. Information may no longer be current.
Strike the Stroke: Left Atrial Appendage Occlusion
Click Here to Manage Email Alerts
A stroke frequently has catastrophic consequences for patients and their families and is accompanied by a significant financial burden to society. It remains the leading cause of disability and third leading cause of death in developed countries.
Therefore, treatment and prevention should be a top priority. A number of technological advances have brought both acute stroke therapy and stroke prevention strategies into focus for interventional cardiologists. A recent preventive strategy has been percutaneous or minimally invasive left atrial appendage (LAA) occlusion in patients with atrial fibrillation.
Why We Need Percutaneous LAA Occlusion
Approximately 15% of all strokes are attributed to AF. Ninety percent of left atrial thrombi are located in the LAA in patients with nonvalvular AF. AF is associated with a 5% annual stroke risk with variations depending on age and risk factor profile. Although oral anticoagulation with warfarin dramatically reduces this risk by more than half, it is also associated with a significant bleeding risk. In fact, most conditions that increase the risk for stroke (prior stroke, hypertension and age) are, at the same time, important predictors for hemorrhage while on oral anticoagulation. Hence, patients who would benefit most from oral anticoagulation are also most likely not to receive it.
Although novel anticoagulants, such as dabigatran (Pradaxa, Boehringer-Ingelheim) and rivaroxaban (Xarelto, Janssen Pharmaceuticals), appear to have at least equivalent efficacy for stroke prevention and are not accompanied by logistical challenges related to monitoring or drug-drug interactions, the risk for major bleeding at the currently recommended dosages does not differ much from that seen with conventional warfarin. Therefore, despite novel anticoagulant agents, a considerable number of patients at risk for stroke will remain untreated. Likewise, pulmonary vein isolation for AF, although promising in those who are fortunate to have a permanent cure, is frequently unsuccessful, and randomized trials demonstrating equivalency to anticoagulation have not yet been performed. As a result, alternative treatment strategies are needed.
Current LAA Occlusion Strategies
Currently, options that provide LAA occlusion include surgical closure with suture or staple ligation or excision, percutaneous device closure and minimally invasive suture closure. Although surgical closure is frequently performed in patients with AF and open chest surgery for other reasons, it has little merit as a stand-alone therapy because the risk of open chest surgery for the sole purpose of LAA ligation or excision would offset any benefits.
More importantly, open surgical approaches have been fraught with high rates of incomplete closures either due to a residual leak or pouch. Residual leaks have been demonstrated to cause thrombus formation in the supplied cavity and may be associated with an increased embolic event rate. Reasons for incomplete surgical closures are technical challenges, including the proximity to the left circumflex coronary and pulmonary artery, as well as the absence of real-time imaging of a beating heart with a well-filled (undecompressed) appendage. Importantly, despite making sense intuitively and its routine use in patients with AF concomitantly with other open chest surgeries, to date, there are no data to support its role in stroke prevention.
With these limitations, the prospect of a percutaneous approach to LAA occlusion becomes even more appealing, especially given its two main advantages: It can be performed without the use of general anesthesia and cardiopulmonary bypass; and it is performed on a beating heart with real-time imaging of a well-filled LAA with direct feedback of treatment results prior to device release. Three devices have been used in larger clinical trials: the percutaneous left atrial appendage transcatheter occlusion system (PLAATO, ev3), Watchman LAA occluder (Atritech/Boston Scientific) and Amplatzer Cardiac Plug (ACP; St. Jude Medical).
Figure 1. PLAATO device was shown to be
feasible and safe in nonrandomized studies.
However, the device was withdrawn by the
manufacturer due to commercial reasons.
For a larger image, click here.
Image: ev3
PLAATO
Although the PLAATO device (Figure 1) is no longer available, nonrandomized trials have demonstrated the feasibility and safety of percutaneous LAA closure and suggested a more than 50% lower stroke risk at short- and long-term follow-up compared with the risk calculated based on the CHADS2 risk score. It is noteworthy that the lower risk was observed despite an absence of routine, including temporary, anticoagulation, as enrolled patients were considered at prohibitive risk for anticoagulation. With the use of the PLAATO device, several challenges of device-based closure surfaced, including the risk of device embolization, LAA injury causing pericardial tamponade, incomplete closure and peri-device leaks.
Watchman
The limitations of PLAATO led to the design of newer-generation occluders, including the Watchman occluder (see Figure 2). Initial results of feasibility and safety trials with the Watchman occluder were encouraging and prompted the currently largest randomized trial comparing it to conventional anticoagulation: PROTECT AF. This trial included 707 patients and demonstrated for the first time the noninferiority of percutaneous LAA device closure compared with conventional anticoagulation using warfarin. The overall stroke rate was equivalent to warfarin (2.3% after device closure vs. 3.2% at 18 months of follow-up). Trial participation mandated continuation of warfarin for 45 days after device implantation with discontinuation only if no thrombus was detected and any peri-device leak was less than 5 mm in diameter on color Doppler examination at 45 days. Thereafter, patients were treated with aspirin and clopidogrel (Plavix, Sanofi-Aventis) until 6 months post-procedure followed by aspirin indefinitely. Importantly, at a mean follow-up of 45 months, device closure was superior to warfarin with a significant reduction in all-cause mortality and equivalent safety profile.
The Watchman device has a CE mark and is routinely used in some European countries in patients with AF who have high stroke and bleeding risks. Design improvements have been made, including to the nitinol cage (18 compared with its previous 10 cell design), facilitating compression and perhaps apposition to the LAA ostium, and to the fixation barbs and bumper system potentially improving safety of delivery. These improvements also allow recapture prior to release and redeployment. The newest-generation device is currently undergoing a clinical trial (EVOLVE; www.clinicaltrials.gov identifier: NCT01196897).
The Watchman occluder does not have FDA approval for routine use pending data from the PREVAIL trial. This trial had a very similar design to PROTECT AF, randomizing patients with AF and indications for anticoagulation either to closure with the Watchman device (followed by 45 days of warfarin and, provided adequate closure is demonstrated at follow-up, antiplatelet therapy thereafter) or anticoagulation. The purpose was to determine whether the efficacy of PROTECT AF and the low complication rate of the subsequent CAP registry (described below) can be reproduced with newer-generation Watchman occluders and other operators.
Figure 2. Watchman left atrial appendage
occluder is CE-mark approved and currently
being tested in the PREVAIL trial.
For a larger image, click here.
Image: Boston Scientific
The preliminary results of PREVAIL were publically released during the American College of Cardiology Scientific Sessions in March. Four hundred and seven patients were enrolled. The implantation success rate was higher compared with PROTECT AF (95% vs. 91%), and the composite of pericardial effusion, stroke, device embolization and other vascular complications was significantly lower than in PROTECT AF (4.4% vs. 8.7%). Of note, the incidence of pericardial effusion requiring surgical repair was also lower than in PROTECT AF (0.4% vs. 1.6%). Importantly, there was no difference in the success and complication rates between operators with prior experience in LAA closure and new operators. The 18-month stroke, systemic embolism or CV/unexplained death rate did not differ between the device and control group in those patients (n=88) who completed follow-up at this point, despite a lower than expected stroke rate in the control group.
In both PROTECT AF and PREVAIL, temporary (45-day) anticoagulation was standard therapy and patients with absolute contraindications to warfarin were excluded. This raises the question whether LAA occluders can be implanted safely in the absence of anticoagulation (eg, with antiplatelet therapy only). The question has not been addressed in randomized trials. However, results from the ASAP registry suggest that Watchman occluder implantation in the absence of even temporary anticoagulation is safe. One hundred and fifty patients (mean CHADS2 score, 2.8) underwent Watchman occluder implantation. Aspirin and clopidogrel were administered for 6 months followed by aspirin indefinitely. Mean follow-up was 14 months. The annual ischemic stroke rate was 1.7%, substantially (77%) lower than predicted based on the CHADS2 score. Furthermore, nonrandomized data using the PLAATO device suggest a low event rate. As previously mentioned, in the PLAATO trial, patients did not routinely receive warfarin.
Amplatzer Cardiac Plug
Another newer-generation occluder is the ACP (Figure 3), which has received a CE mark and is used in some European countries. Randomized controlled data are not yet available; however, in a retrospective study including 137 ACP patients, 96% underwent successful implantation. Newer-generation devices are undergoing clinical studies, including the randomized ACP trial, a randomized trial comparing percutaneous LAA occlusion with the ACP to anticoagulation (www.clinicaltrials.gov identifier: NCT01118299).
Figure 3. Amplatzer Cardiac Plug obtained a
CE mark and is used in some European
countries.
For a larger image, click here.
Image: St. Jude Medical
Lariat
The most recent addition to the concept of LAA closure is minimally invasive suture ligation of the LAA using the Lariat (SentreHEART) device. Femoral venous access is obtained, followed by transseptal puncture. A wire, at the tip of which is a magnet, and a compliant balloon just proximal to the magnet are advanced into the LAA. After epicardial access, a sheath is advanced into the pericardial space and a wire with a magnet located at its tip is advanced to connect with the wire located within the LAA. Using both wires as a rail, a loop is positioned around the LAA ostium and the balloon in the LAA is inflated at very low pressure to mark the appendage ostium and guide loop positioning. Subsequently, the loop is tightened and the balloon deflated under transesophageal and fluoroscopic guidance and the LAA wire removed. Finally, a suture is delivered at the cinched loop to ligate the appendage.
A nonrandomized study including 89 patients has been performed using this concept demonstrating successful implantation in 96% of patients and complete closure (defined as no leak or leaks <1 mm) in 98%. The obvious advantage is closure without implantation of potentially thrombogenic foreign material. Suitability for this concept currently depends to great extent on the LAA anatomy (size, configuration and its relationship to the pulmonary artery), mandating prior CT imaging and reconstruction of the LAA to determine candidacy for this approach. Lariat is currently FDA approved and the only device that can be used on a routine basis for LAA closure in the United States.
Potential Complications with LAA Closure
There are several potential problems that can arise with LAA closure that operators should keep in mind. First and foremost is perforation, either of the appendage (eg, during catheter and device manipulation within the appendage) or atrial free wall (eg, during attempts of transseptal puncture) causing pericardial tamponade, device embolization and periprocedural stroke (either due to air embolism or equipment associated thrombus formation). During PROTECT AF, the incidence of pericardial tamponade was 4.8%, 0.6% for device embolization and 1% for periprocedural stroke.
It has become clear, however, that with increasing operator experience, these complications are considerably less frequent. For example, a lower complication rate was seen toward the end compared with the beginning of PROTECT AF. Further, in the subsequent CAP registry, the above complications were substantially lower in the 406 patients treated than in PROTECT AF (pericardial hemorrhage, 2.3% vs. 4.8%; perioperative strokes, 0% vs. 1%).
The periprocedural complication rate reported with ACP implantation is similar to that reported with the Watchman occluder. In 137 attempted (132 successful) LAA closures with the ACP, there were three ischemic strokes, two device embolizations with successful percutaneous recapture and five clinically significant pericardial effusions. Although unproven, technical success and device efficacy may also be dependent on operator experience because of a learning curve regarding optimal device positioning. Satisfactory device positioning may reduce the incidence of residual leaks, pouches or shoulder overhang potentially lowering device-associated thrombus formation. However, this requires familiarity with the appropriate transseptal puncture location, delivery sheath choice and catheter manipulation to allow coaxial alignment of the delivery system. It also mandates an understanding of the variable 3-D LAA configuration.
Another concern is that there are several potential problems extending beyond the perioperative period that may occur. These include device-associated thrombus formation and incomplete closure. The incidence of device-associated thrombus formation is small. It occurred in 20 of 478 patients (4%) in PROTECT AF, of whom three had an ischemic stroke. Of note, thrombus formation is not specific to the device, but has also been reported with the ACP and PLAATO devices. The obvious concern is cerebral embolization. Most thrombi are clinically silent and detected on echocardiographic follow-up. Therefore, routine 45-day, 6-month and 12-month transesophageal echocardiography is recommended. Most thrombi resolve with temporary anticoagulation.
Incomplete closure, defined as any peri-device leak, is a more common phenomenon. It has been described in 57% (at 1-month follow-up with a lower rate of 32% at 12 months) of patients after Watchman and 80% after PLAATO implantation and is the consequence of a discrepancy between the natural configuration of the LAA orifice (typically oval) and the shape of both currently available occluders (round). The overwhelming majority is small (≤3 mm). Larger leaks (>3 mm) occur in approximately 9% of all closures with the Watchman device. In PROTECT AF, patients with leaks >5 mm were treated as if the LAA had not been occluded (ie, it was recommended to continue anticoagulation indefinitely). Patients with leaks <5 mm were treated identically to those without leaks (45 days of warfarin, followed by antiplatelet therapy only thereafter). Importantly, the presence of leaks did not affect the risk for cerebral and systemic embolic events. However, these data were limited by the small patient number and low event rate.
As a result, definitive conclusions regarding the safety of antiplatelet therapy only, particularly in those with large leaks, cannot be made. Therefore, in the absence of absolute contraindications to anticoagulation, it may be prudent to consider a device closure with a leak >5 mm as unsuccessful and to continue oral anticoagulation until we have a better understanding of the consequences of large leaks. An alternative, although not well studied yet, may be the implantation of a second device provided the anatomic configuration allows it. Of note, similar to device-associated thrombi, leaks are not unique to the Watchman device and have been reported after ACP implantation, as well as minimally invasive suture closure (described above).
Conclusion
Major advances have recently been made to occlude the LAA percutaneously or use a minimally invasive suture technique. Data are now available demonstrating superiority of device closure using the Watchman device over traditional anticoagulation. More data with this device and others are pending completion of ongoing trials. However, only if we continue our efforts to improve device designs, facilitating implantation and assuring optimal device positioning and minimizing potential problems, including peri-procedural complications and incomplete closure, as well as device-associated thrombus formation, will this technology meet up to its promise of routine, safe and reliable stroke prevention in patients with AF.
Disclosure: Sievert reports that his institution has ownership interest in or has received consulting fees, travel expenses, stock options and/or study honoraries from Abbott, Access Closure, AGA, Angiomed, Arstasis, Atritech, Atrium, Avinger, Bard, Boston Scientific, BridgePoint Medical, Cardiac Dimensions, CardioKinetix, CardioMEMS, Coherex, Contego, CSI, EndoCross, EndoTex, Epitek, Evalve, ev3, FlowCardia, Gore, Guidant, Guided Delivery Systems, HLT, InSeal Medical, Lumen Biomedical, Kensey Nash, Kyoto Medical, Lifetech, Lutonix, Medinol, Medtronic, NDC, NMT, OAS, Occlutech, Osprey, Ovalis, Pathway, PendraCare, Percardia, Pfm Medical, Rox Medical, Sadra, SMT, Sorin, Spectranetics, SquareOne, Trireme, Trivascular, Velocimed, Veryan and Vessix. All other authors report no relevant financial disclosures.