Indication

Oral anticoagulation is contraindicated for at least an estimated 30% of patients with AF. Furthermore, it has been shown that contraindications for OAC are particularly common in older patients.⁵ Interventional occlusion of the LAA is suggested for patients with a high risk of embolism and a contraindication for OAC. Often the so-called CHADS2 score is used for defining the risk of embolism.

In addition to assessing the patient’s individual thromboembolic risk, the following factors need to be addressed:

Hence, any decision for an interventional occlusion of the LAA has to be an individualized one: It has to balance the advantages and possible disadvantages of the therapy.

Recently, the U.S. Food and Drug Administration advisory panel recommended the approval of the Watchman LAA closure device, stating that it is comparable to long-term warfarin therapy for the prevention of stroke in warfarin-eligible patients with nonvalvular AF.

Devices and Implantation Techniques

The first device for the interventional occlusion of the LAA was introduced by Lesh and co-workers (animal studies).⁹ The device was called PLAATO (percutaneous left atrial appendage transcatheter occlusion). It consists of a self-expanding, balloon-shaped nitinol cage with an expanded polytetrafluoroethylene (ePTFE) membrane with three rows of anchors stabilizing the device in the appendage (Figure 99-1). The membrane covers the atrial surface of the device, whereas the opposite surface is not covered, which allows secondary thrombosis of the lumen by the device. The morphology of the LAA is evaluated by angiography and simultaneous TEE. The maximal diameter of the LAA orifice is determined by both methods and a mean diameter obtained. A 20% to 30% oversized occlusion device is selected. The device is introduced via a 14-F femoral vein sheath and advanced to the LAA after trans-septal puncture of the atrial septum. Then, the device is expanded and the position is controlled using both angiography and TEE (Figure 99-2). The mean size of the devices used is 29 mm. After verifying the optimal position and adequate sizing of the device, it is released and the final position assessed by repeat angiography.

FIGURE 99-1 First left atrial appendage occlusion device (PLAATO, ev3). It consists of a self-expanding, balloon-shaped nitinol cage with an expanded ePTFE membrane with three rows of anchors.

FIGURE 99-2 Left atrial appendage occlusion device as imaged by fluoroscopy and echocardiography.

A similar device, called Watchman device, was introduced by Atritech (Plymouth, MN) (Figure 99-3).¹² It is similar to the PLAATO device and consists of a self-expanding nickel-titanium frame structure with external fixation barbs and a permeable polyester fabric cover (sizes ranging from 21 to 33 mm). The procedure of implanting the device is similar to that of the implantation of the PLAATO device.

FIGURE 99-3 Watchman device (Atritech) with a self-expanding nitinol frame structure with external fixation barbs.

Meier and coworkers published a study on the successful occlusion of the LAA using an Amplatzer device, which was not designed uniquely for this purpose.¹³ In December 2008, AGA Medical (Plymouth, MN) received the European Conformite Europeene mark of approval for an LAA occlusion device that is different from the previously described ones. It is called a cardiac plug and is designed to achieve optimal occlusion of the LAA; it consists of a distal lobe and disc that is designed to cover the orifice of the LAA (Figure 99-4). The diameter of the lobe ranges from 16 to 30 mm, and the corresponding disc diameter ranges from 20 to 36 mm.

FIGURE 99-4 Cardiac plug left atrial appendage occlusion device (Atritech). It consists of a distal lobe and a disc.

Safety and Results

In various studies, data about the safety and success of LAA occlusion were collected for both the PLAATO device (Figure 99-5) and the Watchman device.^12,14

FIGURE 99-5 Postmortem image of PLAATO device in the left atrial appendage.

The researchers evaluating the PLAATO device showed that the procedure is feasible and may be performed within 2 hours.¹³ Detailed results concerning 111 patients with attempted LAA occlusion were published in 2005.¹³ The procedure was successful in 97% of the patients. Migration of the device or formation of mobile thrombi on the surface of the device were not observed during follow-up 10 months later. One of the patients in the study group died because of procedural complications. Pericardiocentesis was performed in three patients for cardiac tamponade. In addition, one patient developed left-sided hemothorax. Hence, relevant procedural complications were not uncommon. However, they occurred less frequently with increasing experience of the operator.

Two patients had a stroke, and two patients experienced transient ischemic attacks during the observation period. No thrombi formed on the device in any of the patients. Another study followed up the patients closely via echocardiography and showed that the implant was covered with a neointima after 6 months and that the device did not alter pulmonary vein flow.¹¹ Furthermore, no significant atrial septal defects were detected after the puncture of the intra-atrial septum with an F-14 sheath.¹⁵ Because of the high rate of complications, a randomized study comparing anticoagulation and device therapy was requested. Such a study has not been performed yet, and the company has halted distribution of the device. Long-term follow-up data have not systematically been published yet.^16–18

Similar experiences were reported for the Watchman device from Atritech. Sick et al provided information on 75 patients who had the device implanted.¹² A failure rate of 10% was reported for implantation of the device. Two patients experienced device embolization. The devices were successfully retrieved percutaneously in both cases. Two cardiac tamponades and one air embolism were reported as well. One device had a wire fracture and needed to be explanted surgically. In four patients, flat thrombi had formed on the atrial surface of the device by the time of the 6-month echocardiographic follow-up. Two neurologic complications occurred during the follow-up period. To assess how effectively strokes and bleeding events are reduced in patients with AF by occluding the LAA with the Watchman device, the prospective Embolic Protection in Patients with Atrial Fibrillation (PROTECT-AF) study was performed in patients with nonvalvular AF and a CHADS2 score greater than 1.¹⁹ It is a key study for evaluating the concept of LAA occlusion for the prevention of thromboembolism in patients with AF, and hence the study will be discussed in detail in this chapter.

The trial included patients aged 18 years or older with all types of nonvalvular AF and at least one risk factor on the CHADS2 risk score. Patients were excluded if they had contraindications for warfarin or had comorbidities requiring the use of long-term OAC. In addition, all patients with an LAA thrombus, a patent foramen ovale with atrial septal aneurysm and right to left shunt, mobile atheroma, and symptomatic carotid artery disease were excluded from the study. Patients were randomized to either interventional occlusion of the LAA or warfarin therapy in a 2 : 1 ratio. After implantation of the Watchman device, patients were treated for 45 days with warfarin. Thereafter, clopidogrel (75 mg) and aspirin (81 to 325 mg) once daily were prescribed for the remaining period up to the 6-month follow-up visit. Then clopidogrel was discontinued. The primary endpoint of the study was to show non-inferiority of the device compared with warfarin therapy by use of a composite endpoint consisting of efficacy (ischemic and hemorrhagic stroke), cardiovascular or unexplained death, or systemic embolism. The safety endpoint consisted of excessive bleeding events and procedure-related complications. Patients were followed up for an aggregate of 1065 patient-years. Successful implantation was performed in 88% of patients scheduled for intervention. An international normalized ratio range of 2 to 3 was achieved in 66% of patients in the medical treatment group. The primary efficacy event rate was 3 per 100 patient years in the intervention group and 4.9 in the control group. The probability of non-inferiority of the intervention was 99.9%. Hemorrhagic strokes occurred in 0.1% in the intervention group and in 1.6% in the control group.

Primary safety endpoints occurred at a significantly higher rate in the intervention group than in the control group (7.4 vs. 4.4) and were associated mainly with procedural complications, that is, 22 patients had serious pericardial effusions. Device embolization was observed in three patients. The rate of serious pericardial effusion declined with increasing experience of the operator.

In summary, the results of the PROTECT-AF trial demonstrated the non-inferiority of LAA occlusion compared with warfarin therapy for the primary efficacy endpoint. However, a higher initial safety event rate is observed for device implantation.

Other studies on LAA occlusion included far fewer patients. Meier et al provided information on the use of the Amplatzer device for occluding the LAA in 16 patients.¹³ One patient had an embolization of the device. All other patients had no serious complications.

Other devices are currently being assessed in animal models. Fumoto et al reported on the successful implantation of a third-generation atrial exclusion device in 14 dogs.²⁰ Jayakar et al investigated a tissue welding technology to obliterate the LAA.²¹

Importantly, long-term data or data from randomized prospective trials are not available yet.