We conducted 2 feasibility studies to assess the performance and safety of the percutaneous left atrial appendage (LAA) transcatheter occlusion system for the prevention of stroke in high-risk patients with nonrheumatic atrial fibrillation. Results showed that percutaneous LAA occlusion can be performed using the device at acceptable risk. This procedure may be particularly useful for patients at increased risk of ischemic stroke with a contraindication to anticoagulation therapy.
Patients with atrial fibrillation are at increased risk for cardioembolic stroke because insufficient contraction of the left atrium leads to stasis of blood and the subsequent development of thrombi.1,2 Up to 90% of these thrombi begin in a small sacculation that has been described as “our most lethal human attachment”: the left atrial appendage (LAA).3,4 The LAA is a derivative of the primitive atrium and is located between the left upper pulmonary vein and the left ventricle. Its surgical closure during mitral valve surgery as a stroke preventive measure has been performed for many decades and is recommended in the American College of Cardiology/American Heart Association guidelines.5,6 Recently, it has been suggested that LAA occlusion be performed during coronary artery bypass graft surgery.7
Percutaneous occlusion of the LAA is a new, independent, minimally invasive approach to stroke prevention in patients with atrial fibrillation. In 2 prospective multicenter trials, we assessed whether LAA occlusion using the new percutaneous LAA transcatheter occlusion (PLAATO) system is feasible and can be performed at acceptable risk.8
Patients and methods
The PLAATO system is comprised of a self-expandable nitinol cage, 18 to 32 mm in diameter, which is coated with expanded polytetrafluoroethylene. The device is inserted into the orifice of the LAA with a venous 12-French transseptal sheath using fluoroscopy and transesophageal echocardiography (TEE). To gain access to the left atrium, transseptal puncture is performed. At all times, during the procedure until release, the occluder may be traded for a different size, collapsed, or repositioned. It is necessary for the selected occluder to be 20% to 50% greater in size than the ostium of the LAA, as calculated using TEE and angiography.
Left atrial appendage occlusion was evaluated immediately after de­vice expansion by proximal and distal dye injections as well as TEE examination. After the device was released, a final proximal contrast study showed the degree of LAA occlusion. Postprocedure, patients received 300 to 325 mg/day of as­pirin indefinitely and 75 mg/day of clopidogrel bisulfate (Plavix) for up to 6 weeks. They also received prophylaxis treatment for subacute endocarditis for 6 months. At 1 month, 6 months, and 12 months, patients received clinical and echo­cardiographic examinations, as well as chest x-rays.
Percutaneous LAA occlusion using the PLAATO system was undertaken in 111 patients (mean age, 71 ± 9 years). All patients had nonrheumatic atrial fibrillation for 3 or more months, experienced a prior stroke or transient ischemic attack, and/or had at least 1 of the following risk factors for ischemic stroke: a blood flow velocity of 20 cm/sec in the LAA or dense or moderate spontaneous
contrast, coronary artery disease, age of 65 years or older, hy­pertension, diabetes, and congestive heart failure. All of the patients had contraindications to lifelong anticoagulation therapy according to the warfarin sodium (Coumadin) product label.
Patients were excluded from the studies if they had mitral or aortic valve stenosis or regurgitation, other sources of possible thromboembolic events (such as aortic plaques and carotid disease), and thrombus formation in the left atrium or LAA.
The PLAATO system was successfully implanted in 97.3% of patients (108 of 111). Slightly more than half of the implantations were done using general anesthesia, and almost half were done under conscious sedation. The procedure was not completed in 3 patients because of groin vessel perforation in 1, presence of a thrombus formation in the left atrium in another, and a cardiac tamponade in the third. Four patients experienced pericardial effusion or cardiac tamponade, which was treated by pericardiocentesis in 3 of the patients. One pa­tient also experienced a hemothorax and another a pleural effusion.
Most patients received a 29-mm device (median size, 29 mm), and the mean LAA diameter was 22 ± 4 mm. The mean fluoroscopy time was 18 ± 9 minutes, and the mean procedure time was 68 ± 28 minutes. Angiography immediately after the procedure showed a sufficient seal of the LAA in all 108 patients in whom the PLAATO device was successfully implanted. At 6 months, 98% of the patients with assessable TEE showed successful LAA occlusion.
One patient developed a laminar thrombus formation on the occluder, which was detected at the 6-month follow-up visit. The material seemed to be firmly attached to the device surface and was not floating or mobile. The thrombus resolved following 6 months of treatment with clopidogrel and aspirin. No dislocation of the device or impairment of the mitral valve or pulmonary vein inflow was shown in any patient on follow-up echocardiograms.
During a follow-up period of 91 patient-years, 6 patients died. These deaths were not associated with the procedure or the PLAATO system. Two patients had a stroke, which resulted in an annual stroke rate of 2.2% after successful LAA occlusion.
The LAA is the most common source of cardiac thrombus formation in patients with atrial fibrillation. Although more than half (57%) of left atrial thrombi are located in the LAA in patients with valvular atrial fibrillation, they almost exclusively originate in the LAA in patients with nonrheu­matic atrial fibrillation (91%).3 The use of anticoagulation therapy is extremely effective and superior to treatment with aspirin for avoiding this thrombus formation and subsequent cardioembolic events in patients with atrial fibrillation.9 However, warfarin is often underprescribed or ad­ministered in subtherapeutic dosages in patients with atrial fibrillation. Even among patients who are at moderate or high risk for ischemic stroke, warfarin is prescribed for only 40% of patients.10 The reasons for this are diverse and range from clinical uncertainty or the patient’s refusal of treatment, to the concern that the benefit of anticoagulation does not outweigh the increased risk of bleeding complications, especially intracranial hemorrhage.11 Furthermore, up to almost 40% of atrial fibrillation patients have relative contraindications to anticoagulation therapy.10
A new pharmacologic approach to the prevention of thrombus formation is ximelagatran, an oral direct thrombin inhibitor. Even though this medication was shown to be noninferior to warfarin during the initial trials, its administration was associated with a possible increased risk of coronary artery disease and hepatotoxicity, and therefore, it has not yet been ap­proved.12 Furthermore, pa­tients with contraindications to warfarin will not be optimal candidates for this drug.
Because the LAA plays a major role in cardioembolic events, its occlusion is supposed to reduce the risk of ischemic stroke. Garcia-Fernandez and colleagues showed that effective LAA ligation in patients with rheu­matic valvular disease who underwent mitral valve replacement was associated with a 12-fold reduction in em­bolic risk.13 However, surgical ligation does not always lead to an effective occlusion; it may even promote en­hancement of spontaneous echo­con­trast and thrombus formation inside the remaining cavity of the LAA and thus embolic stroke.14
Percutaneous occlusion of the LAA is a new minimally invasive stand-alone approach for stroke prevention in atrial fibrillation patients without the burden of lifelong anticoagulation treatment. The technique is similar to other interventions for structural heart disease, such as closing interatrial communications (patent foramen ovale or atrial septal defects) with self-expandable umbrella-like devices.
During the initial study, percutaneous LAA occlusion was achieved in all but 3 patients. The occurrence of pericardial effusion and hemopericardium were major periprocedural complications. Nevertheless, there were no sequelae in most patients. The relative reduction of stroke achieved by the PLAATO system has been 65% so far, assuming our patients would have been placed on aspirin-only therapy instead of undergoing LAA occlusion.15
Percutaneous occlusion with the PLAATO device was shown in 2 studies to be feasible with an acceptable risk profile. This procedure may be a desirable option for patients at high risk for ischemic stroke but who are unable to receive long-term anticoagulant treatment. Further studies are needed to evaluate the performance of this new technique, not only with regard to long-term tolerance of the implanted device but, more importantly, whether or not the PLAATO system is comparable to anticoagulation therapy for stroke prevention.