A randomized, prospective pilot comparison of 3 atrial appendage elimination techniques: Internal ligation, stapled excision, and surgical excision

Background: Elimination of the left atrial appendage (LAA) attempts to reduce stroke in patients with atrial fibrillation (AF). A retrospective review suggests that various surgical techniques are often unsuccessful and may leave a stump or gap. In a pilot study, we prospectively evaluated 3 surgical techniques with long-term follow up to define effectiveness. Methods: At a single institution, 28 patients undergoing concomitant AF surgery were randomized prospectively into 1 of 3 techniques of LAA elimination: internal suture ligation (IL), external stapled excision (StEx), and surgical excision (SxEx). The success of LAA elimination was assessed by transesophageal echocardiography (TEE) in all patients at the time of surgery. Failure of LAA closure consisted of either a stump (residual appendage tissue > 1 cm in maximum length) or a gap (persistent flow between the left atrium [LA] and the LAA). Failure was treated intraoperatively when recognized. Late follow-up was obtained using a TEE at a mean of 0.4 years in 21/28 (75%) of patients. Results: Early failure was recognized and treated in 1 patient in the IL group (13%), 6 patients in the StEx group (60%), and 2 patients in the SxEx group (20%) (P = .06). On follow-up TEE, 4 of 7 patients in the IL group (57%) had developed gaps, 3 of whom (43%) with greater than mild flow. No patients in the StEx or SxEx groups had a gap (P = .03). In late follow-up, 1 of 7 patients in the IL group (14%) had a stump, compared with 2 of 8 (25%) in the StEx group and 3 of 6 (50%) in the SxEx group (P = .35). The overall failure rate was 57%: 5 of 8 (63%) in the IL group, 6 of 10 (60%) in the StEx group, and 5 of 10 (50%) in the SxEx group (P = .85). No patient had a stroke at any time during follow-up. Conclusions: LAA elimination is often incomplete and goes undetected. If the LAA is eliminated at the time of surgery, then TEE should be used intraoperatively to assess effectiveness and reintervention performed if warranted. Late assessment for completeness of closure should be considered before cessation of anticoagulation until more effective LAA techniques can be developed.