Late Neurologic Events After Surgery for Atrial Fibrillation: Rare but Relevant

Abstract

The classic cut and sew maze is thought to reduce stroke, in part because of left atrial appendage (LAA) elimination. Multiple LAA elimination techniques have evolved with the introduction of new surgical treatment options for atrial fibrillation (AF), but the impact on stroke remains unknown. We studied the rate of late neurologic event (LNE) in the era of contemporary AF surgery. From April 21, 2004, to June 30, 2011, 773 patients underwent surgery for AF. In 131 patients, the LAA was excised. In 579, alternative elimination techniques were used (97 external ligation, 313 internal ligation, 126 stapled excision, 23 stapled excision plus internal ligation, 5 internal plus external ligations, and 15 that did not fit into any category); 63 LAAs were left intact and excluded from analyses. Complete follow-up was obtained by medical record review and phone call. Median survival follow-up was 3.3 years (first and third quartiles, 1.6 and 5.0). An LNE was defined as either a documented stroke or transient ischemic attack 30 or more days after surgery. Baseline characteristics and outcomes between LAA techniques were compared using χ(2), Fisher's exact tests, and Student's t tests. There were 25 LNEs (3.5%) overall; the median occurrence time was 3.6 years (first and third quartiles, 1.9 and 5.4) after surgery. There were 17 strokes and 8 transient ischemic attacks. Of 45 demographic and surgical variables, only age, aortic valve surgery, and perioperative neurologic event (<30 days after cardiac surgery) independently predicted LNE (p = 0.003, 0.021, and 0.010, respectively). Late neurologic events occurred with an annual rate of 1.13% in patients with alternative elimination techniques, and 0.20% in patients with excised LAA (p = 0.001). Patients in AF at any time were more likely to have LNE, but this was not an independent predictor. After surgery for AF ablation, there is ongoing low risk of LNE even when the LAA is surgically excised. Further investigation should be pursued to clarify whether a difference exists with alternative elimination techniques and in patients in whom AF is successfully eliminated.