Morphological risk factors of stroke during thoracic endovascular aortic repair

This study aims to identify independent factors correlating to an increased risk of perioperative stroke during thoracic endovascular aortic repair (TEVAR). A prospective maintained TEVAR database, medical records, and imaging studies of 300 patients (205 men; median age of all, 66 years, range 21-89), who underwent TEVAR between March 1997 and February 2011, were reviewed. Preoperative CT data sets were reviewed by two experienced radiologists with focus on the atheroma burden in the aortic arch (grade I, normal, to grade V, ulcerated or pedunculated atheroma). Aortic arch geometry (arch types I-III) was documented. Further parameters included in the univariate analysis were age, gender, urgency of repair, duration of procedure, adenosine-induced cardiac arrest or rapid pacing, proximal landing zone, left subclavian artery (LSA) coverage, and number of stent grafts. Multivariate logistic regression analysis was performed to assess the independent correlations of potential risk factors. Atherosclerotic aneurysm was the most common pathology (44 %). One hundred and fifty-four of our patients (51 %) were treated under urgent or emergent conditions. Seventeen percent of all patients had significant arch atheroma (grade IV or V), and 43 % had a steep type III aortic arch. The perioperative stroke was 4 % (12 patients; median age, 73 years, range 31-78). Two strokes were lethal (0.7 %). All strokes were classified as embolic based on imaging characteristics. In eight patients, strokes were located in the left cerebral hemisphere (seven of them in the anterior and one in the posterior circulation). Four stroke patients (one in the left posterior circulation) underwent LSA coverage without revascularization. Three stroke patients had severe arch atheroma grade V. Five patients suffering stroke were recognized to have a type III aortic arch. Strokes were equally distributed between zones 0-2 vs. 3-4 (n = 6 each, 5 vs. 3.3 %). The highest incidence was found in zone 1 (11.4 %). In univariate analysis, grade V arch atheroma (odds ratios (OR), 5.35; 95 % confidence intervals (CI), 1.00-25.87; P = 0.035) and zone 1 deployment (OR, 5.03; 95 % CI, 1.19-20.03; P = 0.021) were significantly associated with perioperative stroke. In multivariate analysis, both parameters were confirmed as independent significant risk factors for stroke during TEVAR. Stroke risk during TEVAR is directly associated with the atheroma burden of the aortic arch and the proximal landing zone. These factors should be considered during patient selection, planning, and implantation strategies of TEVAR.