The elephant trunk technique: operative results in 100 consecutive patients

Objectives: To describe morbidity and mortality in patients undergoing the elephant trunk (ET) implantation as the first stage in the repair of their mega aorta and to assess determinants for the occurrence of complications. Methods: One hundred consecutive patients undergoing an ET implantation between 1984 and June 2001 were retrospectively analyzed. The ET was implanted as an extension of an isolated aortic arch (1%), an aortic valve replacement + ascending aorta + arch (14%), a root replacement + ascending aorta + arch (37%) and an ascending aorta + arch (48%). Indications for surgery were acute aortic dissection (1%), an inflammatory aneurysm (3%), chronic post-dissection (31%) or degenerative (65%) aneurysm. Marfan syndrome was present in six patients. For cerebral protection, we used isolated deep hypothermic circulatory arrest (7%), deep hypothermic circulatory arrest combined with uni- or bilateral antegrade cerebral perfusion (18%) or isolated uni- or bilateral antegrade cerebral perfusion (75%). Uni- and multivariate analysis was used. Results: There were no intraoperative deaths. Hospital mortality was 8%. The causes of death were cardiac in one, rupture of a remote aneurysm in three, tamponade in one and sepsis in three. After multivariate analysis, no single factor emerged as a risk factor for hospital mortality. Permanent and transient neurologic dysfunction occurred in 4 and 2%, respectively. Univariate analysis showed the operative period before 1990 (P = 0.029) and emergency (P = 0.018) as significant factors for postoperative neurologic dysfunction; after stepwise logistic regression analysis, only emergent operation retained significance (P = 0.005). Permanent hoarseness, total atrioventricular block requiring pacemaker implantation and rethoracotomy for bleeding occurred in 17, 2 and 30%, respectively. Conclusions: The first step in the repair of a mega aorta, the implantation of an ET, can be performed with a low mortality and an acceptable morbidity. The risk of central neurologic damage is higher in emergency interventions. (C) 2002 Elsevier Science B.V. All rights reserved.