Protection of the spinal cord during surgery of thoraco-abdominal aortic aneurysms.

OBJECTIVE: To assess the risk of ischemic cord injury, we have retrospectively studied the 115 patients who underwent a replacement of the thoracic descending or thoraco-abdominal aorta between January 1980 and December 1994. METHODS: In 72 patients the aortic lesion was located above the diaphragm. The aortic replacement was performed with the aid of extracorporeal circulation in all but 2 patients (97.2%). Only two cases of postoperative paraplegia were observed (2.7%). In 43 patients (10 females and 33 males aged from 26 to 69 years), the occurrence of postoperative paraplegia was considered as a major risk, because of the extension of the aortic lesions (Crawford types I, II and III). Twenty-six patients (60.4%) suffered from chronic dissection and 17 patients had atheromatous aneurysms. Sixteen patients (37.2%) had Marfan syndrome. Twelve patients (27.9%) had already undergone aortic replacement. A preoperative study of the spinal cord vascularization was carried out in 36 patients (83.6%) and the Adamkiewicz artery was visualized in 28 patients (77.8%). In 17 patients (39.5%, group I), the surgical procedure was performed without the aid of extracorporeal circulation. In the remaining 26 patients (60.5%, group II), the surgical procedure was carried out with the aid of cardiopulmonary bypass and profound hypothermic circulatory arrest. Sequential unclamping of the aorta was used in all patients. The cord vascularization was surgically restored in 32 patients (74.4%). When the Adamkiewicz artery was identified, the critical intercostal artery was reimplanted together with the two pairs of adjacent intercostal arteries (25 patients). When the origin of the Adamkiewicz artery remained unknown, the two or three most important patent pairs of intercostal arteries were reimplanted (7 patients). In 8 patients (18.6%) there were no patent intercostal arteries. RESULTS: Hospital mortality accounted for 37.2% (16 patients, including 5 patients with paraplegia). On univariate analysis, extension of the aortic lesions, emergency and redo surgery were the only significant risk factors of mortality (P = 0.05). Cord ischemia was observed in 9 patients (21%): permanent paraplegia in 7 patients (16.2%) and transient medullar disturbance in 2 patients (4.6%). The occurrence of paraplegia was reduced, though not significantly, in group II (16%) vs group I (29%) and in patients with preoperative assessment of the cord vascularization (18% vs 38%). CONCLUSIONS: In our experience: 1) The risk of paraplegia is related to the extension and the type of the aortic lesions. 2) The preoperative study of the medullar vascularization and the use of extracorporeal circulation with deep hypothermia and sequential aortic unclamping, reduce the risk of severe cord ischemia, and 3) Occurrence of postoperative paraplegia depends on several factors and cannot be totally prevented by the surgical technique.