Usefulness and related problems of somatosensory evoked potential monitoring for prevention of spinal cord injury secondary to operation of the aorta / 日本心臓血管外科学会雑誌

ABSTRACT

Spinal cord injury is a dreaded and serious complication of operative procedures on the descending aorta. To avoid this serious complication, 53 patients underwent somatosensory evoked potential (SEP) monitoring during operations on the aorta which required cross-clamping of the descending aorta. 38 patients whose SEPs were kept normal during and after operations did not develop spinal cord injury. Among the 14 patients who developed both abnormal decrease in amplitude and elongation of peak latency, 13 lost their SEPs during aortic cross-clamping. Peripheral nerve ischemia seemed to be the cause of those abnormalities in 8 to whom cross-clamping was given to the abdominal aorta. Inadequate perfusion of the distal aorta was suspected in 6 to whom cross-clamping was given to the descending thoracic aorta. In these cases, however, SEP monitoring was not specific in differentiating spinal cord ischemia from peripheral nerve ischemia. Spinal cord injury was noted in only one of the 6 patients. The remaining one patient developed complete loss of SEP and spinal cord injury on the first postoperative day despite the well preserved SEP during the operation. Since this patient underwent flow reversal and thromboexclusion method for the dissecting aneurysm, gradual thrombotic occlusion of the important radicular arteries draining to spinal cord might have resulted delayed appearance of the spinal cord injury. In conclusion, SEP monitoring is the reliable method to detect the spinal cord ischemia which might be developed during cross-clamping of the descending aorta. However, this method bears limitation in its clinical application due to the following reasons. First, intraoperative SEP monitoring cannot predict delayed occurence of spinal cord injury. Secondly, this method cannot detect the qualitative extent of ischemia of spinal cord and the safe range of the cross-clamp time.