Outcomes of a Spinal Drain and Intraoperative Neurophysiologic Monitoring Protocol in Thoracic Endovascular Aortic Repair

Abstract

Adjuncts for early detection and treatment of spinal cord ischemia (SCI) in thoracic aortic surgery are supported by robust clinical experience in open repair. The utility of cerebrospinal fluid (CSF) drainage and neurophysiologic monitoring (NPM) in thoracic endovascular aortic repair (TEVAR) is less clear. The purpose of this investigation is to determine the influence of a selective institutional spinal cord protection protocol using prophylactic NPM and CSF on outcomes for standard TEVAR. Patients undergoing standard TEVAR entered into a prospectively maintained database from a single institution from 2007 to 2016 were retrospectively reviewed. Preoperative characteristics, aneurysm extent, and etiology were reviewed. Utilization of CSF drains including volume of fluid removed, duration of drainage, and catheter-related complications were collected. NPM data were reviewed to determine the influence on intraoperative management. Exact logistic regression was used to identify independent predictors of SCI. Of 223 patients undergoing TEVAR, 130 met inclusion criteria for the study. CSF drains were used in 71 patients (54.6%), and 56 of 130 (43%) had NPM. SCI occurred in 7 patients (5.4%), of whom 5 had partial or complete recovery. Median time to symptoms of SCI was delayed in all cases (median 52hr, range 8-312), and none of the 4 of 7 patients with adjunct NPM demonstrated intraoperative changes. Intraoperative changes in NPM occurred in 26 (46%), and represented unilateral leg ischemia in all but 2 cases. In both patients, changes consistent with SCI were associated with intraoperative hypotension and resolved with blood pressure augmentation. Neither patient developed postoperative SCI. Median length of stay (22 vs. 9days, P=0.012), operative room time (262 vs. 209, P=0.040), and perioperative mortality (28.6% vs. 4.1%, P=0.046) were significantly higher for patients with SCI versus those without. Length of aortic coverage was found to be the sole independent predictor of SCI (odds ratio 8.2, P=0.026). Complications related to CSF drainage occurred in 4 patients (5.6%) with major complications occurring in 2 patients (2.8%), including 1 with an intrathecal hematoma and permanent bilateral paraparesis. Selective use of prophylactic CSF drainage in TEVAR was associated with moderate risk and questionable benefit. The use of neurophysiological monitoring allowed for early detection and treatment of spinal ischemia, but its utility is limited by logistical factors and to the minority of patients with intraoperative spinal ischemic events.