Utility of neurophysiological monitoring using dorsal column mapping in intramedullary spinal cord surgery

Abstract

Intramedullary spinal cord tumors can displace the surrounding neural tissue, causing enlargement and distortion of the normal cord anatomy. Resection requires a midline myelotomy to avoid injury to the posterior columns. Locating the midline for myelotomy is often difficult because of the distorted anatomy. Standard anatomical landmarks may be misleading in patients with intramedullary spinal cord tumors due to cord rotation, edema, neovascularization, or local scar formation. Misplacement of the myelotomy places the posterior columns at risk of significant postoperative disability. The authors describe a technique for mapping the dorsal column to accurately locate the midline. A group of 10 patients with cervical and thoracic intramedullary spinal cord lesions underwent dorsal column mapping in which a strip electrode was used to define the midline. After the laminectomy and durotomy, a custom-designed multielectrode grid was placed on the exposed dorsal surface of the spinal cord. The electrode is made up of 8 parallel Teflon-coated stainless-steel wires (76-microm diameter, spaced 1 mm apart) embedded in silastic with each of the wires stripped of its insulating coating along a length of 2 mm. This strip electrode maps the amplitude gradient of conducted spinal somatosensory evoked potentials elicited by bilateral tibial nerve stimulation. Using these recordings, the dorsal columns are topographically mapped as lying between two adjacent numbers. The authors conducted a retrospective analysis of the preoperative, immediate, and short-term postoperative neurological status, focusing especially on posterior column function. There were 8 women and 2 men whose mean age was 52 years. There were 4 ependymomas, 1 subependymoma, 1 gangliocytoma, 1 anaplastic astrocytoma, 1 cavernous malformation, and 2 symptomatic syringes requiring shunting. In all patients the authors attempted to identify the midline by using anatomical landmarks, and then proceeded with dorsal column mapping to identify the midline electrophysiologically. In the 2 patients with syringomyelia and in 5 of the patients with tumors, the authors were unable to identify the midline anatomically with any certainty. In 2 patients with intramedullary tumors, they were able to identify the midline anatomically with certainty. Dorsal column mapping allowed identification of the midline and to confirm the authors' anatomical localization. In 2 patients with intramedullary tumors, posterior column function was preserved only on 1 side. All other patients had intact posterior column function preoperatively. Dorsal column mapping is a useful technique for guiding the surgeon in locating the midline for myelotomy in intramedullary spinal cord surgery. In conjunction with somatosensory evoked potential, motor evoked potential, and D-wave recordings, we have been able to reduce the surgical morbidity related to dorsal column dysfunction in this small group of patients.