Somatosensory evoked potential monitoring of lumbar pedicle screw placement for in situ posterior spinal fusion

Background context: Somatosensory evoked potentials (SSEP) are commonly used to monitor the spinal cord and nerve roots during operative procedures that put those structures at risk. The utility of SSEPs to evaluate cauda equina and nerve root function during posterior spinal arthrodesis with pedicular fixation for degenerative lumbar disease has been reported anecdotally and remains controversial. Purpose: An institution-wide review of the ability of SSEP readings to monitor nerve function during posterior lumbar spinal arthrodeses with transpedicular fixation for degenerative lumbar spinal disorders was undertaken. Study design/setting: A retrospective review was undertaken. Patient history, preoperative physical examination, intraoperative anesthesia, SSEP records and the postoperative course were reviewed. Methods: A total of 186 consecutive arthrodeses as described above were reviewed. Patients who had anterior procedures, spondyloreduction or scoliosis correction were excluded from the study. There were 76 male and 110 female patients. Five fellowship-trained spine surgeons placed a total of 888 pedicle screws. Sixty-five percent of the patients had a principal preoperative diagnosis of spinal stenosis with degenerative spondylolisthesis. Other common diagnoses were isthmic spondylolisthesis and degenerative scoliosis. Ninety-three percent of the cases involved decompressive laminectomy. Eight percent had posterior interbody fusions. All pedicle screws were placed without the assistance of fluoroscopy or stereotactic computer-assisted guidance. Screw position was evaluated intraoperatively with standard posteroanterior and lateral radiographs.Anesthetic agents compatible with SSEP monitoring were used in all patients. SSEP baseline readings were obtained in all patients in the operating room soon after induction of general anesthesia. An acute and sustained loss of 50% of the SSEP amplitude and/or increase by 10% of latency from baseline was considered to be pathologic. Results: None of the 186 patients had significant SSEP changes. There were, however, 5 patients with postoperative radiculopathies distinct from their preoperative presentations. Early postoperative plain radiographs and computed assisted tomography (CAT) scans revealed malpositioned pedicle screws. Consequently, eight pedicle screws were either revised or removed. All patients had partial or full recovery of their new deficits after revision surgery. Conclusion: We conclude that the use of SSEPs in evaluating pedicle screw placement during lumbar arthrodesis is limited. In this setting, if monitoring is required, alternative methods with greater sensitivity and efficacy should be explored.