Results of Spinal Fusion After Spinal Nerve Sheath Tumor Resection

Abstract

Intradural extramedullary spine tumors, approximately one-half of which are peripheral nerve sheath tumors (PNSTs), comprise two-thirds of primary spinal neoplasms. Given the rarity of PNSTs and the restricted indications for adding fusion to laminectomy for tumor resection, analyses of spinal fusion outcomes are limited. Demographics, clinical presentation, tumor characteristics, extent of resection, spinal fusion, complications, and clinical follow-up were recorded retrospectively. A total of 221 tumors in 199 patients were identified (53 neurofibromas, 163 schwannomas, 5 malignant PNSTs); 78 patients underwent fusion (70 instrumented; 8 noninstrumented). Fusion rates were higher for extradural versus intradural lesions (60% vs. 29%; P= 0.001) and for tumors involving the cervicothoracic junction (88% vs. 31%, P < 0.001). There was no difference in fusion rates based on pathology. Rates of new or worsening sensory (19% in fusion vs. 13% in nonfused) or motor deficits (8% in fused vs. 4% in nonfused), wound infection (3% in fused vs. 6% in nonfused) and cerebrospinal fluid (CSF) leak or pseudomeningocele (6% in fused vs. 4% in nonfused) were not statistically different. There were 10 fusion-related complications: 6 adjacent segment disease, 3 implant failures, and 1 pseudoarthrosis. Mean time from surgery to last follow-up was 32 months. In this cohort, PNSTs in the cervical spine, spanning the cervicothoracic junction, and extradural tumors were associated with higher rates of spinal fusion. Fusion was not associated with new or worsening motor/sensory deficits, CSF leak, pseudomeningocele, wound infection, or spinal deformity. Overall, spinal fusions were well tolerated and did not increase the risk of postoperative complications.