Perioperative outcomes and survival after surgery for intramedullary spinal cord tumors: a single-institution series of 302 patients

Abstract

OBJECTIVE Intramedullary spinal cord tumors (IMSCTs) are rare neoplasms whose treatment is often technically challenging. Given the low volume seen at most centers, perioperative outcomes have been reported infrequently. Here, the authors present the largest single-institution series of IMSCTs, focusing on the clinical presentation, histological makeup, perioperative outcomes, and long-term survival of surgically treated patients. METHODS A cohort of patients operated on for primary IMSCTs at a comprehensive cancer center between June 2002 and May 2020 was retrospectively identified. Data on patient demographics, tumor histology, neuraxial location, baseline neurological status, functional deficits, and operative characteristics were collected. Perioperative outcomes of interest included length of stay, postoperative complications, readmission, reoperation, and discharge disposition. Data were compared across tumor histologies using the Kruskal-Wallis H test, chi-square test, and Fisher exact test. Pairwise comparisons were conducted using Tukey’s honest significant difference test, chi-square test, and Fisher exact test. Long-term survival was assessed across tumor categories and histological subtype using the log-rank test. RESULTS Three hundred two patients were included in the study (mean age 34.9 ± 19 years, 77% white, 57% male). The most common tumors were ependymomas (47%), astrocytomas (31%), and hemangioblastomas (11%). Ependymomas and hemangioblastomas disproportionately localized to the cervical cord (54% and 59%, respectively), whereas astrocytomas were distributed almost equally between the cervical cord (36%) and thoracic cord (38%). Clinical presentation, extent of functional dependence, and postoperative 30-day outcomes were largely independent of underlying tumor pathology, although tumors of the thoracic cord had worse American Spinal Injury Association (ASIA) grades than cervical tumors. Rates of gross-total resection were lower for astrocytomas than for ependymomas (54% vs 84%, p < 0.01) and hemangioblastomas (54% vs 100%, p < 0.01). Additionally, 30-day readmission rates were significantly higher for astrocytomas than ependymomas (14% vs 6%, p = 0.02). Overall survival was significantly affected by the underlying pathology, with astrocytomas having poorer associated prognoses (40% at 15 years) than ependymomas (81%) and hemangioblastomas (66%; p < 0.01) and patients with high-grade ependymomas and astrocytomas having poorer long-term survival than those with low-grade lesions (p < 0.01). CONCLUSIONS The neuraxial location of IMSCTs, extent of resection, and postoperative survival differed significantly across tumor pathologies. However, perioperative outcomes did not vary significantly across tumor cohorts, suggesting that operative details, rather than pathology, may have a stronger influence on the short-term clinical course, whereas pathology appears to have a stronger impact on long-term survival.

Surgical approaches to intramedullary spinal cord astrocytomas in the age of genomics

“…Consequently, surgeons must balance the potential improvement in survival from more extensive resections with the increased risk of worsening and/or new neurological deficits ( 27 ). Reported rates of GTR are lower compared to intramedullary ependymomas, which often present with distinct cleavage planes ( 17 , 28 , 29 ). Patients with low-grade astrocytomas, particularly pilocytic astrocytomas, may present with dissection planes that facilitate gross total resection; however, GTR can be difficult to achieve in high-grade astrocytomas ( 17 , 28 , 30 ).…”

“…Reported rates of GTR are lower compared to intramedullary ependymomas, which often present with distinct cleavage planes ( 17 , 28 , 29 ). Patients with low-grade astrocytomas, particularly pilocytic astrocytomas, may present with dissection planes that facilitate gross total resection; however, GTR can be difficult to achieve in high-grade astrocytomas ( 17 , 28 , 30 ). Using intra-operative findings and neuromonitoring, and considering the patient’s pre-operative baseline deficits, surgeons must determine the maximal safe resection feasible.…”

“…Studies have suggested improved functional outcomes for patients with intra-operative monitoring compared to those without monitoring ( 37 , 42 ). Institutional studies have reported using somatosensory evoked potentials, motor evoked potentials, and transcranial motor evoked potential monitoring ( 17 , 39 ). Loss of evoked potentials without recovery is often considered an endpoint for resection due to its ability to predict new post-operative deficits, although protocols differ ( 17 , 32 , 43 ).…”

“…Intramedullary tumors can present with nonspecific symptoms, such as axial back pain and radiculopathy, complicating diagnosis ( 10 ). Growth of the neoplasm results in progressive deformation of the spinal cord, contributing to numbness and sensory deficits, weakness and motor deficits, and bowel and bladder dysfunction ( 17 , 18 ). Astrocytomas typically present hypo-or isointense on T1-weighted and hyperintense on T2-weighted magnetic resonance imaging (MRI) and can include syrinx formation ( 18 ) ( Figure 1 ).…”

“…Survival for patients with intramedullary astrocytomas is poorer compared to ependymomas and hemangioblastomas, with reported overall rates ranging from 40%–65% over 15 years ( 17 ). These poorer outcomes reflect the infiltrative nature of astrocytomas, resulting in low rates of gross total resection (GTR) or near total resection.…”