Impact of extent of resection and adjuvant therapy in diffuse gliomas of the spine.

Abstract

Diffuse gliomas of the spine (DGS)-consisting of intradural intramedullary glioblastoma, astrocytoma, and oligodendroglioma-are exceedingly rare tumors that account for about 2% of primary spinal cord tumors. Much is unknown about their optimal treatment regimen due to a relative lack of clinical outcome data. To provide an updated analysis on treatment and outcomes in DGS. Observational cohort study using The National Cancer Database (NCDB), a multicenter prospectively collected oncology outcomes database. A systematic literature review was also performed to compare the resulting data to previous series. Patients with histologically confirmed DGS from 2004 to 2018. Long-term overall survival and short-term 30/90-day postsurgical mortality, 30-day readmission, and prolonged hospital length of stay. Impact of extent of resection and adjuvant therapy on overall survival was evaluated using Kaplan-Meier estimates and multivariable Cox proportional hazards regression. Univariate and multivariate logistic regression was used to analyze covariables and their prognostic impact on short-term surgical outcomes. Of the 747 cases that met inclusion criteria, there were 439 astrocytomas, 14 oligodendrogliomas, and 208 glioblastomas. Sixty percent (n=442) of patients received radiation, and 45% (n=324) received chemotherapy. Tumor histology significantly impacted survival; glioblastoma had the poorest survival (median survival time [MS]: 12.3 months), followed by astrocytoma (MS: 70.8 months) and oligodendroglioma (MS: 71.6 months) (p<.001). Gross total resection (GTR) independently conferred a survival benefit in patients with glioblastoma (hazard ratio [HR]: 0.194, p<0.001) and other WHO grade four tumors (HR: 0.223, p=.003). Adjuvant chemotherapy also improved survival in patients with glioblastoma (HR: 0.244, p=.007) and WHO grade four tumors (HR: 0.252, p<.001). Systematic literature review identified 14 prior studies with a combined DGS mortality rate of 1.3%, which is lower than the 4% real-world outcomes calculated from the NCDB. This difference may be explained by selection biases in previously published literature in which only centers with favorable outcomes publish their results. There remains a paucity of data regarding treatment paradigms and outcomes for DGS. Our analysis, the largest to date, demonstrates that GTR and adjuvant therapy independently improve survival for certain high-grade subgroups of DGS. This best-available data informs optimal management for such patients.