Intracranial Gliosarcoma: A National Cancer Database Survey of Clinical Predictors for Overall Survival

• Published in: World neurosurgery 2023-09, Vol.177, p.e621-e629
• Main Authors: Bachu, Vismaya S., Alem, Danayt, Jimenez, Miguel, Lehner, Kurt, Porras, Jose Luis, Mukherjee, Debraj
• Format: Article
• Language: English
• Subjects: Gliosarcoma; National Cancer Database; Predictors; Survival; Treatment
• ISSN: 1878-8750; 1878-8769
• DOI: 10.1016/j.wneu.2023.06.111

Gliosarcoma (GSM) is a variant of glioblastoma, 1 of the most common and aggressive primary brain tumors in adults. Our study seeks to analyze a large cohort of patients with GSM in the National Cancer Database (NCDB) to elucidate clinical predictors of overall survival (OS). Data was collected on patients diagnosed with histologically-confirmed GSM using the NCDB (2004–2016). OS was determined via univariate Kaplan-Meier analysis. Bivariate and multivariate Cox proportional-hazards analyses were also utilized. Our cohort of 1015 patients had a median age at diagnosis of 61 years. Six hundred thirty-one (62.2%) were male, 896 (89.0%) were Caucasian, and 698 (68.8%) lacked any comorbidities. Median OS was 11.5 months. Regarding treatment, 264 (26.5%) patients underwent surgery (S) only (OS = 5.19 months), 61 (6.1%) underwent surgery and radiotherapy (S + RT) (OS = 6.87 months), 20 (2.0%) underwent surgery and chemotherapy (S + CT); (OS = 15.51 months), and 653 (65.4%) underwent S + CT + RT (triple) combination therapy (OS = 13.8 months). Notably, on bivariate analysis, S + CT (Hazard ratio [HR] = 0.59, P-value = 0.04) and triple therapy (HR = 0.57, P < 0.01) were associated with increased OS. S + RT was not significantly associated with OS. Similarly, on multivariate Cox proportional-hazards analyses, gross total resection (HR = 0.76, P = 0.02), S + CT (HR = 0.46, P < 0.01), and triple therapy (HR = 0.52, P < 0.01) predicted significantly increased OS. Furthermore, age >60 years old (HR = 1.03, P < 0.01) and the presence of comorbidities (HR = 1.43, P < 0.01) predicted significantly decreased OS. Despite maximal multimodal treatment, GSMs have poor median OS. NCDB data suggest age, comorbidities, extent of resection, and adjuvant treatment each minimally delays poor outcomes.