Prognostic Implications of Extent of Resection in Glioblastoma: Analysis from a Large Database.

Abstract

To analyze the predictors for and clinical impact of gross total resection (GTR) in patients with glioblastoma (GBM). The National Cancer Database was queried for patients with GBM diagnosed from 2004 to 2013 with known survival and extent of resection. Patients were grouped based on extent of resection into biopsy alone, subtotal resection (STR), and GTR. Univariable analyses and multivariable analyses (MVAs) were performed to investigate factors associated with the likelihood of GTR and overall survival (OS) after diagnosis. A total of 27,865 patients met inclusion criteria. Factors associated with increased odds of GTR on MVA included later year of diagnosis, younger age, higher performance status, nonright-sided tumors, multifocal tumors, and O6-methylguanine-methyltransferase gene promoter non-hypermethylated tumors (each P < 0.020). Factors associated with improved OS on MVA included younger patient age, female sex, race, lower comorbidity score, higher performance score, smaller tumor size, unifocality, O6-methylguanine-methyltransferase hypermethylation, radiotherapy, chemotherapy, and facility volume (each P < 0.005). After we adjusted for each of these factors, compared with biopsy alone, GTR was associated with improved OS (hazard ratio 0.768, P<0.001), whereas STR (grouped together) was not (hazard ratio 0.995, P= 0.930). Although a prospective randomized trial on this topic is unlikely to be completed, this large retrospective analysis provides evidence to support the recommendation of GTR in patients with GBM. This study does not support a survival benefit of STR over biopsy alone (when grouped into these 2 groups), although there may be a subset of patients with near total resection who would benefit.