Prognostic implications of resection extent for patients with glioblastoma multiforme: a meta-analysis.

INTRODUCTION Surgery is the primary treatment of glioblastoma multiforme (GBM), and a greater extent of resection (EOR) has been shown to be associated with improved survival. Our objective was to perform a meta-analysis comparing the 1-year overall survival (OS) and 1-year progression-free survival (PFS) of GBM patients who receive total resection, incomplete resection, or biopsy only. EVIDENCE ACQUISITION PubMed and the Cochrane databases were searched until May 19th, 2015 using the terms "glioblastoma/glioblastoma multiforme," "extent of resection," "surgery prognosis/prognostic," "survival rate." Randomized controlled trials (RCTs), two-arm prospective studies, retrospective studies, and cohort studies reporting OS and/or PFS data were included. One-year OS and 1-year PFS were compared. EVIDENCE SYNTHESIS Three prospective/RCTs, and 3 retrospective studies were included. The 6 studies included 1618 patients: 523 underwent total resections, 857 underwent incomplete resections, and 238 had biopsies. Total resection was associated with greater 1-year OS than incomplete resection (pooled odds ratio [OR] 1.89, 95% confidence interval [CI]: 1.35-2.64, P<0.001), and greater 1-year PFS than incomplete resection (pooled OR 2.11, 95% CI: 1.44-3.09, P<0.001). Analysis by study type (RCT or retrospective) produced similar results, although only one RCT provided 1-year PFS data and there was no significant difference between total resection and incomplete resection in that study. All analyses showed that total resection was associated with greater survival than biopsy only. CONCLUSIONS Total resection of GBM is associated with improved OS and PFS as compared to incomplete resection or biopsy.