Association of Surgical Resection, Disability, and Survival in Patients with Glioblastoma

Abstract

Objective Extent of resection (EOR) and Karnofsky Performance Status (KPS) are at odds in glioblastoma (GBM) surgery, that is, the anticipated postoperative disability limits the EOR. This study analyzes the correlation of different surgical modalities with the resulting physical status and survival of patients with GBM. Methods A total of 565 patients with primary GBM were operated on in a single institution between 2006 and 2014. Possible surgical modalities comprised supratotal resection (SLR), gross total resection (GTR; ≥ 95% by volume), tumor debulking (TDB; ≤ 95% by volume), and stereotactic biopsy (SB). Pre- and postoperative KPS before and up to 4 weeks after surgery as well as overall survival (OS) rate were determined retrospectively. Hazard ratio (HR) and 95% confidence intervals were calculated using a Cox proportional hazards model. Results Median postoperative KPS was ≥ 70, irrespective of surgical modality. Mean OS was 12.5 months. Multivariate analysis revealed age ≥ 70 years (HR: 1.93), preoperative KPS < 70 (HR: 2.15), and unmethylation in MGMT promoter (HR: 1.27) as independent factors for worse OS. Regarding surgical modality, SB was associated with the worst survival (HR: 2.3) followed by TDB (HR: 1.36). SLR was inferior to GTR (HR: 1.27). Conclusion Higher EOR in patients with GBM does not seem inevitably correlated with increasing functional impairment, but better survival, provided there is a balanced preoperative indication. Nevertheless, SLR does not seem to be superior to GTR. Whenever possible, maximal safe resection should be considered in patients with GBM, even if an EOR ≥ 95% is not possible.