Gross total but not incomplete resection of glioblastoma prolongs survival in the era of radiochemotherapy.

ABSTRACT

BACKGROUND: This prospective multicenter study assessed the prognostic influence of the extent of resection when compared with biopsy only in a contemporary patient population with newly diagnosed glioblastoma. PATIENTS AND METHODS: Histology, O(6)-methylguanine-DNA methyltransferase (MGMT) promoter methylation status, and clinical data were centrally analyzed. Survival analyses were carried out with the Kaplan-Meier method. Prognostic factors were assessed with proportional hazard models. RESULTS: Of 345 patients, 273 underwent open tumor resection and 72 biopsies; 125 patients had gross total resections (GTRs) and 148, incomplete resections. Surgery-related morbidity was lower after biopsy (1.4% versus 12.1%, P = 0.007). 64.3% of patients received radiotherapy and chemotherapy (RT plus CT), 20.0% RT alone, 4.3% CT alone, and 11.3% best supportive care as an initial treatment. Patients ≤60 years with a Karnofsky performance score (KPS) of ≥90 were more likely to receive RT plus CT (P < 0.01). Median overall survival (OS) (progression free survival; PFS) ranged from 33.2 months (15 months) for patients with MGMT-methylated tumors after GTR and RT plus CT to 3.0 months (2.4 months) for biopsied patients receiving supportive care only. Favorable prognostic factors in multivariate analyses for OS were age ≤60 years [hazard ratio (HR) = 0.52; P < 0.001], preoperative KPS of ≥80 (HR = 0.55; P < 0.001), GTR (HR = 0.60; P = 0.003), MGMT promoter methylation (HR = 0.44; P < 0.001), and RT plus CT (HR = 0.18, P < 0.001); patients undergoing incomplete resection did not better than those receiving biopsy only (HR = 0.85; P = 0.31). CONCLUSIONS: The value of incomplete resection remains questionable. If GTR cannot be safely achieved, biopsy only might be used as an alternative surgical strategy.