Repeat-surgery at Glioblastoma recurrence, when and why to operate?

Objective: Glioblastoma (GB) recurrence is inevitable; guidelines for treatment at disease recurrence are deficient. Clinicians are faced with deciding whom to choose for repeat-surgery. This study analyzes recurrence therapy modalities, investigates characteristics of patients operated on at recurrence and evaluates outcome benefit. Methods: Consecutive adult patients operated on for de novo GB at a single institution from 2006 to 2011 were reviewed. Clinical, radiographic and molecular data of 141 patients diagnosed of recurrent disease were assessed. Reasons for recurrence therapy and therapy modalities were reviewed. Univariate analysis was used to analyze differences in parameters of patients operated on at recurrence and those not. Impact of re-operation on survival was evaluated by the Kaplan-Meier method and Log-rank test. Results: 53 (38%) patients were selected for repeat surgery upon recurrent disease, this was followed by either chemotherapy (CT) (40%), radiotherapy (8%) or both (49%). 57 (40%) patients received CT alone, which was the most frequent mono-second-line therapy opted for. Most frequent indications for repeat-surgery were maximum possible tumor resection mass reduction and symptom relief (62% and 21%, respectively). Univariate analysis of re-operated vs. not operated patients, showed significant differences for age (p = 0.0001*) and Karnofsky Performance status (KPS) > 70 at both primary and repeat tumor resection (p = 0.013* and 0.0001*, respectively). The operated group had a significantly lower Charlson-comorbidity-index <= 3 (p = 0.004*) and larger tumor size (p = 0.0001*). Complication risk at recurrence was not significantly different between groups (p = 0.069). However, patients chosen for repeat surgery had significantly less complications at index surgery (p = 0.006*). Median time from recurrence to death was 11 months (range, 1-33 months) for operated patients as opposed to 5 months (range, 0-22 months) for not operated patients. The former survived significantly longer; 19 months compared to 13 months for those not operated upon (p = 0.002*). Conclusions: Our study depicts that patients eligible for repeat-surgery at GB recurrence are characterized by a KPS > 70% before primary and repeat-surgery, Charlson-comorbidity-index <= 3, large tumor size and young age. These well-selected patients survive significantly longer after repeat-surgery without being at a higher complication risk in comparison to patients not operated upon. (C) 2015 Published by Elsevier B.V.