[Single center experience with bevacizumab in a cohort of patients treated for a relapsing glioblastoma].

Purpose: Although there is no standard treatment for recurrent glioblastoma, prospective data in selected patients have suggested the usefulness of bevacizumab. We report our single center experience with bevacizumab in a cohort of patients treated for a relapsing glioblastoma.

Methods: We performed a retrospective analysis of consecutive patients treated with bevacizumab for a relapsed glioblastoma, between 2008 and 2013. Tumor responses, toxicities, time to progression and overall survival rates were analyzed.

Results: Thirty-five consecutive patients were identified. They were treated with bevacizumab 10mg/kg biweekly, associated with irinotecan (n=29; 84%), temozolomide (n=3; 9%) or as single agent (n=3; 9%) for a glioblastoma relapsing after chemoradiation (n=29) or after first line temozolomide only because of a poor general health status or because of multifocal tumor. Two (6%), 28 (80%) and five (14%) patients presented with Recursive Partitioning Analysis (RPA) III, IV and V-VI, respectively. After 2-3 months of treatment, median dose of prednisolone per patient was decreased three times. Clinical improvements or stability were reported in eight (23%) and 17 patients (49%). The best tumor response was partial response in 14 patients (40%), stable disease in nine patients (26%) and tumor progression in 11 patients (31%). Toxicities requiring treatment disruption were reported in five patients (14%). Median survival was 18.4 months (5-41 months). Median time interval between bevacizumab initiation and its disruption because of clinical/radiological progression and/or toxicity was 5.0 months (0.6-21.4 months). Median survival from bevacizumab initiation was 8.1 months (1.4-34 months).

Conclusion: This single center retrospective experience suggests that bevacizumab is active for recurrent glioblastoma, in a series of poorly selected patients. Median survival times were in the range of those reported in therapeutic trials. This study questions the validity of usual predictive factors in the era of bevacizumab.