INTRODUCTION: Treatment options for recurrent glioblastoma remain limited, with only modest activity of second line chemotherapy. Despite the absence of well controlled trials, bevacizumab is widely used for recurrent glioblastoma. Nonetheless, it remains unclear if the high response rates observed after treatment with bevacizumab translate into an overall survival (OS) benefit. We report the first randomized phase II trial on bevacizumab in recurrent glioblastoma with a control arm without bevacizumab. METHODS: This was multicenter phase II study conducted in 14 Dutch sites. Patients with a first recurrence of a glioblastoma after chemo-irradiation with temozolomide were randomized between treatment with oral lomustine 110 mg/m2 once every 6 weeks, with bevacizumab 10 mg/kg once every 2 weeks intravenously or with the combination lomustine 110 mg/m2 every 6 weeks and bevacizumab 10 mg/kg every 2 weeks. Response was assessed using the RANO criteria. Primary endpoint was OS at 9 months. A safety analysis was planned after the first ten patients completed two cycles of 6 weeks in the combination arm. RESULTS: 153 patients were randomized, of which 148 were eligible. After the safety analysis the lomustine dosage in the combination arm was reduced to 90 mg/m2 because of thrombocytopenia, resulting in bevacizumab dose delays. After reduction the treatment in the combination arm was well tolerated. The percentage 9 month OS (95% confidence interval) was 43% (29, 57) in the lomustine arm, 38% (25, 51) in the bevacizumab arm, and 59% (43, 72) in the bevacizumab/lomustine 90 arm. Response rates were 35-40% for bevacizumab treated patients, and 5% in the lomustine single agent arm. CONCLUSION: This study does not show a meaningful OS at 9 months after treatment with bevacizumab alone. However, the results of combination therapy of bevacizumab with lomustine met the prespecified criterion for further phase III studies. AKNOWLEDGMENTS: This study was supported by an unrestricted grant from Roche Nederland bv. The study was also supported by grant nr DDHK 2010-4678 from the ‘KWF Kankerbestrijding’ (Dutch Cancer Society).