The role of bevacizumab in the setting of recurrent glioblastoma (GBM) is still an argument of debate. In Europe, the EMA did not approve this agent despite the promising results in terms of response rate and progression-free survival provided by early Phase II studies without a calibration arm [1,2]. Therefore, new prospective randomized trials with bevacizumab in the recurrent setting have been conducted, and have been recently reported: the BELOB [3] and the AVAREG [4] trials.Both these trials adopted overall survival rates as primary end points, to avoid issues related to response assessment [5], and included a nitrosourea calibration arm, generating novel information in this setting.For many years response rate was considered the primary end point for Phase II studies in medical oncology. In neurooncology progression-free survival (PFS) and, in particular the incidence of patients free of progression at 6 months (PFS-6) have been considered the most appropriate end point for Phase II studies since the nineties [6], radiological responses being scarce (6-8%) [7,8], and there being a correlation between PFS and overall survival [9].Yet the action of bevacizumab on blood-brain barrier permeability directly affects the MRI evaluation made in these cases, thus potentially compromising the reliability of the assessment of the response rate and PFS. Therefore PFS-6 was no longer considered a sound end point [5].In the BELOB trial [3], a randomized Phase II study that enrolled 153 patients to receive bevacizumab alone or in combination with lomustine, or lomustine alone, Taal and Colleagues adopted the Response Assessment in Neuro-Oncology (RANO) criteria [10] for disease assessment, and reported that the radiological response following bevacizumab was correlated with survival (HR: 0.37; 95% CI: 0.23-0.58). These results are consistent with those reported in previous retrospective analyses in which imaging-based end points (response rate or PFS), obtained with bevacizumab were found to correlate with survival [11,12].Does this suggest that the best end point in Phase II studies should be re-reviewed?As yet, these findings are suggestive that neuroradiology still has a role in disease assessment for patients treated with bevacizumab, but further data are needed.