The monoclonal anti-vascular endothelial growth factor (VEGF) antibody bevacizumab has recently received considerable attention from neuro-oncologists. Bevacizumab has demonstrated high radiographic response rates, a decreased need for corticosteroids and temporary improvement in neurological function in glioblastoma patients in several uncontrolled studies. In a pivotal trial overall survival of patients with recurrent glioblastoma treated with bevacizumab was 9.2 months (95% confidence interval 8.2 to 10.7), with 42.6% (confidence interval 29.6 to 55.5%) of patients alive and progression-free at six months. Based on these results, the US Food and Drug Administration (FDA) provisionally approved bevacizumab for the use in recurrent glioblastoma. Two large multinational phase III trials are currently evaluating the efficacy of bevacizumab in addition to standard adjuvant chemo-radiotherapy in newly diagnosed glioblastoma.
A number of new radiological phenomena including changes in contrast media uptake, perfusion, apparent diffusion coefficients and metabolic activity have recently been described in glioblastoma patients undergoing bevacizumab therapy. In the paper by Bähr et al., another previously unknown and particularly interesting neuroradiological feature is documented in such patients. Bähr et al. demonstrate that up to 60% of glioblastoma patients develop tumour calcifications within few weeks after initiation of bevacizumab therapy. These changes are apparent as hyperintense lesions on standard T1-weighted magnetic resonance images and were confirmed as tumour calcifications using computed tomography. Intriguingly, the prognosis of patients developing such tumour calcifications was significantly better than that of patients without calcifications both with regard to progression-free and overall survival times. The mechanism of tumour calcification as an effect of bevacizumab therapy in glioblastoma has not been experimentally addressed in the present work, but the authors speculate that they may origin from changes in tumour blood vessels induced by VEGF blockage. In sum, the data presented by Bähr et al. are interesting, as they may introduce a new and easily assessable neuroradiologcial biomarker for bevacizumab-treated glioblastoma patients. It will be interesting to validate their findings in further prospectively collected patient cohorts, e.g. the patients enrolled in the two ongoing large phase III trials evaluating bevacizumab in newly diagnosed glioblastoma.