Although promising preliminary results have been widely observed with bevacizumab for recurrent malignant gliomas, many unanswered questions remain to be resolved to achieve an optimal outcome. No predictive biomarkers of a survival benefit from bevacizumab have been established, and no consensus exists about the response or survival benefit regarding the prior recurrence pattern or tumor location. Here we retrospectively analyzed the clinical benefit from bevacizumab for recurrent malignant gliomas in relation to the prior recurrence pattern or tumor location. Thirty-one consecutive patients with recurrent malignant gliomas who were treated with bevacizumab were investigated. The treatment response and survival benefit from bevacizumab were analyzed in association with age, sex, Karnofsky performance status, prior pathological diagnosis, prior recurrence pattern, primary location of tumor, recurrence status, and expression of angiogenic and hypoxic markers. The group with leptomeningeal dissemination had a significantly shorter median overall survival with bevacizumab (OSBev) (6.0months, 95% confidence interval (CI) 1.4-10.7) compared to those in the local/distant group (11.8months, 95% CI 6.1-17.4). The median OSBev of the infratentorial tumor group and supratentorial tumor group were 9.2months (95% CI 5.0-13.4) and 10.4months (95% CI 6.6-14.3), respectively. With multivariate analysis, the prior recurrence pattern was the only independent prognostic factor of OSBev. Patients with leptomeningeal dissemination of recurrent malignant glioma experienced minimal benefit from bevacizumab. Therefore, in the context of cost effectiveness, bevacizumab is not recommended for patients with leptomeningeal dissemination.
Keywords: Bevacizumab; Efficacy; Prior recurrence pattern; Tumor location.
Copyright © 2017 Elsevier Ltd. All rights reserved.