Bevacizumab is widely used for recurrent high-grade glioma, but the real-world effectiveness of bevacizumab with or without irinotecan remains uncertain. We evaluated outcomes of bevacizumab-based regimens in a large multicenter Turkish cohort.

In this retrospective study from 30 centers, adults with recurrent glioblastoma or other primary brain tumors treated with a bevacizumab-containing regimen at first or second progression were included. Patients received bevacizumab monotherapy, bevacizumab plus low-dose weekly irinotecan, or bevacizumab plus standard-dose irinotecan every 14 days. Tumor response, progression-free survival (PFS), overall survival (OS), and toxicity were assessed. Prognostic factors were analyzed using Cox regression.

A total of 437 patients were included; 78.0% had glioblastoma. Treatment consisted of bevacizumab monotherapy in 9.4%, bevacizumab plus weekly irinotecan in 8.5%, and bevacizumab plus irinotecan every 14 days in 82.2% of patients. The objective response rate was 41.6%, and the disease control rate was 80.1%. Median OS was 10.77, 7.37 and 9.77 months (log-rank p = 0.024), and median PFS was 5.77, 3.93 and 6.43 months (p = 0.005), respectively. On multivariable analysis, glioblastoma histology independently predicted shorter PFS and OS, whereas a higher number of treatment cycles and antiepileptic drug use were associated with longer PFS. For OS, the irinotecan–bevacizumab every-14-day regimen and a higher number of treatment cycles were associated with improved survival compared with bevacizumab monotherapy, while baseline corticosteroid use and discontinuation of bevacizumab-containing therapy were independent adverse prognostic factors.

In this large real-world cohort, bevacizumab-based therapy achieved meaningful disease control and survival in recurrent primary brain tumors. An irinotecan–bevacizumab regimen administered every 14 days was associated with superior OS at the expense of increased but manageable chemotherapy-related toxicity, supporting its use in appropriately selected patients.