Bevacizumab alternating chemotherapy for improving the survival of patients with recurrent high-grade glioma.

BACKGROUND

High-grade glioma (HGG) is an aggressive tumor for which there are no effective therapies at recurrence, especially for isocitrate dehydrogenase (IDH)-wild-type glioblastoma. This retrospective study compared survival outcomes between patients receiving bevacizumab alternating chemotherapy (BAC) and those receiving bevacizumab (BEV) alone.

METHODS

We collected data from 95 adult patients with rHGG who were treated at our institute between January 2018 and August 2023. The patients were divided into 3 groups based on treatment and glioma grade: BAC regimen to treat grade 3 gliomas ( = 23), BAC regimen to treat grade 4 gliomas ( = 29), and treatment with BEV alone ( = 43). The BAC regimen included 2 cycles of etoposide + carboplatin, followed by 1 cycle of cyclophosphamide + vinblastine, with bevacizumab (10 mg/kg) every 4 weeks. One full cycle lasted approximately 3 months. We analyzed overall survival (OS) and postrecurrence survival (PRS).

RESULTS

In patients with grade 4 gliomas, the BAC regimen significantly improved survival compared with BEV alone, with a median OS of 29 versus 19 months and a PRS of 16 versus 10 months (both  < .05). In the IDH-wild-type subgroup, the BAC regimen produced a median OS of 27 versus 19 months and a PRS of 16 versus 10 months ( < .05). The 2-year OS and PRS rates were also higher in the BAC groups. Notably, patients with MGMT-methylated grade 4 gliomas treated with the BAC regimen had the longest median OS, 33 months.

CONCLUSIONS

The BAC regimen appears effective and well tolerated in adult patients with rHGG, particularly in younger patients. Its alternating design may improve the median OS (29 vs. 19 months) and PRS (16 vs. 10 months) of patients with grade 4 gliomas while maintaining safety. As a practical option for those ineligible for clinical trials, BAC warrants further evaluation in prospective randomized studies to confirm its benefits and address the limitations of retrospective analysis.