Background: The use of alkylating chemotherapy versus bevacizumab for recurrent glioblastoma remains controversial. Here, we tested the hypothesis that the activity of alkylators, but not that of bevacizumab, would be associated with the O6-methylguanine DNA methyltransferase (MGMT) promoter methylation status.
Methods: We analyzed a cohort of patients treated at centers of the German Glioma Network or the University Hospital Zurich with alkylating agent-based chemotherapy (n = 260) or bevacizumab without or with irinotecan (n = 84) for first recurrence of glioblastoma. Outcome was stratified for O6-methylguanine DNA methyltransferase (MGMT) status and crossover to bevacizumab or alkylators at further progression.
Results: Median post-recurrence survival-1 (PRS-1) for patients receiving alkylating agents at first recurrence was longer than with bevacizumab (11.1 versus 7.4 months, p < 0.001). The use of alkylators was associated with longer PRS-1 for patients with a methylated versus unmethylated MGMT promoter (p = 0.017). For patients receiving bevacizumab, PRS-1 was not different with or without MGMT promoter methylation. PRS-1 was longer in patients receiving alkylating chemotherapy compared to bevacizumab for patients with methylated (p < 0.001) or unmethylated MGMT promoter (p = 0.034). For patients with alkylators at first recurrence receiving bevacizumab at any further recurrence, PRS-1 was longer than in patients receiving bevacizumab first and alkylators thereafter (p = 0.002).
Conclusions: This study confirms limited value of bevacizumab in recurrent glioblastoma independent of MGMT status. Alkylating agents have activity in recurrent glioblastoma, especially in the context of MGMT promoter methylation.
Keywords: Bevacizumab; Chemotherapy; Glioblastoma; Nitrosourea; Temozolomide.