Neurosurg Rev. 2026 Feb 2;49(1):186. doi: 10.1007/s10143-025-04088-7.
ABSTRACT
Bevacizumab only improves progression-free survival (PFS) but not overall survival (OS) in glioblastoma (GBM) patients. Drug-induced hypertension is a common adverse event associated with bevacizumab in GBM, and it may paradoxically be associated with a favorable treatment response. However, the prognostic role of hypertension as a biomarker for bevacizumab efficacy in GBM remains unresolved. This study aimed to systematically evaluate the prognostic role of drug-induced hypertension in GBM patients treated with bevacizumab. We included studies on hypertension and survival outcomes in GBM patients treated with angiogenesis inhibitors from PubMed, Cochrane Library, and Web of Science databases. We extracted time-to-event data, including hazard ratios, and reconstructed individualized patient data from Kaplan-Meier curves. We used a meta-analysis approach to analyze pooled hazard ratio outcomes. A total of 1082 patients were included from 7 studies. Of these, 215 (24.8%) patients developed drug-induced hypertension, while 867 (75.2%) patients were normotensive. Compared to normotensive patients, patients who developed drug-induced hypertension showed a median benefit of PFS ranging from 2 to 8 months and OS ranging from 4 to 10 months in individual studies. Pooled time-to-event analysis showed that drug-induced hypertension significantly prolonged both PFS (HR = 0.44; 95% CI:0.28-0.70; p = 0.008) and OS (HR = 0.50; 95% CI:0.30-0.83; p = 0.015). Meta-regression demonstrated that earlier onset of hypertension may confer a greater survival benefit (PFS: β = 0.0078, OS: β = 0.0056), and subgroup analysis indicated that a ≥ 140/90 mmHg threshold may serve as a practical biomarker cutoff. In conclusion, this meta-analysis suggest that drug-induced hypertension is significantly associated with improved PFS and OS in bevacizumab-treated GBM patients. These findings suggest its potential as a positive prognostic biomarker, warranting further prospective validation.
PMID:41622339 | DOI:10.1007/s10143-025-04088-7