Abstract
Surgical re-resection for recurrent IDH–wildtype glioblastoma remains controversial, and no molecular biomarkers currently inform on tailoring the extent of cytoreduction and hence onco-functional balance at recurrence. On the other hand, while the prognostic relevance of O6-methylguanine-DNA-methyltransferase (MGMT) promoter methylation is established in glioblastoma. Its role in surgical decision-making at recurrence remains unclear. We analyzed a cohort of 153 patients who underwent re-resection for IDH-wildtype recurrent glioblastoma (WHO classification 2021) between 2015 and 2024. Volumetric analysis of contrast-enhancing (CE) tumor on pre- and early postoperative MRI was performed. Patients were stratified by MGMT promoter methylation status and residual tumor volume (RTV): 0 ml, < 1 ml, > 1 ml. Functional outcomes up to one year after surgery were longitudinally assessed using the National Institutes of Health Stroke Scale (NIHSS). There was no significant difference in RTV between MGMT-methylated (n = 58, 38%) and unmethylated patients (p > 0.999). The benefit from complete or near-complete resection was more pronounced in MGMT-unmethylated tumors, with a median postoperative survival of 288 days for 0 to 1 ml RTV versus 190 days > 1 ml (p = 0.024; difference = 98 days), compared to MGMT-methylated tumors with a median survival of 411 days versus 378 days (p = 0.043; difference = 33 days). MGMT-unmethylated patients with 0 ml RTV showed significantly higher KPS values six weeks postoperatively (p < 0.05) compared to higher RTVs. Postoperative deficits were comparable across groups. These findings identify MGMT promoter methylation as a clinically actionable molecular modifier of the oncological benefit of cytoreduction and support MGMT-stratified, precision neurosurgical strategies for recurrent glioblastoma, especially around functionally eloquent tumors. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s40478-026-02234-w.