Background/Objectives: Patients with unresectable glioblastoma (GBM) undergoing biopsy alone represent a clinically heterogeneous population with poor survival and limited access to standard oncological treatment. We investigated clinical, radiological, and histomolecular factors associated with early mortality after biopsy. Methods: We retrospectively analyzed 93 consecutive cases of patients with IDH-wildtype GBM who underwent biopsy without tumor resection between 2013 and 2025. Early death was defined as death within three months of biopsy. Baseline clinical, radiological, and molecular variables were evaluated using univariable analyses and exploratory multivariable logistic regression. Results: Median overall survival was 3 months, and 48/89 patients with available survival data (53.9%) died within three months. Preoperative Karnofsky Performance Status (KPS) and MGMT promoter methylation status (available for 71/93 patients) were nominally associated with early mortality in univariable analyses, although these associations did not remain significant after FDR correction. Age was not significant in univariable analysis but was associated with early mortality after adjustment in the exploratory multivariable model based on 62 complete cases. Conventional radiological parameters showed no consistent association with early death. Exploratory analysis suggested a high-risk phenotype combining KPS ≤ 60 and FLAIR signal crossing the midline. Overall, 23/93 patients (24.7%) did not initiate any adjuvant treatment, while only 17/93 (18.3%) completed the Stupp regimen. Conclusions: Early mortality is frequent in biopsy-only GBM. Clinical and molecular factors may help identify particularly vulnerable patients at diagnosis, supporting individualized treatment planning and shared decision-making. External validation in larger multicenter cohorts is warranted.

Background/Objectives: Recurrent glioblastoma has a poor prognosis and no universally accepted standard of care. Regorafenib has been investigated in this setting, but its use may be limited by treatment-related toxicity. This study evaluated the feasibility, safety, and clinical outcomes of reduced-dose regorafenib in patients with recurrent IDH-wildtype glioblastoma. Methods: We retrospectively analyzed 21 patients with recurrent IDH-wildtype glioblastoma (WHO 2021) treated at a single center after progression according to the Stupp protocol. Regorafenib was started at 80 mg/day on the standard 3-weeks-on/1-week-off schedule. We assessed overall survival from diagnosis (OS1) and from first progression (OS2), progression-free survival from diagnosis (PFS1) and from first progression (PFS2), treatment exposure, dose modifications, and treatment-related adverse events (CTCAE v5.0). Results: Median PFS2 was 5 months (95% CI, 3–8), median OS2 was 6 months (95% CI, 5–10), and median OS1 was 21 months (95% CI, 15–27). No grade ≥ 3 adverse events occurred, and no patient discontinued treatment permanently because of toxicity. Conclusions: In this retrospective single-center cohort, reduced-dose regorafenib was feasible and well tolerated in selected patients. These descriptive, hypothesis-generating findings warrant prospective studies to define the role of individualized regorafenib dosing in recurrent glioblastoma.