Prognostic Impact of Clinical Diagnosis-based Treatment in Glioblastoma Patients Aged ≥80 Years: A Single-center Cohort Study Focused on Performance Status Maintenance

In aging societies such as Japan, the number of patients aged ≥80 years with glioblastoma who are unsuited for surgery is rapidly increasing. At our institution, we adopted a therapeutic strategy based on clinical diagnosis without histological confirmation when surgery was not feasible. This study aimed to retrospectively evaluate the short-term outcomes and clinical value of surgical intervention and histological confirmation in this population. Patients aged ≥80 years diagnosed with glioblastoma between 2011 and 2025 were reviewed. Clinical diagnosis was applied when patients were unfit for surgery. Patient data, including age, sex, Karnofsky performance status, extent of resection, treatment options, and complications, were analyzed for associations with overall survival, Karnofsky performance status improvement, and discharge home. Among 203 cases, 31 patients (15.3%) were included. Their mean age was 84 years; the median Karnofsky performance status scores at admission and discharge were 50 and 60, respectively. Six patients underwent resection, 14 biopsy, and 11 clinical diagnosis. The clinical diagnosis group showed better functional outcomes, including improved Karnofsky performance status, shorter hospitalization, and higher discharge home rates than the biopsy group. Bevacizumab contributed significantly to these outcomes. Perioperative complications occurred in 4 patients; none recovered functionally or were discharged home. Low Karnofsky performance status at discharge was the only independent predictor of poor overall survival. Surgical intervention did not significantly affect the functional and survival outcomes. In conclusion, clinical diagnosis-based treatment, particularly with bevacizumab, could help preserve function and facilitate discharge in very elderly patients with glioblastoma. Biopsies might negatively impact the outcomes in the present cohort.