2626 - Comorbidity Burden and Radiotherapy in Elderly Glioblastoma Patients
Presenter(s)
S. M. R. Leiss1, H. X. Hou2, D. Schmottermeyer1, C. Diehl1, B. Wiestler3, J. C. Peeken1, K. Borm4, C. Negwer5, A. Wagner5, I. Yakushev6, C. Delbridge7, M. Mitsdörffer8, F. Schmidt-Graf8, B. Meyer9, S. E. Combs10,11, and D. Bernhardt12,13; 1Department of Radiation Oncology, TUM University Hospital, Munich, Germany, 2Department of Radiation Oncology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, 3Department of Diagnostic and Interventional Neuroradiology, echnical University of Munich, Klinikum rechts der Isar, Munich, Germany, 4Department of Radiation Oncology, TUM Unviversity Hospital, Munich, Germany, 5Department of Neurosurgery, TUM University Hospital, Munich, Germany, 6Department of Nuclear Medicine, TUM University Hospital, Munich, Germany, 7Department of Pathology and Neuropathology, TUM University Hospital, Munich, Germany, 8Department of Neurology, TUM University Hospital, Munich, Germany, 9Department of Neurosurgery, Klinikum rechts der Isar, Technical University of Munich, Munich, Germany, 10Department of Radiation Oncology, School of Medicine and Health and Klinikum rechts der Isar, Technical University of Munich (TUM), Munich, Germany, 11Institute of Radiation Medicine (IRM), Department of Radiation Sciences (DRS), Helmholtz Center Munich, Munich, Germany, 12Department of Radiation Oncology, TUM University Hospital, München, Germany, 13Institute of Radiation Medicine (IRM), Department of Radiation Sciences (DRS), Neuherberg, Germany
Purpose/Objective(s): Elderly glioblastoma (GBM) patients face challenging treatment conditions due to comorbidities and age-related frailties. The age-adjusted Charlson Comorbidity Index (ACCI) accounts for age and comorbidities and serves as a predictive tool for survival rates. We hypothesize that ACCI predicts survival outcomes in elderly GBM patients undergoing radiotherapy and radio-chemotherapy. The primary outcomes assessed were overall survival (OS) and progression-free survival (PFS).
Materials/Methods: We retrospectively analyzed 166 elderly GBM patients (=61 years) treated with radiotherapy at a single institution (2001–2021). Follow-up data were collected from electronic records and clinical follow-ups. Treatment protocols were standardized, but variations in adjuvant therapies and comorbidities were considered. Kaplan-Meier analysis estimated OS for ACCI groups (=5 vs. =6). Cox proportional hazards models assessed associations between ACCI score, age, comorbidities, and survival. Harrell’s C-Index evaluated model discrimination. A Random Survival Forest identified key survival predictors using permutation importance with bootstrapped confidence intervals. The multivariate models and RSF were used for accounting for biases inherent in retrospective analyses, to adjust for confounders and identify independent predictors of survival.
Results: Among the 166 patients, 84 had an ACCI total score of = 5, and 82 had a score of = 6. Kaplan-Meier analysis revealed a median OS of 18.17 months for ACCI = 5 and 11.01 months for ACCI = 6, showing a significant difference (p=0.0337). The Cox model identified diabetes without complications as significantly associated with higher mortality (HR = 2.86, 95% CI: 1.57–5.44, p=0.001), whereas the ACCI total score (HR = 1.00, p=0.997) was not significant.
ACCI was not significantly associated with OS (HR = 1.08, p = 0.23) or PFS (HR = 1.11, p = 0.23), with Harrell’s C-Index indicating poor discriminative ability (0.589 for OS, 0.531 for PFS). In contrast, diabetes without complications was significantly associated with worse OS (HR = 2.06, p = 0.01) and PFS (HR = 2.15, p = 0.01), though its predictive value remained low (C-Index: 0.540 for OS, 0.543 for PFS). In the RSF model, ACCI contributed to both OS (0.027, 95% CI: 0.003–0.073) and PFS (0.017, 95% CI: 0.003–0.045) but had smaller effects on PFS, where other variables were stronger predictors.
Conclusion: A clinically relevant survival difference of approximately seven months was observed between ACCI groups. High ACCI scores correlate with poorer survival rates and could aid in treatment stratification for elderly glioblastoma patients. However, prospective validation is needed to confirm its clinical utility.