Access-related disparities and survival in high-grade glioma: A social vulnerability analysis at a Texas tertiary cancer center.

Abstract

e14043 Background: Glioblastoma (GBM) and astrocytoma, IDH-mutant grade 4, are aggressive brain tumors requiring timely access to care. While social vulnerability has been linked to cancer outcomes at the population level, its relevance among patients treated at tertiary referral centers remains unclear. Methods: We conducted a retrospective cohort study of adults with GBM (N=400) and astrocytoma, IDH-mutant grade 4 (N=84) treated at a comprehensive cancer center from 2020–2025. OS was defined from diagnosis to death or last follow-up. Social vulnerability was measured using the CDC Social Vulnerability Index (SVI), assigned at the county level for all patients and at the census tract level for Houston metropolitan patients. SVI was analyzed continuously and by quartiles. Kaplan–Meier and multivariable Cox models evaluated associations with OS, adjusting for age, sex, Karnofsky Performance Status (KPS), extent of resection, MGMT promoter methylation, insurance type, and distance to center. Results: SVI was not independently associated with OS after adjustment (GBM county-level HR per 0.1 increase 1.02, 95% CI 0.97–1.08; tract-level HR 1.05, 95% CI 0.97–1.12; astrocytoma grade 4 county-level HR 1.04, 95% CI 0.95–1.14). Higher SVI was associated with non-private insurance, lower KPS, and reduced likelihood of gross total resection (all p<0.05). Distance to the tertiary cancer center was not independently associated with OS. Houston metropolitan and non-Houston patients had similar outcomes once treated (GBM median OS 28.8 vs 28.3 months; 24-month OS 48.7% vs 47.9%; astrocytoma grade 4 median OS not reached; 24-month OS 84.0% vs 86.5%). In contrast, a county-based comparison across Texas showed a modest survival difference (log-rank χ²=4.704, p=0.030), with one region experiencing ~24% lower hazard of death (HR 0.76, 95% CI 0.58–0.99) despite a small absolute median OS difference (23.36 vs 21.16 months). Conclusions: Among patients with high-grade glioma treated at a tertiary referral cancer center, neighborhood-level social vulnerability was not independently associated with survival after adjustment, whereas clinical and tumor-related factors predominated. Although regional survival differences were observed across Texas, these differences attenuated after tertiary referral, indicating that disparities primarily arise upstream—at diagnosis, referral, and access to definitive treatment—rather than during specialty neuro-oncologic care. Together, these findings suggest that improving timely access to tertiary neuro-oncologic care at diagnosis represents a critical opportunity to reduce survival disparities and supports further investigation of interventions that expedite referral and treatment initiation.