Social vulnerability and survival among patients with high-grade glioma treated at a tertiary cancer center.
We examined associations of county- and tract-level Social Vulnerability Index (SVI) with survival and treatment-related outcomes in adults with IDH-wildtype glioblastoma (GBM) or IDH-mutant grade 4 astrocytoma.
Adults treated at MD Anderson from 2020 to 2025 with PROACTIVE consent, survival follow-up, and Texas residential data were included. The 2022 CDC/ATSDR SVI was modeled per 0.1 increase (10 percentile points). Primary Cox models for GBM adjusted for age, sex, performance status, extent of resection, MGMT methylation, insurance, and distance. Proportional hazards were assessed using Grambsch-Therneau tests.
Cohorts included 398 patients with GBM and 62 with IDH-mutant grade 4 astrocytoma. In GBM, SVI was not associated with survival in county-level analyses (HR per 0.1 increase, 1.00; 95% CI, 0.96-1.05; P = 0.920) or Houston-area tract-level analyses (HR, 1.05; 95% CI, 0.98-1.13; P = 0.176). In exploratory time-varying analysis, the SVI-mortality association attenuated over follow-up (interaction P = 0.028). Higher tract-level SVI was associated with greater odds of not undergoing gross-total resection (OR, 1.19; 95% CI, 1.07-1.33; q = 0.010) and lower odds of clinical-trial treatment (OR, 0.81; 95% CI, 0.71-0.92; q = 0.008). No significant survival association was identified in exploratory astrocytoma analyses.
Among patients who reached tertiary neuro-oncology care, primary Cox models did not identify a statistically significant association between SVI and survival. Higher tract-level SVI was associated with less extensive resection and lower odds of clinical-trial treatment.
Adults treated at MD Anderson from 2020 to 2025 with PROACTIVE consent, survival follow-up, and Texas residential data were included. The 2022 CDC/ATSDR SVI was modeled per 0.1 increase (10 percentile points). Primary Cox models for GBM adjusted for age, sex, performance status, extent of resection, MGMT methylation, insurance, and distance. Proportional hazards were assessed using Grambsch-Therneau tests.
Cohorts included 398 patients with GBM and 62 with IDH-mutant grade 4 astrocytoma. In GBM, SVI was not associated with survival in county-level analyses (HR per 0.1 increase, 1.00; 95% CI, 0.96-1.05; P = 0.920) or Houston-area tract-level analyses (HR, 1.05; 95% CI, 0.98-1.13; P = 0.176). In exploratory time-varying analysis, the SVI-mortality association attenuated over follow-up (interaction P = 0.028). Higher tract-level SVI was associated with greater odds of not undergoing gross-total resection (OR, 1.19; 95% CI, 1.07-1.33; q = 0.010) and lower odds of clinical-trial treatment (OR, 0.81; 95% CI, 0.71-0.92; q = 0.008). No significant survival association was identified in exploratory astrocytoma analyses.
Among patients who reached tertiary neuro-oncology care, primary Cox models did not identify a statistically significant association between SVI and survival. Higher tract-level SVI was associated with less extensive resection and lower odds of clinical-trial treatment.
Authors
Hsi Hsi, Wathoo Wathoo, Campbell Campbell, Nassif Nassif, Buchold Buchold, Nguyen Nguyen, McNeill McNeill, Hawk Hawk, Puduvalli Puduvalli, Kamiya-Matsuoka Kamiya-Matsuoka
View on Pubmed