Trends in Treatment and Survival Among Patients with Glioblastoma in the United States from 2000 to 2020

Abstract

Background: Population-level patterns in the real-world use of initial treatment combinations for glioblastoma (GBM) remain poorly characterized. Materials and Methods: We assessed temporal trends in treatment and survival among patients with GBM in the SEER registry. Guideline-concordant multimodal therapy was defined as cancer-directed surgery combined with radiotherapy and chemotherapy. Outcomes were further stratified by extent of resection, including gross total resection (GTR) versus subtotal resection. Results: Among 46,186 patients, 43.5% underwent GTR, 32.0% subtotal resection, and 24.5% no surgery; 71.8% received radiotherapy and 61.3% received chemotherapy. From 2000 to 2020, utilization increased for any surgery (OR/year 1.03, 95% CI 1.028–1.036), radiotherapy (1.01, 1.010–1.017), and chemotherapy (1.08, 1.081–1.088), whereas the likelihood of GTR vs. subtotal resection declined (0.92, 0.915–0.922). Triple therapy increased modestly (1.01, 1.005–1.012) but plateaued at 30%. Older age, non-lobar tumors, unmarried status, and lower income predicted lower odds of triple therapy; after adjustment, calendar year showed a marginal decline in the odds of receiving triple therapy (aOR/year 0.99, 0.99–1.00). Median overall survival improved from 6.0 to 10.0 months, with gains in fixed-time survival. Patients receiving multimodal therapy demonstrated the longest survival overall; within this group, GTR-based multimodal therapy was associated with longer survival than subtotal resection-based multimodal therapy across age strata. Conclusions: Guideline-concordant multimodal therapy was associated with the longest survival. GTR rates did not rise in recent years, and substantial disparities in the deployment of GTR-based multimodal therapy persisted. Efforts to expand equitable access and prioritize maximal safe resection are essential to achieve greater population-level survival gains. These differences should be interpreted cautiously, as SEER cannot distinguish appropriate clinical treatment selection from limited access to care or treatment ineligibility.