How survival time reshapes prognostic risk in giant cell glioblastoma

To address the lack of dynamic prognostic tools for the rare Giant Cell Glioblastoma (gcGBM), this study utilized conditional survival (CS) analysis and a population-based nomogram to provide evolving, personalized survival estimates beyond baseline diagnosis.

We identified 490 patients with gcGBM from the Surveillance, Epidemiology, and End Results (SEER) database (2000-2022). The cohort was split into training (n = 343) and validation (n = 147) sets. Conditional survival probabilities-CS(t|s) = S(t + s)/S(s)-were calculated to assess the likelihood of surviving an additional t years given s years already survived. A Random Survival Forest with Recursive Feature Elimination (RSF-RFE) algorithm was employed to select the most parsimonious predictors. A CS-nomogram was constructed and evaluated using calibration curves, time-dependent AUC, and Decision Curve Analysis (DCA). An interactive web-based calculator was also developed.

The baseline 5-year overall survival (OS) was 14%. However, CS analysis revealed that the prognosis improved significantly over time: for patients who survived 3 years, the probability of surviving to year 5 rose to 69%, and reached 83% for those surviving to year 4. The RSF-RFE identified six key predictors: age, sex, tumor site, surgery, radiotherapy, and chemotherapy. The resulting CS-nomogram demonstrated high discriminatory power and clinical net benefit in both cohorts. Risk stratification effectively distinguished high-risk and low-risk groups.

Prognosis in gcGBM is a dynamic process. The risk of mortality is highest in the early years post-diagnosis but decreases substantially for long-term survivors. Our validated CS-nomogram and web-based tool provide clinicians with a practical means to offer personalized, time-adjusted survival estimates, facilitating better patient counseling and refined follow-up strategies.