Evaluating short-term survivors of glioblastoma: A proposal based on SEER registry data

Background: Glioblastomas (GBMs) are central nervous system tumors with a poor prognosis and limited treatment options. Although small subsets of GBM patients survive longer than 3 years, there is little evidence regarding the prognostic factors of GBM. Therefore, we conducted a thorough characterization of GBM in the United States.
Methods: We queried the Surveillance, Epidemiology, and End Results database between 2000 and 2021 to extract age-adjusted incidence rates (AAIRs), age-adjusted mortality rates (AAMRs), and survival data for GBM. We compared trends in AAIR, AAMR, and survival time across age groups 0–14, 15–39, 40–69, and 70+ years. Also, we employed the Fine–Gray competing risk model among short-term survivors (STSs), defined as those with a survival time of 6 months or less, and long-term survivors (LTSs), defined as those with a survival time of 3 years or more.
Results: This study included 60 615 incident GBM cases, 54 998 GBM-specific deaths, and 47 207 GBM patients with available survival time between 2000 and 2021. The mortality-to-incidence ratio was constant among STSs, whereas it increased with age among LTSs. Higher age and male sex were significantly associated with GBM-specific death among LTSs, whereas non-Hispanic White and less intensive treatments were associated with GBM-specific deaths among STSs. Interestingly, higher age was significantly associated with other causes of death among STSs.
Conclusions: STSs partially consist of populations who died from causes other than GBM. It is important to include only GBM-specific deaths in STS groups to conduct reproducible research comparing STSs and LTSs.