Background: Glioblastoma Multiforme (GBM) is one of the most aggressive primary brain tumors and is associated with a dismal prognosis. The median survival after the primary diagnosis remains poor, even after multimodal treatment approaches. However, a few patients have been reported to have long term survival greater than three years. A number of studies have attempted to define factors capable of predicting long term outcomes in specific patient groups. This article reports the outcomes of a very large group of patients diagnosed with GBM, and analyzes specific prognostic factors known to influence survival in these patients.
Methods: We used the Surveillance, Epidemiology, and End Results (SEER) database of the US National Cancer Institute (NCI) to investigate various patient-related and treatment-related factors that could influence the long term survival in patients diagnosed with glioblastoma. A total of 34,664 patients aged 20 years or older with a diagnosis of GBM during the years 1973 to 2008 were studied. Overall survival outcomes were examined with Kaplan-Meier survival analysis and Cox hazard models.
Results: Asian/Pacific Islanders had a better survival compared to the white population (P = <0.001). Patients diagnosed with GBM during the years 2000 to 2008 had a superior survival rate when compared with earlier decades (P = <0 .001). Statistically significant improvements in overall survival were also found for patients who received surgical resections, and adjuvant radiation treatment versus no radiation (P-values <0.001). Young age was also found to be highly predictive of improved overall survival rates when separated into age groups as well as when studied as a continuous variable.
Conclusions: Clinical pretreatment and treatment factors, including young age at diagnosis, Asian/Pacific Islander ethnicity, recent year of diagnosis, surgical resection and the use of adjuvant radiation therapy favorably influence survival in patients diagnosed with glioblastoma.
Trial registration: All data were obtained from the United States Surveillance Epidemiology and End Results (SEER) database.