Prognosis in glioblastoma: insight gained from recent prospective trials

Gittleman et al. report on the creation and independent validation of a nomogram estimating the survival of patients with glioblastoma multiforme (GBM) following chemoradiation therapy (CRT) (1). The nomogram was developed using three statistical approaches (cox proportional hazards regression, random survival forests, and RPA) from available data points of patients accrued to Radiation Therapy Oncology Group (RTOG) 0525, a clinical trial evaluating the benefit of dose-intensified temozolomide (TMZ) in patients with newly diagnosed GBM (2). The algorithm was validated in an independent population of patients who completed therapy on the RTOG 0825 clinical trial (2,3), which asked if the addition of bevacizumab to standard therapy for GBM improved survival. Neither of these randomized trials demonstrated a statistically significant difference in their primary endpoint (survival), so the inclusion of all patients in both cohorts is valid. The resultant nomogram provides an individualized tool for predicting 6, 12, and 24-month survival in patients with newly diagnosed GBM based on age, sex, performance status, extent of surgical resection, and O6-methylguanine-DNA-methyltransferaase (MGMT) methylation status.