BIOS-07. SURVIVAL TIME OF GLIOBLASTOMA PATIENTS ENROLLED IN RANDOMIZED CONTROLLED TRIALS: A NEXT GENERATION META-ANALYSIS OF SURVIVAL CURVES

Abstract

Abstract INTRODUCTION Glioblastoma is the most common primary malignant brain tumor and remains an incurable disease with high mortality. Randomized controlled trials (RCTs) have investigated novel therapies that have shown promise, but there have been no major additions in the treatment armamentarium in almost two decades. OBJECTIVE In a first study of its kind, we investigate the cumulative change in survival times over years for GBM patients by combining the survival outcome data published by multiple trials. METHODS A systematic database search of PubMed for GBM RCTs with significant outcomes published after 2005 was conducted. All Kaplan-Meier (KM) survival curves were analyzed. Individual patient data was extracted from the survival curves and a cumulative survival curve was generated with corresponding times calculated. The primary outcome for all included RCT was mortality at longest follow-up. RESULTS We identified 14 RCT which contained KM curves presenting sufficient data for individual patient survival data extraction. Survival data of a total of 2294 patients was extracted with 1396 (60.85%) patients in the treatment arms and 898 in the control arms (39.15%). Patients in the control arm received the standard care of GBM management, whereas those in the treatment arms received differing experimental regimens. After generating the cumulative survival curve, the median survival time for patients in the treatment arm was found to be 15.2 months and 11.2 months for the control arm (P < 0.0001). The two-year survival rate was 31.0% for the treatment arm and 15.6% for the control arm. CONCLUSION The survival times observed in GBM patients receiving the standard care of management were worse than anticipated. In general, however, overall estimates of survival among patients with GBM have improved with more patients surviving to 2 years.