EPID-20. THE SURVIVAL BENEFIT OF VALPROIC ACID IN GLIOBLASTOMA TRENDS AWAY FROM SIGNIFICANCE WITH NEWER STUDIES: A SYSTEMATIC REVIEW AND META-ANALYSIS

Abstract

BACKGROUND: Glioblastoma (GBM) is a dismal disease that presents often with seizure. The use of anti-epileptic medication valproic acid (VPA) has been suggested to provide survival benefit. This study aims to quantify the survival benefit of VPA in GBM on overall survival (OS) and determine if a trend in reported effect over time exists. METHOD(S): Searches of 7 electronic databases from inception to April 2018 were conducted following PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) guidelines. There were 1498 articles identified for screening. Prognostic hazard ratios (HRs) derived from multivariate regression analysis were extracted and analyzed using meta-analysis of proportions and linear regression. RESULT(S): Seven observational studies reporting prognostic HRs satisfied selection criteria. They described 3631 primary GBM diagnoses, with 603 (17%) receiving VPA in their treatment. VPA was shown to confer a statistically significant OS advantage the pooled cohort (HR, 0.709; 95% CI, 0.551-0.914; I2=60.1%; p=0.008) without specific bias concerns. However, this survival advantage trended towards null significance in newer studies (effect coefficient, 1.151; p=0.023). CONCLUSION(S): The current literature suggests that VPA confers a significant, prognostic OS advantage in GBM independent of other prognostic factors. This may be through a variety of possible biological mechanisms and clinical consequences. However, newer studies are significantly trending away from such a positive association, towards null significance. Larger, prospective randomized controlled studies are needed to validate if VPA itself provides true OS benefit in GBM patients.