GE-29 * EXPRESSION SUBCLASS PROFILE IN PSEUDOPROGRESSION AND TRUE PROGRESSION IN NEWLY DIAGNOSED GBM

Abstract

INTRODUCTION: The hallmark of glioblastoma multiforme (GBM) is its penchant for relentless progression. Pseudoprogression describes a post-treatment reaction demonstrating increased edema and contrast enhancement similar to typical tumor progression except that on subsequent imaging without escalation of antitumor therapy these changes stabilize or revert [1]. Accurate identification of pseudoprogression has important implications for therapy and research and potentially prognosis as well. Increased cellular proliferation (Ki-67 indices) and the presence of a methylated O-6-methylguanine-DNA methyltransferase (MGMT) promoter have been associated with higher rates of pseudoprogression [2,3]. However, more sensitive and specific biomarkers of pseudoprogression are needed. This study seeks to identify novel indicators of pseudoprogression. METHODS: Patients were identified using the Hermelin Brain Tumor Center database at Henry Ford Hospital. Tissues from 52 patients with newly diagnosed GBM between 1992 and 2011 were gathered and whole genome sequencing and subtyping was performed by The Cancer Genome Atlas researchers. Retrospective chart review was carried out. Patients were assigned to either pseudoprogression (PP) or true progression (TP) groups based on whether changes suggestive of disease progression on MRI within 2 months of post-operative therapy initiation regressed without additional antitumor therapy during the ensuing 4 months. The incidence of pseudoprogression and GBM subclass were correlated using Fisher's Exact Test. RESULTS: Forty-one of 52 (79%) cases were identified as TP while 11/52 (21%) were found to have PP. In our study population, PP was associated with significantly increased median survival compared with TP (735 versus 313 days, respectively, p < 0.0012). The molecular subclass profile for both groups included a predominance of Mesenchymal and Neural subtypes, revealing no correlation between GBM subclass and the risk of pseudoprogression (p < 0.8069). CONCLUSIONS: Pseudoprogression correlated with improved survival. Interestingly, Mesenchymal and Neural GBM subclasses predominated in our PP group, suggesting a factor independent of molecular subclassification may predict pseudoprogression.