P01.103 Molecular landscape of diffuse gliomas associated with seizures

Abstract

BACKGROUND: Glioma is the most common primary brain tumour and often confers a grave prognosis. In low-grade gliomas seizures occur frequently (70-80%) and usually are pharmacoresistant. While genes coding for interleukins and implicated in neurotransmitters signalling have been associated with focal seizures, a molecular profile of glioma-associated epilepsy (GAE) remains largely understudied. To fll this gap, we analysed whole-genome expression and DNA methylation profiles of almost 500 glioma samples to find seizure-related correlates. Uncontrolled seizures are associated with 1% annual mortality, but GAE correlates with a favourable prognosis. Therefore, characterization of epileptogenesis in GAE could allow to identify new glioma prognostic markers. MATERIAL AND METHODS: Gene expression and DNA methylation profiles of WHO grade II and III diffuse gliomas samples were downloaded from The Cancer Genome Atlas website and analysed using R statistical software. The genes expressed differentially between samples with and without seizure history were identified using multtest. Fast Gene Set Enrichment Analysis followed by motif enrichment analysis done using PWMEnrich and Differential Gene Correlation Analysis were performed to extract functional correlations. The expression levels of the identified genes were compared with DNA methylation profiles, using minf, and heterogeneity of cell subpopulations was assessed using xCell. RESULTS: Majority of the upregulated genes were positive prognostic factors, whereas the downregulated correlated with shorter survival. Among the differentially expressed genes, FOXG1 and SEZ6 have established roles in epileptogenesis and cancer biology. We found that genes encoding components of G protein-coupled receptors signalling and development of neural cells characterised GAE history, whereas genes related to angiogenesis, immune cell infiltration, protein synthesis and epithelial-mesenchymal transition were associated with seizure free samples. Moreover, VEPH1, a gene involved in most of the former processes, and DGCR5, a lncRNA associated with prolonged survival in other neoplasms, were correlated with GAE and better survival. Overall, differentially expressed genes were enriched in the Neuron-Restrictive Silencer Factor (REST) motif and its expression had a prognostic role. CONCLUSION: This study identified molecular features characteristic for GAE. The majority of found features correlated with prolonged survival. Our results provide novel insight into pathobiology of GAE and its potential treatment in the future. This study was supported by TEAM TECH CORE FACILITY fin P: NGS platform for comprehensive diagnostics and personalized therapy in neuro-oncology.