TMIC-41. LATERAL VENTRICLE PROXIMITY OF GLIOBLASTOMA IS ASSOCIATED WITH LOCALIZED MALIGNANCY-PROMOTING MOLECULAR SIGNATURES IN IMAGE-GUIDED BIOPSIES

Abstract

Abstract Glioblastoma (GBM) is the most devastating and common form of primary brain cancer in adults. Tumor location plays a significant role in patient prognosis; in particular, GBM tumors contacting the lateral ventricles (LVs) are more aggressive than LV-distal counterparts. This may be due to interaction of tumors with unique features of the LV microenvironment, including the cerebrospinal fluid (CSF) and neural progenitor cells (NPCs) of the subventricular zone. Despite LV contact having a role on tumor malignancy, studies using bulk tumor samples to identify molecular contributors to LV-contacting prognosis have been only moderately successful, in large part due to intratumoral heterogeneity. Here, we leverage intraoperative surgical navigation to collect matched biopsies of adult LV-contacting primary GBM at locations proximal (< 2 cm) and distal ( > 2 cm) from the LV. Matched samples were collected from ten patients with equal sex distribution and an age range of 46 to 81 years. RNA sequencing of biopsies revealed a localized transcriptional signature of gene expression in LV-proximal locations. These genes include those identified to be associated with GBM malignancy and upregulated in primary GBM cells by CSF exposure or co-culture with NPCs, such as SERPINA3, CD44, and CTSB, as well as many newly identified genes. Implicated biological pathways involved in this transcriptional signature involve inflammation, chemotaxis, Notch signaling, angiogenesis, integrin signaling, and more. The top 50 upregulated genes by LV-proximity of GBM were used to cluster patient transcriptomes from the TCGA GBM database and evaluate the contribution to patient outcome. This analysis revealed a cluster with significantly extended overall survival that displayed a general downregulation of LV-proximity genes. Ultimately, these studies provide evidence that regional tumoral analysis is essential in identifying molecular signatures associated with microenvironmental influence, and that LV-contacting GBM has regional gene expression changes in many genes associated with malignancy.