Abstract
While it is known that glioblastoma (GBM) intratumoral heterogeneity contributes to therapeutic resistance, how this extends to spatial heterogeneity is less clear. Adjuvant treatment post-surgical resection is informed by analysis of the resected enhancing tumor but is expected to target residual non-enhancing tumor. To assess intratumoral spatial heterogeneity, we examined multi-region samples taken from 18 GBM patients undergoing resection. Tissue was collected from the central core, enhancing rim, and non-enhancing rim of the tumor using a stereotactically-navigated handpiece and automated preservation system. Samples were flash frozen and underwent bulk RNA sequencing, whole exome sequencing, and single nucleus RNA sequencing. Paired-end reads were aligned to the reference human genome. Raw read counts were estimated and differential expression and gene ontology analysis was performed along with detection of disease-causing mutation variants. The single nucleus RNA sequencing is in process and data will be available soon. The bulk RNA-Seq data demonstrates different gene ontologies between non-enhancing rim and the remaining tumor. The tumor is enriched in nuclear division and vasculature development pathways, while the non-enhancing rim is enriched in gliogenesis and myelination pathways. The non-enhancing rim is enriched for pathogenic variants in RGS10, involved in G-protein signaling. Nine driver genes were found to be preferentially mutated across different regions, including KMT2C, a known tumor suppressor, mutated in 24% of non-enhancing rim samples but in only 6% of the remaining regions. These results highlight the degree of intratumoral heterogeneity that extends to distinct anatomical regions and will be further delineated using single nucleus RNA sequencing. Our findings describe the spatial heterogeneity present between the enhancing and non-enhancing regions in GBM, which can be studied to inform targeted adjuvant therapies.