P02.14.B SPATIAL TRANSCRIPTOMIC ARCHITECTURE OF TUMOURS OF THE PERIPHERAL NERVOUS SYSTEM

Abstract

Abstract BACKGROUND Studies into the genetics of peripheral nerve tumors have been focused on their association with known gene mutations and syndromes as well as on the tumor microenvironment of malignant peripheral nerve sheath tumors. Still, not many advances have been made into clarifying the structure and diverse pathophysiology of these tumors. Here we present spatially resolved transcriptomic profiling of benign and malignant entities and describe the topographical architecture of the neoplastic and microenvironmental diversity. MATERIAL AND METHODS We employed spatially resolved transcriptomics using the 10X Visium technology from paraffine embedded specimens. We included hybrid peripheral nerve sheath tumors (HPNSTs) plexiform neurofibromas and malignant peripheral nerve sheath tumors (MPNST). Computational analysis was employed by the SPATA2 toolbox. Transcriptomic niches were identified by graph-based spatial clustering. Correlation of histomorphological features and transcriptomic programs was performed by a co-expression network-based analysis. RESULTS We identified hub genes associated with neurofibroma and schwannoma histopathology across primary pathologies and discovered that neurofibroma regions of HPNSTs exhibited unique transcriptional programs marked by a significant higher local accumulation of bone-derived macrophages and lymphoid cells compared to primary neurofibromas. A novel marker, Apolipoprotein D (APOD), was found to accurately identify neurofibroma subregions in HPNSTs, which was confirmed by immunostaining in a control cohort. In MPNSTs, we observed a high grade of inter- and intrapatient heterogeneity driven by unique transcriptional niches and genetic variability. Subsequent pathway analysis revealed distinct oncogenic signaling driver across regional niches along with variance in immune infiltration. Our findings highlight the disbalance between histological annotation and actual tumor cell abundance in MPNSTs indicated by inferred copy number alterations. CONCLUSION Our integrated analysis pipeline revealed distinct transcriptional signatures and cellular interactions within benign PNSTs, including neurofibromas and hybrid peripheral nerve sheath tumors (HPNSTs) and sought to expose the tumor heterogeneity of MPNST and identify various cancer niches. Our study underscores the potential of spatially resolved transcriptomics as a powerful tool for advancing our understanding of PNST biology, and ultimately improving diagnosis, prognosis, and treatment strategies.