720 - BJS Prize entrant: society of British Neurological Surgeons (SBNS) prize winner: axonal injury is a targetable driver of glioblastoma progression

Abstract Background Glioblastoma (GBM) is the commonest primary brain malignancy which, despite increasing understanding of advanced disease, remains incurable. A lack of relevant disease models has meant that the mechanisms driving earlier stages of GBM development and progression remain poorly understood. This is a key gap in knowledge as it opens new avenues for treatment and disease control of earlier grade disease. Somatic mouse models developed in our lab have now allowed us to explore the earliest phases of disease progression. Here, we explore what changes in the local tumour microenvironment (TME) drive gliomagenesis, and identify axonal injury as a key driver of GBM progression. We find that a programmed form of axonal injury is triggered by infiltrating tumour cells exerting mechanical strain on axons, triggering Sarm1 dependent axonal degeneration. We show that interference with this pathway leads to slowed disease progression, demonstrating the therapeutic potential of targeting axonal death in gliomagenesis. Material and Methods We combine a murine somatic GBM model with PDX models and patient tissue datasets. We use immunohistochemistry, spatial transcriptomics, single cell RNA sequencing and flow cytometry, as well as behavioural and survival studies to explore tumour development and progression in wild type (WT) and Sarm1−/− mice—which are resistant to Wallerian degeneration. Results Time-course analysis of tumourigenesis showed that early GBM preferentially infiltrates the white matter. Spatial transcriptomics identified axonal loss as an early event which strongly correlates with the level of tumour cell infiltration and is accompanied by neuroinflammation. This was confirmed at the tissue level using immunohistochemistry and electron microscopy. Axonal injury was shown to drive tumour cell proliferation and neuroinflammation through injury experiments where a transection of corpus callosum axons accelerated tumourigenesis. Conversely, blocking Sarm1 dependent axonal degeneration using a Sarm1−/− mouse model led to a marked delay in GBM development. Consistently, terminal Sarm1−/− tumours were less dense, contained more immature tumour cells, and were marked as lower grade on histopathological assessment. Strikingly, Sarm1−/− mice had a significantly prolonged survival, and behavioural studies revealed that Sarm1−/− tumour-bearing mice experienced less neurological decline than WT mice. Conclusions Our work provides insights into GBM development, identifying axonal injury as a key tumour promoting event. This is elicited by early tumour infiltration into the white matter and orchestrated by Sarm1 dependent programmed axonal degeneration, a druggable pathway already in clinical trials for neurodegenerative diseases. Hence, this is a potential novel therapeutic target for GBM providing a clinical opportunity for disease control.