Tmic-23. unravelling the heterogeneity of glioblastoma's vascular landscape

Abstract BACKGROUND Glioblastoma tumours are highly vascular and angiogenic. However, they are also heterogeneous, comprising many different macro- and microenvironments. Reflecting this, tumour-associated blood vessels are variable in their distribution, structure, and function. Despite the significance of vasculature in the development and progression of glioblastoma, its complexity remains poorly defined. Therefore, characterising vascular diversity across the glioblastoma landscape is a crucial step in the development of targeted therapies. METHODS 40 glioblastoma and 5 epilepsy cases were immunohistochemically labelled with a vascular antibody panel (PDGFRβ, ColIV, αSMA, UEA-1 and GFAP) to identify and subtype vessels. Single-vessel image analysis was performed to analyse vascular coverage and to classify vascular subtypes based on marker co-expression. Labelling with haemoglobin, ZO-1, and claudin-5 allowed for the evaluation of vascular leakage and blood-brain barrier integrity. Tissue segmentation was performed to analyse vascular coverage, vascular subtypes, and vascular integrity across the tumour macro-environment. RESULTS No significant difference in gross vascular coverage was found between tumour and epilepsy tissue. However, vascular coverage within glioblastoma varied significantly depending on the tumour macro-environment. Whilst vascular coverage in the tumour periphery was comparable to epilepsy tissue, increased vessel density was observed in the intermediate tumour zone, followed by decreased vessel density towards the tumour core. Reflecting this variation, multiple vessel subtypes, including previously described glomerular microvascular proliferation, mother vessels, vascular malformations and hyperplastic vessels, were identified. These tumour vessels display highly irregular shapes, sizes, and cellular compositions. Finally, glioblastoma tissue displayed a significant increase in haemoglobin leakage and a decrease in vascular integrity compared to epilepsy tissue. CONCLUSION Characterising this vascular landscape provides valuable insights for treatment strategies related to hypoxia and angiogenesis, two key factors driving tumour growth. Indeed, targeting specific vessel subtypes to address compromised vascular function and aberrant vascularisation may improve outcomes for glioblastoma patients.