EXPLORING THE IMMUNE LANDSCAPE OF THE TUMOUR MICROENVIRONMENT IN 3-D USING MULTI-MODAL IMAGING

Central nervous system tumours, including meningiomas and glioblastomas, exhibit highly heterogeneous microenvironments that contribute to therapeutic resistance and poor clinical outcomes. Conventional two-dimensional histological approaches provide limited insight into the spatial organization of these tumours, highlighting the need for three-dimensional (3D) methods that preserve intact tissue architecture.

Here, we present a scalable workflow combining tissue clearing with light sheet microscopy (LSM) to achieve cellular-resolution mapping of vascular and immune components within human CNS tumours. Freshly resected meningioma and glioblastoma specimens were cleared and immunolabelled for endothelial and immune cell markers, including CD31 (endothelial cells), CD3 (T cells), and Iba1 (myeloid cells), followed by LSM imaging.

A proof-of-concept segmentation and analysis pipeline was established using Imaris software, enabling quantitative assessment of key spatial parameters such as immune cell abundance, spatial distribution, and proximity to blood vessels. Marker validation and segmentation were performed across meningiomas grades I–III and extended to glioblastoma samples using an expanded immune marker panel, including macrophage and microglia markers. In ongoing studies, we are integrating macroscopic pre- and postoperative magnetic resonance imaging (MRI) by co-registering high-field MRI data with LSM datasets to establish cellular-level ground truth for radiological features.

Together, these results demonstrate the feasibility of a robust 3D imaging and analysis platform for characterizing complex tumour microenvironments in human brain tumours. This approach overcomes the limitations of conventional 2D analyses and provides a foundation for future integration with spatial transcriptomics, enabling comprehensive investigation of tumour heterogeneity, immune–tumour interactions, and spatial biomarkers relevant to therapeutic targeting.