pathoDISCO-HE: Towards light sheet microscopy enabled volumetric histopathology of human gliomas

Histopathological assessment of human tumours relies heavily upon traditional slide-based techniques such as H&E staining and immunohistochemistry. Whilst providing excellent resolution, traditional histological techniques require destruction of the specimen and do not provide details of three-dimensional tissue architecture. Glioblastomas are amongst the most spatially heterogeneous of human tumours, making them a prime target for the growing field of volumetric histopathology. Tissue clearing and light sheet microscopy are complementary technologies allowing high-resolution volumetric imaging of biological specimens. In this study we aimed to explore the possibility of using these techniques as tools for volumetric histopathology of human gliomas. Starting with the previously published pathoDISCO protocol, we added pigment bleaching and fluorescent nuclear labelling stages to develop a protocol capable of clearing glioma specimens up to several millimetres thick within a few days. Using cresyl violet as a nuclear label and autofluorescence as an eosin analogue we were able to produce realistic “virtual H&E” images using a simple image post-processing pipeline. Virtual H&E images accurately represented several key pathological features of gliomas including necrosis, hypercellularity, microvascular proliferations and glomeruloid and garland vascular structures. The virtual H&E images can be viewed three dimensionally with only limited loss of resolution in z-projections. The volumetric imaging data could also be visualised by maximum intensity projection, allowing tracing of vascular structures throughout the tissue and identification of discrete zones of differing cellular or vascular density. Finally, we showed that our virtual H&E protocol is reversible and post-hoc traditional histology can be applied if necessary. Overall, our study represents a novel contribution to the palette of volumetric histopathology techniques which may eventually allow more accurate and detailed characterisation of human tumours, thus facilitating more accurate diagnosis and prognosis and selection of the most effective personalised therapies, potentially improving patient outcomes.