Spatial Radiogenomics of Ovarian Cancer: Implementation of a Lesion-specific 3D-printed Mould Pipeline in the Clinical Workflow for Image-guided Tissue Multi-sampling of Ovarian Tumours

The biological spatial and temporal heterogeneity of High Grade Serous Ovarian Carcinoma (HGSOC) severely impacts the effectiveness of therapies and is a determinant of poor outcomes.

Current histological evaluation is made on a single tumour sample from a single disease site per patient thus ignoring molecular heterogeneity at the whole-tumour level, key for understanding and overcoming chemotherapy resistance. Imaging can play a crucial role in the development of personalised treatments by fully capturing the disease's heterogeneity.

Radiomics quantify the image information by capturing complex patterns related to the tissue microstructure. This information can be complemented with clinical data, liquid biopsies, histological markers and genomics ("radiogenomics") potentially leading to a better prediction of treatment response and outcome. However, the extracted quantitative features usually represent the entire tumour, ignoring the spatial context.

On the other hand, radiomics-derived imaging habitats characterize morphologically distinct tumour areas and are more appropriate for monitoring the changes in the tumour microenvironment over the course of therapy. In order to successfully incorporate the habitat-imaging approach to the clinic, histological and biological validation are crucial. However, histological validation of imaging is not a trivial task, due to issues such as unmatched spatial resolution, tissue deformations, lack of landmarks and imprecise cutting. Patient-specific three-dimensional (3D) moulds are an innovative tool for accurate co-registration between imaging and histology. The aim of this study is to optimize and integrate such an automated computational 3D-mould co-registration approach in the clinical work-flow in patients with HGSOC. The validated radiomics-based tumour habitats will also be used to guide tissue sampling to decipher their underlying biology using genomics analysis and explore novel prediction markers.