Abstract PO-043: Combining multiplexed immunohistochemistry and deep learning to spatially map the tumor microenvironment

Abstract

Abstract The tumour microenvironment(TME) is a highly complex mixture containing epithelium, stroma and a diverse network of immune cells and the spatial organization of these immune cells within the TME reflects a crucial process in anti-tumor immunity. The usual standard of care for assessing if a patient has cancer, its stage and its likely future biological behaviour is visual examination of one or more H&E and/or Immunohistochemistry(IHC) stained sections. The paradigm of digital pathology has changed, moving from single-marker IHC towards multiplexed labeling, increasing the need for more advanced techniques that can be easily integrated in routine clinical pathology. Although recent advances in multiple immunostaining have enabled characterization of several parameters on a single tissue section. For a higher dimensional chromogen based methodology, we have developed a multiplexed IHC procedure combining multiple labels per round with several sequential rounds, where multiple is 12-14 chromogen based antibodies on a single tissue section. Thus, enabling analysis of complex immune cell population’s on a single slide through consecutive cycles of staining, destaining, hyperspectral imaging and spectral unmixing of the chromogen biomarkers in each round. The process presented is using absorption microscopy, enabling these images to be done in a reasonable time frame. Not only does this allow us to completely visualize and spatially map the TME, but through the application of AI we can further recognize common histological features. Stoichiometric DNA following mIHC, improves nuclei identification so we can identify various cell populations present in the tissue. Thus, providing us with an accurate spatial cell level representation of any tissue section using only a single staining process. Although there have been several uprising technologies which characterize the complexities of the TME, our process is capable of doing so in a shorter time frame and at a reduced cost using absorption based methods. Robust, accurate, segmentation of cell nuclei for overlapping nuclei is one of the most significant unsolved issues in digital pathology. By combining a multiplexed IHC technique which enables the detection of multiple markers on a single slide with deep learning segmentation methods to segment every individual cell nuclei in tissue sections with an accuracy comparable to human annotation. We can analyze the cell-cell interactions between immune and tumour cells and identify clinically relevant patterns that may improve patients outcome by informing on the likelihood of success of possible treatments. These two techniques joined can be scaled up to the entire tissue section level, improving our understanding of the biological aggressiveness of specific cancers and enabling an accurate spatial cell level representation of the tissue. Citation Format: Kouther Noureddine, Paul Gallagher, Martial Guillaud, Calum MacAulay. Combining multiplexed immunohistochemistry and deep learning to spatially map the tumor microenvironment [abstract]. In: Proceedings of the AACR Virtual Special Conference on Artificial Intelligence, Diagnosis, and Imaging; 2021 Jan 13-14. Philadelphia (PA): AACR; Clin Cancer Res 2021;27(5_Suppl):Abstract nr PO-043.