Integration of epidemiology, pathology, immunology and outcomes in colorectal cancer - ABSTRACT Advances in epidemiology, pathology and immunology offer new insights into the diagnosis and treatment of colorectal cancer (CRC). For CRC, accurate pathologic diagnosis is essential in the selection of appropriate therapy and is becoming more complex, especially with the explosion of newly introduced prognostic and predictive biomarkers. Machine learning has the potential to transform pathologic diagnosis and to address very limited accessibility of expert pathology in low-income countries. Routine histology images of solid tumors contain an immense number of visual features that can be extracted and processed by artificial intelligence tools like machine learning, which can also predict clinically relevant features including microsatellite instability and specific immune features. This large study uses images of whole slides and tissue microarrays from CRC cases, and applies computational pathology methods and digital spatial expression profiling to quantifiably improve CRC diagnosis, prognosis and predictive models augmented by clinical, epidemiologic and genetic data. The study goals will be accomplished through three specific aims. In Aim 1, we will apply novel machine learning algorithms from whole slide images to reproducibly identify MSI, histopathologic, genomic and immune features of colorectal cancer in representative populations. We will study H&E slides from 6,751 CRC cases, digitizing existing slides from 5,551 CRC cases and 1,200 new cases of CRC with contemporaneous clinical and epidemiologic data. Then, we will apply deep learning methods to accurately identify histopathologic features and immune characteristics of CRC. We will use a robust training validation, and testing design (70%/15%/15%) to ensure the rigor and reproducibility of our findings. In Aim 2, we will test whether clinical, epidemiologic, and germline genetic data significantly contribute beyond digital pathology machine learning algorithms to improve: a) prognostic models of overall and disease-specific survival, and b) predictive models of response to therapy. We will use machine learning statistical methods to test whether algorithms developed in Aim 1 improve prediction of overall survival and response to therapy with the addition of supplemental information beyond whole slide digital images. Finally, in Aim 3, we will compare the information derived from digital spatial profiling of expressed proteins in CRC with information derived from Immunoscore measures of lymphocyte populations at the tumor center and the invasive margin and explore whether these measures improve models developed in Aims 1 and 2 in a subset of samples. We will compare GeoMx digital spatial profiling to Immunoscore, a scoring system relying exclusively on expression patterns of CD3+ and CD8+ T cells. This study takes advantage of pathologic, epidemiologic, molecular, clinical, and germline genetic data from representative CRC patients from California, Detroit, New York, Florida, Puerto Rico, Israel and Spain. Our overarching goal is to shift the paradigm of how CRC is diagnosed and molecularly characterized through histologic, genomic, and immune features derived from routinely collected images.
