Eco-Evolutionary Guided Pathomic Analysis Detects Biomarkers to Predict Ductal Carcinoma In Situ Upstaging.

UNLABELLED

Cancers evolve in a dynamic ecosystem. Thus, characterizing the ecological dynamics of cancer is crucial to understanding cancer evolution, which can lead to the discovery of biomarkers to predict disease progression. Ductal carcinoma in situ (DCIS) is an early-stage breast cancer characterized by abnormal epithelial cell growth confined within the milk ducts, and biomarkers are needed to predict which cases will progress to aggressive disease. In this study, we showed that ecological analysis of hypoxia and acidosis biomarkers can significantly improve prediction of DCIS upstaging. Quantitative analyses were performed on immunohistologic images from a retrospective cohort of DCIS specimens collected from biopsy samples. First, an eco-evolutionary designed approach was developed to define habitats in the tumor intraductal microenvironment based on oxygen diffusion distance. Then, cancer cells with metabolic phenotypes attributed to their habitats were identified, including a hypoxia-responding CA9+ phenotype and an acid-adapted LAMP2b+ phenotype. Whereas these markers have traditionally shown limited, if any, predictive capabilities for DCIS progression when analyzed from an ecological perspective, their power to differentiate between non-upstaged and upstaged DCIS increased significantly. Additionally, the distribution of distinct niches with specific spatial patterns of these biomarkers predicted patient upstaging. The niches were characterized by pattern analysis of both cellular and spatial features. A random forest classifier that was trained and underwent a five-fold validation on the biopsy cohort achieved an AUC of 0.74 for predicting clinical outcome. These results affirm the importance of tumor ecological features in eco-evolutionary-designed approaches for biomarker discovery.

SIGNIFICANCE

Evolutionary dynamics of the various niches composing the tumor ecosystem can be harnessed for predicting cancer progression, demonstrating how eco-evolutionary-designed approaches can guide biomarkers discovery studies in the era of digital pathology. This article is part of a special series: Driving Cancer Discoveries with Computational Research, Data Science, and Machine Learning/AI.