Eco-evolutionary Guided Pathomics Analysis to Predict DCIS Upstaging

Area of Science:

• Oncology
• Cancer Biology
• Tumor Microenvironment Ecology

Background:

• Cancers evolve within dynamic ecosystems, necessitating characterization of their ecological dynamics for understanding evolution and discovering predictive biomarkers.
• Ductal carcinoma in situ (DCIS) is an early-stage breast cancer with abnormal epithelial cell growth confined to milk ducts, where progression prediction remains challenging.

Purpose of the Study:

• To investigate if ecological analysis of hypoxia and acidosis biomarkers can improve the prediction of DCIS upstaging.
• To develop and apply an eco-evolutionary approach to identify tumor microenvironment habitats and niches for biomarker discovery.

Main Methods:

• Developed an eco-evolutionary approach to define tumor habitats based on oxygen diffusion distance.
• Identified cancer cell metabolic phenotypes using biomarkers CA9 (hypoxia) and LAMP2b (acidosis).
• Analyzed spatial patterns of biomarkers to define distinct tumor niches and predict patient upstaging using a random forest classifier with 5-fold validation.

Main Results:

• Ecological analysis significantly enhanced the predictive power of hypoxia and acidosis biomarkers for DCIS upstaging compared to traditional methods.
• Distinct tumor niches characterized by specific biomarker spatial patterns were identified.
• A random forest classifier achieved an AUC of 0.74 in predicting patient upstaging based on niche distribution.

Conclusions:

• Tumor ecological features are crucial for eco-evolutionary-designed approaches in novel biomarker discovery.
• This study demonstrates the potential of ecological analysis of biomarkers to improve the prediction of DCIS progression.
• The findings highlight the importance of considering the tumor microenvironment's spatial and metabolic heterogeneity.