Relation of Quantitative Histologic and Radiologic Breast Tissue Composition Metrics With Invasive Breast Cancer Risk

Area of Science:

• Oncology
• Pathology
• Radiology
• Biomarkers
• Machine Learning

Background:

• Benign breast disease (BBD) is a significant risk factor for developing invasive breast cancer.
• Accurate identification of BBD patients at high risk for future breast cancer remains a clinical challenge.
• Novel biomarkers are needed to stratify risk among BBD patients.

Purpose of the Study:

• To investigate the association between quantitative tissue composition metrics from BBD biopsies and mammograms with future invasive breast cancer risk.
• To identify predictive biomarkers for breast cancer development in women with BBD.

Main Methods:

• A case-control study nested within a cohort of 15,395 BBD patients was conducted.
• Machine learning was applied to digitized H&E-stained biopsies to quantify tissue composition.
• Computer-assisted thresholding of mammograms determined percent mammographic density (MBD).
• Quantitative metrics were analyzed for association with incident invasive breast cancer diagnosis.

Main Results:

• Increased epithelial area on BBD biopsies correlated with higher breast cancer risk (OR_{Q4 vs Q1} = 1.85).
• Higher epithelium-to-stroma proportion (OR_{Q4 vs Q1} = 2.06) and percent mammographic density (MBD) (OR_{Q4 vs Q1} = 2.20) were independently associated with increased risk.
• The combination of high epithelium-to-stroma proportion and high MBD significantly elevated breast cancer risk (OR = 2.27), particularly in nonproliferative BBD.

Conclusions:

• Epithelium-to-stroma proportion from BBD biopsies and MBD are independent predictors of future invasive breast cancer.
• These quantitative metrics, especially in combination, offer valuable risk stratification for BBD patients.
• Findings are particularly relevant for nonproliferative BBD, where predictive biomarkers are currently lacking.