VIRTUAL PHYSIOLOGY MODELING OF BREAST TUMOR TISSUE FOR MALIGNANCY ASSESSMENT

Abstract

To understand breast cancer dynamics, our study probes the biomechanical landscape of the tumor microenvironment (TME) using a novel in silico approach, which mirrors the complex interplay within the TME and its influence on tumor malignancy. We examined eight biopsy samples showcasing diverse breast cancer pathologies, including hyperplasia, invasive carcinoma, invasive lobular carcinoma, and normal breast tissue, were analyzed via Fourier Transform Infrared (FTIR) Imaging. Through molecular dynamics with LAMMPS, each sample was segmented into six specific tissue types, cancerous epithelium, noncancerous epithelium, dense stroma, loose stroma, reactive stroma, and other (modeled as red blood cells), allowing precise simulation of their mechanical properties. Young's Modulus values obtained for each specific tissue types informed the computational models, accurately replicating tissue stiffness. Simulations revealed distinct mechanical stress profiles across eight issue samples replicating tumor tissues at different states of malignancy. Notably, noncancerous epithelial tissues consistently exhibited the highest VMS values, followed by reactive stroma as the second highest, highlighting their susceptibility to mechanical forces. Additionally, the absence of dense stroma in certain samples correlated with increased stress in cancerous tissues, while loose stroma and "other" tissue types experienced the lowest VMS values. By bridging patient-specific TME images and mechanics with tumor behavior, our virtual TME models explain the critical role of mechanical stress in cancer progression, paving the way for tailored therapeutic strategies and advancing personalized medicine.