A Fully Coupled Multi-Physics, Multi-Scale In-Silico Model of Coronary Atherosclerosis Progression: Bridging Hemodynamics, Tissue Mechanics, and Pathophysiology

Introduction: Herein we introduce a new multi-physics, multi-scale modeling approach to understand plaque progression during coronary artery disease. Prior works have coupled agent-based models (ABMs) with finite element analysis (FEA) or computational fluid dynamics (CFD) to study the individual contributions of tissue mechanics or hemodynamics to plaque growth, respectively [Figure 1]. However, these approaches were limited by their inability to simultaneously capture the dynamic interplay between all three domains that drive plaque development. This study aims to present a novel method that innovates prior approaches by merging hemodynamics via CFD, biological processes via ABM, and biomechanics via FEA into a single multi-scale, multi-physics simulation (CAFe). Moreover, this approach facilitates spatial and temporal continuity across the different physics domains by utilizing a volumetric tetrahedral geometry that is compatible with CFD, ABM, and FEA. Preliminary results suggest that the progression of coronary atherosclerosis can successfully be simulated utilizing this novel approach. Moreover, these results suggest that plaque enlargement is largely driven by the accumulation of macrophage derived foam cells, as observed in histological studies of advanced plaques. The innovative approach described herein, therefore, marks a significant step forward in accurately modeling the progression of coronary atherosclerosis, paving the way for more effective diagnostic and therapeutic strategies in cardiovascular disease management.

Materials and

Methods: The CAFe model is an in-silico framework designed to explore the mechanisms behind plaque growth in arteries. It integrates three interconnected modules: the Agent-Based Model (ABM), Finite Element Analysis (FEA), and Computational Fluid Dynamics (CFD). The ABM simulates biological processes at the cellular and tissue scale, the FEA calculates mechanical stresses and strains, and the CFD assesses hemodynamic wall shear stress (WSS). The model operates on a volumetric tetrahedral mesh representing vessel geometry, initialized with specific cellular and molecular conditions [Figure 2]. Initially, the FEA and CFD modules use FEBio software to determine baseline mechanical and hemodynamic states. The ABM, implemented in MATLAB, then processes these outputs to simulate biological activities such as cell proliferation, apoptosis, and cytokine dynamics, updating agent counts and locations. These updates modify the mesh geometry, prompting reruns of the FEA and CFD to adjust mechanical and hemodynamic values accordingly. This iterative process continues across time steps, with all modules sharing the same geometry to ensure consistency. The model accommodates different tissue types, including necrotic core, fibrotic, fibrofatty, calcium, and artery wall, using patient-specific clinical data. Changes in tissue composition and geometry are continuously updated in the ABM, influencing stiffness calculations in the FEA to reflect stress and strain changes throughout the mesh.

Results, Conclusions, and Discussions: Preliminary results using the CAFe model highlight the models capability to simulate remodeling paradigms observed in coronary artery atherosclerosis. Dynamic changes in various cellular, molecular, and biochemical factors over time were observed. Smooth muscle cells (SMC) show slight declines over via apoptosis, while fibroblasts, foam cells, and lymphocytes exhibit different trends, with lymphocytes significantly decreasing after an initial peak. Neutrophils, monocytes, and macrophages rapidly decrease, indicating a quick response followed by stabilization. Growth factors (PDGFAB and TGFB) and cytokines (TNFA, IL1, IL10) show varying patterns, with TNFA and IL1 decreasing over time, while IL10 initially increases before declining. Lipids (LDL and OxLDL) remain relatively stable after an initial adjustment. Matrix metalloproteinases (MMP1, MMP2, MMP9) and structural proteins (collagen and elastin) exhibit minimal changes, suggesting steady-state conditions in the extracellular matrix components. These results indicate that the CAFe model effectively captures the complex interactions and temporal dynamics of plaque development and progression.

Acknowledgements (Optional): American Heart Association (17SDG33400239). NHLBI of the NIH(R01HL136776). UTD Department of Bioengineering for startup funding. Vascular Mechanobiology Lab for the opportunity to work on this research through the guidance of Dr. Hayenga, the lab director.