A comparison of two quasi‐static computational models for assessment of intra‐myocardial injection as a therapeutic strategy for heart failure

Yiling, Fan, William, Ronan, Irvin, Teh, Jurgen E., Schneider, Claudia E., Varela, William, Whyte, Peter, McHugh, Sean, Leen, Ellen, Roche

International Journal for Numerical Methods in Biomedical Engineering |

Myocardial infarction, or heart attack, is the leading cause of mortality globally. Although the treatment of myocardial infarct has improved significantly, scar tissue that persists can often lead to increased stress and adverse remodeling of surrounding tissue, and ultimately to heart failure. Intra-myocardial injection of biomaterials represents a potential treatment to attenuate remodeling, mitigate degeneration and reverse the disease process in the tissue. In vivo experiments on animal models have shown functional benefits of this therapeutic strategy. However, a poor understanding of the optimal injection pattern, volume and material properties has acted as a barrier to its widespread clinical adoption. In this study, we developed two quasi-static finite element simulations of the left ventricle to investigate the mechanical effect of intra-myocardial injection. The first model employed idealized left ventricular geometry with rule-based cardiomyocyte orientation. The second model employed subject-specific left ventricular geometry with cardiomyocyte orientation from diffusion tensor magnetic resonance imaging. Both models predicted cardiac parameters including ejection fraction, systolic wall thickening, and ventricular twist that matched experimentally reported values. All injection simulations showed cardiomyocyte stress attenuation, offering an explanation for the mechanical reinforcement benefit associated with injection. The study also enabled a comparison of injection location, and the corresponding effect on cardiac performance at different stages of the cardiac cycle. While the idealized model has lower fidelity, it predicts cardiac function and differentiates the effects of injection location. Both models represent versatile in silico tools to guide optimal strategy in terms of injection number, volume, site, and material properties.