In-stent restenosis represents a major cause of failure of percutaneous coronary intervention with drug-eluting stent implantation. Computational multiscale models have recently emerged as powerful tools for investigating the mechanobiological mechanisms underlying vascular adaptation processes during in-stent restenosis. However, to date, the interplay between intervention-induced inflammation, drug delivery and drug retention has been under-investigated. Here, an original patient-specific multiscale agent-based modelling framework was developed to investigate the interplay between drug release, plaque composition and intervention-induced inflammation on in-stent restenosis following drug-eluting stent implantation. The framework integrated a finite element simulation of stent expansion, with a drug transport simulation and an agent-based model of cellular dynamics. A patient-specific coronary cross-section with heterogeneous diseased tissue was considered and rigorously analyzed through a variety of scenarios, including different plaque compositions and different inflammatory responses. The analysis revealed three significant findings: (i) calcifications substantially impeded drug transport, resulting in drug-depleted regions and reduced stent efficacy; (ii) by impacting drug transport, variations in plaque composition influenced arterial wall response, with the fully-calcific scenario showing the greatest lumen area reduction; (iii) the impact of different drug receptor saturation conditions (obtained with different plaque compositions) was particularly evident under conditions of persistent inflammatory state. This study represents a significant advancement in multiscale modelling of in-stent restenosis following drug-eluting stent implantation. The results obtained provided deeper insights into the complex interactions among patient-specific plaque composition, inflammation and drug retention, suggesting a patient-specific management of the intervention, particularly in cases of complex disease.

Atherosclerotic plaque rupture in coronary arteries, an important trigger of myocardial infarction, is shown to correlate with high levels of pressure-induced mechanical stresses in plaques. Finite element (FE) analyses are commonly used for plaque stress assessment. However, the required information of heterogenous material properties of atherosclerotic coronaries remains to be scarce. In this work, we characterized the component-wise mechanical properties of atherosclerotic human coronary arteries. To achieve this, we performed ex vivo inflation tests on post-mortem human coronary arteries and developed an inverse FE modeling (iFEM) pipeline, which combined high-frequency ultrasound deformation measurements, a high-field magnetic resonance-based artery composition characterization, and a machine learning-based Bayesian optimization (BO) with uniqueness assessment. By using the developed pipeline, 10 cross-sections from five atherosclerotic human coronary arteries were analyzed, and the Yeoh material model constants of the fibrous intima and arterial wall components were determined. This work outlines the developed pipeline and provides the knowledge of non-linear, multicomponent mechanical properties of atherosclerotic human coronary arteries.