Establishing An In Silico Model To Predict The Effect Of Ivabradine On The Probability Of Coronary Plaque Rupture

BackgroundAcute coronary events that are mostly responsible for mortality in patients with ischemic heart disease, including unstable angina, myocardial infarction and sudden death, are usually caused by plaque rupture and consequent thrombosis. Beside the implication of high heart rate in atherogenesis and early atherosclerosis, its role throughout the natural history of atherosclerosis, culminating to the formation of vulnerable atherosclerotic plaque prone to rupture, is also critical. Thus, heart rate is an important therapeutic target and heart rate inhibitors may lower acute coronary events through lessening the likelihood of rupture. Ivabradine specifically inhibits the I(f) current in the sinoatrial node to lower heart rate, without affecting other aspects of cardiac function. As it is difficult to obtain the dynamic composition of coronary plaque and surface stress concentration in vivo, and no representative animal model of plaque rupture, we seek the computer simulation to achieve predicting the impact of ivabradine on plaque rupture.AimThe purpose of our work is to build an in silico model to predict the effect of ivabradine, on the probability of plaque rupture through its exclusive pure heart rate reduction.MethodCoronary plaque rupture depends on both the composition of the plaque and extrinsic triggers that may precipitate plaque disruption in a vulnerable plaque. Elevated heart rate produced increased coronary flow velocity, therefore, stressed on plaque surface. Fatigue, a chronic failure process induced by repetitive loading, might play a role in plaque rupture. Fatigue depended on heart rate. We assumed shear stress on the plaque surface was related to plaque rupture. Whole model was consisted of two sub-models:(1) coronary flow fluid model of left descending artery:to calculate the shear stress at shoulder of plaque under different heart rate and stenosis. (2) in silico model to predict the probability of plaque rupture:we postulated that plaque would rupture when instaneous stress exceeded critical stress. Instaneous stress was mean shear stress per day. Critical stress was the decreasing time function of fatigue, the composition of the plaque (lipid core, fibrous cap thickness and high sensitive C protein), and patient characteristics (age, sex, total cholesterol and body weight index).Results(1) Fluid model demonstrated that shear stress was positively proportional to heart rate no whether the severity of stenosis. Elevated heart rate attributed the shortening of the fatigue life of plaque, impairing the plaque stability.(2) In silico model:Plaque rupture depended on heart rate, lipid core, fibrous cap thickness, high sensitive C protein, and patient characteristics (age, sex, total cholesterol and body weight index). Decreased heart rate produced by ivabradine could postpone the time of plaque rupture. When heart rate decreased by 8 beats per minute in one simulated patient population, plaque rupture would decrease about 2% per year. Patient with thin fibrous cap thickness and rich lipid core were at very high risk of plaque rupture. But when cap thickness was relatively thick, plaque was not prone to rupture. Male patient would get more benefit from heart rate decreasing strategy than female patients.DiscussionIvabradine could decrease the probability of plaque rupture by decreasing heart rate and fatigue, thus acute coronary events. Clinically, the treatment focus should also emphasized on the heart rate reduction. This is the first time to apply computation simulation technology to predict the probability of plaque rupture, which was difficult to investigate in vivo. This model provided a new approach and technology platform to help us whole understanding the mechanism of plaque rupture. Also, this new technology would save the limited resources, shorten the research time. However, the mechanism of plaque rupture remains elusive. Also, other factors influencing plaque rupture still were not included in this model. This model need to be further improved and validated.