A Mechanistic Lumped Parameter Model of the Berlin Heart EXCOR to Analyze Device Performance and Physiologic Interactions

摘要

Abstract Purpose The Berlin Heart EXCOR (BH) is the only FDA-approved, extracorporeal pulsatile ventricular assist device (VAD) for infants and children with heart failure. Clinicians control four settings on the device—systolic and diastolic drive pressures, device pump rate, and systolic time as a percentage of the pump cycle. However, interactions between BH pneumatics and the native circulation remain poorly understood. Thus, establishing appropriate device size and settings can be challenging on a patient-to-patient basis.Methods In this study we develop a novel lumped parameter network based on simplified device mechanics. We perform parametric studies to characterize device behavior, study interactions between the left ventricle (LV) and BH across different device settings, and develop patient-specific simulations. We then simulate the impact of changing device parameters for each of three patients.Results Increasing systolic pressure and systolic time increased device output. We identified previously unobserved cycle-to-cycle variations in LV–BH interactions that may impact patient health. Patient-specific simulations demonstrated the model’s ability to replicate BH performance, captured trends in LV behavior after device implantation, and emphasized the importance of device rate and volume in optimizing BH efficiency.Conclusion We present a novel, mechanistic model that can be readily adjusted to study a wide range of device settings and clinical scenarios. Physiologic interactions between the BH and the native LV produced large variability in cardiac loading. Our findings showed that operating the BH at a device rate greater than the patient’s native heart decreases variability in physiological interactions between the BH and LV, increasing cardiac offloading while maintaining cardiac output. Device rates that are close to the resting heart rate may result in unfavorable cardiac loading conditions. Our work demonstrates the utility of our model to investigate BH performance for patient-specific physiologies.