Use of a comprehensive numerical model to improve biventricular pacemaker temporization in patients affected by heart failure undergoing to CRT-D therapy

Virtual and Artificial Cardiorespiratory Patients in Medicine and Biomedical Engineering

Recently, ‘medicine in silico’ has been strongly encouraged due to ethical and legal limitations related to animal experiments and investigations conducted on patients. Computer models, particularly the very complex ones (virtual patients—VP), can be used in medical education and biomedical research as well as in clinical applications. Simpler patient-specific models may aid medical procedures. However, computer models are unfit for medical devices testing. Hybrid (i.e., numerical–physical) models do not have this disadvantage. In this review, the chosen approach to the cardiovascular system and/or respiratory system modeling was discussed with particular emphasis given to the hybrid cardiopulmonary simulator (the artificial patient), that was elaborated by the authors. The VP is useful in the education of forced spirometry, investigations of cardiopulmonary interactions (including gas exchange) and its influence on pulmonary resistance during artificial ventilation, and explanation of phenomena observed during thoracentesis. The artificial patient is useful, inter alia, in staff training and education, investigations of cardiorespiratory support and the testing of several medical devices, such as ventricular assist devices and a membrane-based artificial heart.

Sensitivity Analysis of Single Beat Left Ventricular Elastance Estimation by Chen Method

INTRODUCTION

Left ventricular (LV) end-systolic elastance (Ees) can be estimated using single-beat (Ees(sb)) Chen method, employing systolic and diastolic arm-cuff pressures, stroke volume (SV), ejection fraction and estimated normalized ventricular elastance at arterial end-diastole. This work aims to conduct a sensitivity analysis of Chen formula to verify its reliability and applicability in clinical scenario.

METHODS

Starting from a baseline condition, we evaluated the sensitivity of Ees(sb) to the parameters contained in the formula. Moreover, a mathematical model of the cardiovascular system was used to evaluate the sensitivity of Ees(sb) to end-diastolic LV elastance (Eed), Ees, arterial systemic resistance (Ras) and heart rate (HR).

RESULTS

In accordance with Ees definition, Ees(sb) increases by increasing aortic pressure and pre-ejection time, reaching the highest value for a pre-ejection time = 40 ms, and then decreases. In contrast with Ees definition, Ees(sb) increases (from 3.21 mmHg/mL to 12.15 mmHg/mL) by increasing the LV end-systolic volume and decreases by increasing the SV. In the majority of the analysis with the mathematical model, Ees was underestimated using the Chen method: by increasing Ees (from 0.5 to 2.5 mmHg/mL), Ees(sb) passes only from 0.56 to 1.54 mmHg/mL. Ees(sb) increases for higher Eed (from 1.03 to 2.33 mmHg/mL). Finally, Ees(sb) decreases (increases) for HR < 50 bpm (< 50 bpm), and for Ras < 1100 mmHg/gcm⁴ (> 1100 mmHg/gcm⁴).

CONCLUSION

Unexpectedly Ees(sb) increases for higher LV end-systolic volume and decreases for higher SV. These results contrast with Ees definition, which is the ratio between the LV end-systolic pressure and the LV end-systolic volume. Moreover, Ees(sb) is influenced by cardiocirculatory parameters such as LV Eed, HR, Ras, ejection time, and pre-ejection time. Finally, Ees(sb) computed with the model output often underestimates model Ees.

Towards a Personalized and Dynamic CRT-D

Background: In spite of cardiac resynchronization therapy (CRT) benefits, 25 – 30% of patients are still non responders. One of the possible reasons could be the non optimal atrioventricular (AV) and interventricular (VV) intervals settings. Our aim was to exploit a numerical model of cardiovascular system for AV and VV intervals optimization in CRT.

Methods: A numerical model of the cardiovascular system CRT-dedicated was previously developed. Echocardiographic parameters, Systemic aortic pressure and ECG were collected in 20 consecutive patients before and after CRT. Patient data were simulated by the model that was used to optimize and set into the device the intervals at the baseline and at the follow up. The optimal AV and VV intervals were chosen to optimize the simulated selected variable/s on the base of both echocardiographic and electrocardiographic parameters.

Results: Intervals were different for each patient and in most cases, they changed at follow up. The model can well reproduce clinical data as verified with Bland Altman analysis and T-test (p > 0.05). Left ventricular remodeling was 38.7% and left ventricular ejection fraction increasing was 11% against the 15% and 6% reported in literature, respectively.

Conclusions: The developed numerical model could reproduce patients conditions at the baseline and at the follow up including the CRT effects. The model could be used to optimize AV and VV intervals at the baseline and at the follow up realizing a personalized and dynamic CRT. A patient tailored CRT could improve patients outcome in comparison to literature data.

Concomitant Pulsatile and Continuous Flow VAD in Biventricular and Univentricular Physiology: A Comparison Study with a Numerical Model

Introduction

To develop and test a lumped parameter model to simulate and compare the effects of the simultaneous use of continuous flow (CF) and pulsatile flow (PF) ventricular assist devices (VADs) to assist biventricular circulation vs. single ventricle circulation in pediatrics.

Methods

Baseline data of 5 patients with biventricular circulation eligible for LVAD and of 5 patients with Fontan physiology were retrospectively collected and used to simulate patient baselines. Then, for each patient the following simulations were performed: (a) CF VAD to assist the left ventricle (single ventricle) + a PF VAD to assist the right ventricle (cavo-pulmonary connection) (LCF + RPF); (b) PF VAD to assist the left ventricle (single ventricle) + a CF VAD to assist the right ventricle (cavo-pulmonary connection) (RCF + LPF)

Results

In biventricular circulation, the following results were found: cardiac output (17% RCF + LPF, 21% LCF + RPF), artero-ventricular coupling (-36% for the left ventricle and -21.6% for the right ventricle), pulsatility index (+6.4% RCF + LPF, p = 0.02; -8.5% LCF + RPF, p = 0.00009). Right (left) atrial pressure and right (left) ventricular volumes are decreased by the RCF + LPF (by RPF + LCF). Pulmonary arterial pressure decreases in the LCF + RPF configuration. In Fontan physiology: cardiac output (LCF + RPF 35% vs. 8% in RCF + LPF), ventricular preload (+4% RCF + LPF, -10% LCF + RPF), Fontan conduit pressure (-5% RCF + LPF, +7% LCF + RPF), artero-ventricular coupling (-14% RCF + LPF vs. -41% LCF + RPF) and pulsatility (+13% RCF + LPF, - 8% LCF + RPF).

Conclusions

A numerical model supports clinicians in defining and innovating the VAD implantation strategy to maximize the hemodynamic benefits. Results suggest that the hemodynamic benefits are maximized by the LCF + RPF configuration.

Application of a Lumped Parameter Model to Study the Feasibility of Simultaneous Implantation of a Continuous Flow Ventricular Assist Device (VAD) and a Pulsatile Flow VAD in BIVAD Patients

The aim of this work is to develop and test a lumped parameter model of the cardiovascular system to simulate the simultaneous use of pulsatile (P) and continuous flow (C) ventricular assist devices (VADs) on the same patient. Echocardiographic and hemodynamic data of five pediatric patients undergoing VAD implantation were retrospectively collected and used to simulate the patients' baseline condition with the numerical model. Once the baseline hemodynamic was reproduced for each patient, the following assistance modalities were simulated: (a) CVAD assisting the right ventricle and PVAD assisting the left ventricle (RCF + LPF), (b) CVAD assisting the left ventricle and PVAD assisting the right ventricle (LCF + RPF). The numerical model can well reproduce patients' baseline. The cardiac output increases in both assisted configurations (RCF + LPF: +17%, LCF + RPF: +21%, P = ns), left (right) ventricular volumes decrease more evidently in the configuration LCF + RPF (RCF + LPF), left (right) atrial pressure decreases in the LCF + RPF (RCF + LPF) modality. The pulmonary arterial pressure slightly decreases in the configuration LCF + RPF and it increases with RCF + LPF. Left and right ventricular external work increases in both configurations probably because of the total cardiac output increment. However, left and right artero-ventricular coupling improves especially in the LCF + RPF (-36% for the left ventricle and -21% for the right ventricle, P = ns). The pulsatility index decreases by 8.5% in the configuration LCF + RPF and increases by 6.4% with RCF + LPF (P = 0.0001). A numerical model could be useful to tailor on patients the choice of the VAD that could be implanted to improve the hemodynamic benefits. Moreover, a model could permit to simulate extreme physiological conditions and innovative configurations, as the implantation of both CVAD and PVAD on the same patient.

Simulation of Acute Haemodynamic Outcomes of the Surgical Strategies for the Right Ventricular Failure Treatment in Pediatric LVAD

Background

Right ventricular failure (RVF) is one of the major complications during LVAD. Apart from drug therapy, the most reliable option is the implantation of RVAD. However, BIVAD have a poor prognosis and increased complications. Experiments have been conducted on alternative approaches, such as the creation of an atrial septal defect (ASD), a cavo-aortic shunt (CAS) including the LVAD and a cavo-pulmonary connection (CPC). This work aims at realizing a lumped parameter model (LPM) to compare the acute hemodynamic effects of ASD, CPC, CAS, RVAD in LVAD pediatric patients with RVF.

Methods

Data of 5 pediatric patients undergoing LVAD were retrospectively collected to reproduce patients baseline hemodynamics with the LPM. The effects of continuous flow LVAD implantation complicated by RVF was simulated and then the effects of ASD, CPC, CAS and RVAD treatments were simulated for each patient.

Results

The model successfully reproduced patients' baseline and the hemodynamic effects of the surgical strategies. Simulating the different surgical strategies, an unloading of the right ventricle and an increment of left ventricular preload were observed with an improvement of the hemodynamics (total cardiac output: ASD +15%, CPC +10%, CAS +70% RVAD +20%; right ventricular external work: ASD -19%, CPC -46%, CAS -76%, RVAD -32%; left ventricular external work: ASD +12%, CPC +28%, RVAD +64%).

Conclusions

The use of numerical model could offer an additional support for clinical decision-making, also potentially reducing animal experiments, to compare the outcome of different surgical strategies to treat RVF in LVAD.

Simulation of Ventricular, Cavo-Pulmonary, and Biventricular Ventricular Assist Devices in Failing Fontan

Considering the lack of donors, ventricular assist devices (VADs) could be an alternative to heart transplantation for failing Fontan patients, in spite of the lack of experience and the complex anatomy and physiopathology of these patients. Considering the high number of variables that play an important role such as type of Fontan failure, type of VAD connection, and setting (right VAD [RVAD], left VAD [LVAD], or biventricular VAD [BIVAD]), a numerical model could be useful to support clinical decisions. The aim of this article is to develop and test a lumped parameter model of the cardiovascular system simulating and comparing the VAD effects on failing Fontan. Hemodynamic and echocardiographic data of 10 Fontan patients were used to simulate the baseline patients' condition using a dedicated lumped parameter model. Starting from the simulated baseline and for each patient, a systolic dysfunction, a diastolic dysfunction, and an increment of the pulmonary vascular resistance were simulated. Then, for each patient and for each pathology, the RVAD, LVAD, and BIVAD implantations were simulated. The model can reproduce patients' baseline well. In the case of systolic dysfunction, the LVAD unloads the single ventricle and increases the cardiac output (CO) (35%) and the arterial systemic pressure (Pas) (25%). With RVAD, a decrement of inferior vena cava pressure (Pvci) (39%) was observed with 34% increment of CO, but an increment of the single ventricle external work (SVEW). With the BIVAD, an increment of Pas (29%) and CO (37%) was observed. In the case of diastolic dysfunction, the LVAD increases CO (42%) and the RVAD decreases the Pvci, while both increase the SVEW. In the case of pulmonary vascular resistance increment, the highest CO (50%) and Pas (28%) increment is obtained with an RVAD with the highest decrement of Pvci (53%) and an increment of the SVEW but with the lowest VAD power consumption. The use of numerical models could be helpful in this innovative field to evaluate the effect of VAD implantation on Fontan patients to support patient and VAD type selection personalizing the assistance.

Simulation of Apical and Atrio-aortic VAD in Patients with Transposition or Congenitally Corrected Transposition of the Great Arteries

Purpose

VADs could be used for transportation of the great arteries (TGA) and for congenitally corrected transposition (ccTGA) treatment. A cardiovascular numerical model (NM) may offer a useful clinical support in these complex physiopathologies. This work aims at developing and preliminarily verifying a NM of ccTGA and TGA interacting with VADs.

Methods

Hemodynamic data were collected at the baseline (BL) and three months (FUP) after apical (atrio-aortic) VAD implantation in a TGA (ccTGA) patient and used in a lumped parameter NM to simulate the patient's physiopathology. Measured (MS) and simulated (SIM) data were compared.

Results

MS and SIM data are in accordance at the BL and at FUP. Cardiac output (l/min): BL_m = 2.9 ± 0.4, BL_s = 3.0 ± 0.3; FUP_m = 4.2 ± 0.2, FUP_s = 4.1 ± 0.1. Right atrial pressure (mmHg): BL_m = 21.4 ± 4.1, BL_s = 18.5 ± 4.5; FUP_m = 13 ± 4, FUP_s = 14.8 ± 3.6. Pulmonary arterial pressure (mmHg): BL_m = 56 ± 6.3, BL_s = 57 ± 2, FUP_m = 37.5 ± 7.5, FUP_s = 35.5 ± 5.9. Systemic arterial pressure (mmHg): BL_m = 71 ± 2, BL_s = 74.6 ± 2.1; FUP_m = 84 ± 9, FUP_s = 81.9 ± 9.8.

Conclusions

NM can simulate the effect of a VAD in complex physiopathologies, with the inclusion of changes in circulatory parameters during the acute phase and at FUP. The simulation of differently assisted physiopathologies offers a useful support for clinicians.

Endothelial dysfunction is a marker of systemic response to the cardiac resynchronization therapy in heart failure

BACKGROUND

Cardiac resynchronization therapy (CRT) induces a significant improvement in patients with heart failure (HF), who are often characterized by the presence of endothelial dysfunction (ED) with impaired flow-mediated vasodilation (FMD). We aimed to study the ED in patients with HF candidates to CRT with defibrillator (CRT-D).

METHODS AND RESULTS

We studied 57 consecutive patients affected by HF and undergoing CRT-D. At the baseline we recorded a high prevalence of ED (64.9%) with impaired FMD (4.1 ± 3.8%). After 12 months of CRT, we reported a marked increase of the mean FMD (8.8 ± 4.8% vs 4.1 ± 3.8%; P < .05) along with significant improvement of left ventricular ejection fraction (LVEF), left ventricular end-systolic volume (LVESV), New York Heart Association (NYHA) functional class, and 6-minute walk test (6MWT); 42 patients (73.7%) were classified as responders according to standard criteria. FMD was related to LVEF (r = 0.169; P < .05), LVESV (r = -0.169; P < .05), NYHA functional class (r = -0.27; P < .051), and 6MWT (r = 0.360; P < .01).

CONCLUSIONS

ED is not an independent predictor of CRT response, but it is able to intercept the systemic effects of CRT and is an affordable marker of response to CRT, especially in patients unable to perform the 6MWT.

A novel methodology for AV and VV delay optimization in CRT: results from a randomized pilot clinical trial

The aim of this work was to determine whether the use of a newly developed methodology (Alg1) for AV and VV optimization improves cardiac resynchronization therapy (CRT) clinical and echocardiographic (ECHO) outcomes. In this single-center pilot clinical trial, 80 consecutive patients (79 % male; 70.1 ± 11.2 years) receiving CRT were randomly assigned to AV and VV optimization using Alg1 (group A) or standard commercial procedures (group B). Clinical status and ECHOs were analyzed at baseline (_0) , 3 (fu1), and 6 months (fu2) of follow-up evaluating left ventricular end systolic (LVESV) and end diastolic (LVEDV) volumes, ejection fraction (EF), Minnesota test, and 6-min walk test (6MWT). Alg1 is based on a cardiovascular model fed with patient data. Baseline characteristics did not differ significantly between groups. Group A had a better clinical outcome and reverse remodeling. Remodeling was calculated as the difference (Δ) between fu1 and _0 and between fu2 and fu1, respectively: [LVESV (ml): ΔA_fu1 = -55.3, ΔB_fu1 = -13.5, p_fu1 = 0.002; ΔA_fu2 = -22.8, ΔB_fu2 = 3.0, p_fu2 = 0.04], [LVEDV (ml): ΔA_fu1 = -61.9, ΔB_fu1 = -16.1, p_fu1 = 0.01; ΔA_fu2 = -30.4, ΔB_fu2 = 11.3, p_fu2 = 0.02]; Minnesota test: total (p_fu1 = 0.01; p_fu2 = 0.04), physical (p_fu1 = 0.01; p_fu2 = 0.03) and emotional scores (p_fu1 = 0.04; p_fu2 = 0.03) and in 6MWT (m) (p_fu2 = 0.008). No statistically significant difference was observed in QRS width. Compared with current standard of care, CRT optimization using Alg1 is associated with better outcomes, showing the power of a tailored CRT.

Assessment of hemodynamic load components affecting optimization of cardiac resynchronization therapy by lumped parameter mode

Timing of biventricular pacing devices employed in cardiac resynchronization therapy (CRT) is a critical determinant of efficacy of the procedure. Optimization is done by maximizing function in terms of arterial pressure (BP) or cardiac output (CO). However, BP and CO are also determined by the hemodynamic load of the pulmonary and systemic vasculature. This study aims to use a lumped parameter circulatory model to assess the influence of the arterial load on the atrio-ventricular (AV) and inter-ventricular (VV) delay for optimal CRT performance.