http://orcid.org/0000-0003-1583-5783
Figures
Abstract
Background
Left ventricular assist devices (LVADs) are an established treatment for end stage heart failure patients. As LVADs do not currently respond to exercise demands, attention is also directed towards improvements in exercise capacity and resulting quality of life. The aim of this study was to explore hemodynamic responses observed during maximal exercise tests to infer underlying patient status and therefore investigate possible diagnostics from LVAD derived data and advance the development of physiologically adaptive LVAD controllers.
Methods
High resolution continuous LVAD flow waveforms were recorded from 14 LVAD patients and evaluated at rest and during maximum bicycle exercise tests (n = 24). Responses to exercise were analyzed in terms of an increase (↑) or decrease (↓) in minimum (QMIN), mean (QMEAN), maximum flow (QMAX) and flow pulsatility (QP2P). To interpret clinical data, a cardiorespiratory numerical simulator was used that reproduced patients’ hemodynamics at rest and exercise. Different cardiovascular scenarios including chronotropic and inotropic responses, peripheral vasodilation, and aortic valve pathologies were simulated systematically and compared to the patients’ responses.
Results
Different patients’ responses to exercise were observed. The most common response was a positive change of ΔQMIN↑ and ΔQP2P↑ from rest to exercise (70% of exercise tests). Two responses, which were never reported in patients so far, were distinguished by QMIN↑ and QP2P↓ (observed in 17%) and by QMIN↓ and QP2P↑ (observed in 13%). The simulations indicated that the QP2P↓ can result from a reduced left ventricular contractility and that the QMIN↓ can occur with a better left ventricular contractility and/or aortic insufficiency.
Conclusion
LVAD flow waveforms determine a patients’ hemodynamic “fingerprint” from rest to exercise. Different waveform responses to exercise, including previously unobserved ones, were reported. The simulations indicated the left ventricular contractility as a major determinant for the different responses, thus improving patient stratification to identify how patient groups would benefit from exercise-responsive LVAD control.
Citation: Gross C, Fresiello L, Schlöglhofer T, Dimitrov K, Marko C, Maw M, et al. (2020) Hemodynamic exercise responses with a continuous-flow left ventricular assist device: Comparison of patients’ response and cardiorespiratory simulations. PLoS ONE 15(3): e0229688. https://doi.org/10.1371/journal.pone.0229688
Editor: Vakhtang Tchantchaleishvili, Thomas Jefferson University, UNITED STATES
Received: October 30, 2019; Accepted: February 11, 2020; Published: March 18, 2020
Copyright: © 2020 Gross et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: The authors CG, TS, KD, CM, MM, BM, DW, DZ and HS received no specific funding for this work. FM received funding from the Austrian Science Fund (FWF, KLI357) and LF received the Frans Van de Werf scholarship.
Competing interests: The authors have declared that no competing interests exist.
Abbrevations: AV, aortic valve; AI, Aortic Insufficiency; AS, Aortic Stenosis; EmaxL/EmaxR, left/right ventricular contractility; HR, heartrate; LVAD, left ventricular assist device; TPR, total peripheral resistance; QMAX, pump flow maximum; QMIN, pump flow minimum; QMEAN, pump flow mean; QP2P, pump flow pulsatility
Introduction
Left ventricular assist devices (LVADs) have become an established therapy to manage end-stage heart failure [1]. Patients receive an LVAD with different treatment intentions: as a bridge to cardiac transplant, as a bridge to further treatment decision, as bridge to cardiac recovery, and even for lifetime implantation (so-called destination therapy). This results in a patient population with a broad spectrum of demographic parameters and comorbidities with individual pathophysiological conditions. Patients typically benefit of hemodynamic normalization at rest after LVAD implantation.
Despite increasing implantation rates over the last decade with improvements in patient outcomes and acceptable durability of currently used LVADs [1], exercise capacity remain substantially lower compared to gender and age predicted values [2]. Exercise capacity, often referred to by measurements of oxygen uptake at peak exercise, involves mechanisms at multiorgan levels with a prominent component related to cardiac function [2,3]. Exercise response in LVAD patients may reflect underlying differences in cardiac and peripheral conditions, as well as in the interaction between the assisted left ventricle and the LVAD, strongly tied to the overall cardiac output. This study is an attempt to characterize these patients’ cardiac and peripheral conditions at the level of the pump flow waveform and hypothesize their determinants.
Currently used LVADs operate at a fixed pump speed with different resulting supporting levels. When the cardiac demand is solely delivered by the pump one speaks of full-support. With a partial ventricular support by the LVAD, an additional volume of blood is ejected through the aortic valve in parallel to the output provided by the LVAD. Improvement in cardiac output with exercise results with fixed LVAD operation from the adaptations of the patient’s cardiovascular system, rather than from an increase in pump output [4], thus indicating possible improvements by exercise-responsive LVADs. However, a proper characterization of patient residual adaptation mechanisms to exercise is important for the design of exercise-responsive LVADs. High-resolution LVAD flow waveforms have proven to be a valuable source of information for the detection of patient’s hemodynamic status [5]: Aortic valve (AV) opening [6,7], heartrate [8], suction events [9], contractility and relaxation parameters [10,11]. This LVAD-based diagnostics can be therefore performed systematically during exercise tests, thus revealing the unobserved information of this complex interaction between heart and LVAD. The aim of this study was to evaluate, for the first time ever, patients’ hemodynamic LVAD flow waveform responses during exercise and compare the results to the responses reproduced with systematic cardiorespiratory numerical simulations (performed independent from responses observed in patients). This should help to design physiologically adaptive controls which take into account different types of exercise as well as individual patient conditions.
Methods
This work consists of a clinical study conducted on 14 patients that underwent a total of 24 maximal bicycle exercise tests while high resolution LVAD data were recorded. The LVAD data of these patients collected from rest to exercise were analyzed and compared to the LVAD data resulting from a cardiorespiratory simulator. The cardiorespiratory simulator reproduced hemodynamics during exercise for the average LVAD patient from literature with additional single cardiovascular parameter changes described in detail below.
Patient analysis
Patient data was collected within a prospective observational study approved by the Institutional Review Board of the Medical University Vienna (EK-243/2011, ClinicalTrial.gov identifier: NCT01981642). High resolution LVAD data (motor current and impeller rotational speed) of patients implanted with an HVAD (Medtronic Inc., Minneapolis, MN, USA) during maximum bicycle exercise tests were recorded continuously with a sampling frequency of 50 Hz. Pharmacologic treatment of patients was in accordance with the guidelines of the European Society of Cardiology [12] and included beta-blocker, ACE-inhibitor, diuretica, calcium-channel blockers and Angiotensin II type I receptor blockers. Mean arterial blood pressure levels from 70–85 mmHG in these patients were desired. Echocardiographic examination of the right ventricular function was assessed according to the recommendations of the American college of cardiology [13]. The classification of the right ventricular function was made based on global visual assessment [13,14] and staged as normal or, in case of dysfunction as mild, moderate or severe dysfunction.
Patients’ exercise tests were performed on a bicycle ergometer and included cardiopulmonary exercise tests (upright position) and bicycle stress-echocardiography (semi-recumbent position). The incremental increase in workload was determined by the clinician on a patient-individual basis, according to previous results of maximum or sub-maximum physical capacity tests. Bicycle ergometer tests were performed until subjectively perceived maximum physical capacity symptoms developed. Additional information concerning the performed tests is provided in reference [15].
Pump flow was estimated from the recorded LVAD motor current and impeller rotational speed [16] and heartrate (HR) was derived subsequently using a previously developed method [8]. For each cardiac cycle the following hemodynamic features were calculated (see Fig 1): pump flow minimum (QMIN), mean (QMEAN), maximum (QMAX) values, flow pulsatility (QP2P) and aortic valve (AV) opening [6,17]. LVAD waveform responses were classified by the changes of the parameters QMIN, QMEAN, QMAX and QP2P. The symbols ↓ and ↑ will be used to indicate negative (<0 L/min) and “positive” (≥0 L/min) parameter changes from baseline to peak exercise.
LVAD pump flow features analyzed to characterize patients’ response to exercise. For description of symbols please refer to the text.
Statistical analysis
Data was processed and analyzed with MATLAB (TheMathworks Inc., Natick, MA, USA). Metric variables were tested for normality with the Shapiro-Wilk test for small sample sizes and are reported as mean ± standard deviation (SD) for normally distributed data. Non-normally distributed data is reported as median with 25th and 75th percentile (quartile 1 & 3). Statistical significance for responses in LVAD parameters was calculated with the two sample t-test for the normal distributed metric data and with Fisher’s exact test for nominal variables. Statistical significance was set at p<0.05.
Exercise simulations
To evaluate possible hemodynamic mechanisms responsible for responses in LVAD parameters to exercise, a computational simulator was used and a sensitivity analysis for systematic single parameter changes performed. The simulator was developed at the Cardiac Surgery Department of the Katholieke Universiteit Leuven in LabVIEW (National Instruments Austin, TX, USA) (see Fig 2). The simulator included a lumped parameters model representing the cardiovascular and respiratory systems combined together and specifically adapted to reproduce heart failure conditions [18,19]. The cardiovascular system included a time varying elastance model for the representation of the atria and ventricles. The circulation is split into different circulatory compartments (ascending and descending aorta, upper body, kidneys, splanchnic circulation, left and right leg, superior and inferior vena cava and pulmonary circulation), in which the metabolic state of oxygen consumption and production of carbon dioxide can be simulated depending on physical exertion.
Schematic diagram of the cardiovascular compartments and biological feedback control loops (arrows) of the cardiorespiratory simulator.
Specific physiological feedback mechanisms were implemented to reproduce the cardiorespiratory changes observed from rest to exercise. Autonomic control regulated sympathetic nerve overstimulation and vagal withdrawal resulting in vasoconstriction of the systemic circulation and positive inotropic and chronotropic responses of the heart during exercise. Metabolic control regulated the vasodilation in the peripheral tissues, if hypoxia was detected. Concerning ventilation, a specific control was implemented that regulates tidal volume and ventilation frequency according to the oxygen and carbon dioxide partial pressure sensed in the upper body.
The model of the LVAD was implemented according to the in-vitro measurements with a HVAD pump reported in [16] and connected between the left ventricle and the aorta (see Fig 2) [20–22].
The average hemodynamic condition of LVAD patients at rest were simulated with a LVAD flowrate of 4.8 L/min (full support, thereby equaling the cardiac output) at a LVAD speed of 2700rpm as described in [23–27]. Then, a bicycle exercise with an intensity of 80 watts was simulated, corresponding to an oxygen uptake of 15.2 ml/min/kg, a typical maximum value reported in LVAD patients at peak exercise. The simulation evolved in order to attain the exercise hemodynamic condition, with similar hemodynamic adaptations as reported by previous clinical studies [23–27]. This set of simulation parameters, that reproduce an average LVAD patient, are referred to as baseline (BLREST and BL, at rest and during exercise respectively). In addition to that, other simulation runs were performed to mimic different LVAD patient profiles. For this purpose, one cardiovascular parameter at a time was decoupled from feedback control mechanisms and manually changed instead. See Table 1 for a list of these parameters and their changes.
Heart rate (HR), left/right ventricular contractility (EmaxL/EmaxR) and total peripheral resistance (TPR) were increased and decreased by 20% compared to BL simulation at exercise. This permitted to mimic a better/poorer chronotropic, inotropic or peripheral circulatory response to exercise. Valvular pathologies, as observed in some LVAD patients, were also simulated such as aortic valve insufficiency (AI) and aortic valve stenosis (AS) [28,29]. The simulated cardiovascular parameters mentioned beforehand are reported in Table 1.
A comprehensive overview of the cardiovascular parameter profile for the baseline simulation and the parameter ranges of the performed simulations during exercise are shown in Table 1. The LVAD flow, resulting from these different simulations, was analyzed with the same signal processing routines used to analyze the clinical data (see Fig 3).
Left heart pressures, pump head pressure (H) and corresponding LVAD pump flow with the analyzed flow waveform parameters from simulation. Hemodynamics corresponding to the time points at QMAX and QMIN are depicted with t1 and t2, respectively. Furthermore resulting pump flows from pump head (H = AoP-LVP) and the LVAD’s characteristic are shown as HQ-Loop (top right).
Results
Patient data
A total of 24 bicycle exercise stress tests, consisting of 18 cardiopulmonary exercise tests and 6 bicycle stress-echocardiography, were analyzed from 14 patients implanted with an LVAD. Demographic, clinical and exercise data are reported in Table 2. For all bicycle exercise tests the response in parameters ΔQMEAN and ΔQMAX was always positive (↑) whereas different responses in ΔQMIN and ΔQP2P occurred. Consequently, three different types of LVAD waveform responses were identified based on the changes of ΔQMIN and ΔQP2P. Response type 1 is characterized by ΔQMIN↑ and ΔQP2P↑ and was found in 70% of all exercise tests. This response occurred in all exercise tests of patients with the AV remaining closed at peak exercise and in 60% of all tests with the AV open at peak exercise. Response type 2 is classified by ΔQMIN↑ and ΔQP2P↓ and was found in 17% of all tests. Response type 3 is characterized by ΔQMIN↓ and ΔQP2P↑ and was found in 13% of all tests. Response type 2 and 3 were only found in tests with the patients’ AV open at peak exercise. The statistics for the subgroups of the three identified responses is provided in Table 3. Additionally examples of the changes in LVAD waveforms from rest to exercise in the three responses are shown in Fig 4. Individual LVAD parameter responses for each bicycle exercise test can be observed in the online available data of this study.
Overview of the three average responses from patients’ LVAD pump flow at rest and during exercise. Parameter values for the responses are depicted in Table 3.
Five of the 14 patients underwent multiple bicycle exercise tests. Two of these five patients showed intra-individual differences in responses, one patient due to different AV status at rest (Patient Nr. 11 in S1 Table) and the other due to the occurrence of suction (Patient Nr. 2 in S1 Table).
Simulation data
Simulations refer to the exercise response of the average LVAD patient from literature (BL) and to the additional responses summarized in Table 1. Simulated LVAD flow was analyzed and classified with the same method used for clinical data. In the simulations the LVAD waveform parameter responses already identified from patient data occurred (see Fig 5).
Time normalized pump flow waveforms for simulations resulting in the LVAD waveform parameter responses observed in patients. Heartrate for BLREST was 75 bpm, for AIREST 80 bpm, for HR- 90 bpm, for HR+ 136 bpm and ranging from 112 to 114 bpm for all other exercise simulations. Abbreviations and overview of simulations are shown in Table 1.
Response type 1 (QMIN↑, QMEAN↑, QMAX↑, QP2P↑) was observed in most of the simulations: BL, HR-, HR+, EmaxR-, EmaxR+, TPR+, TPR-, and AS. Response type 2 (ΔQMIN↑ and ΔQP2P↓) was observed for the simulation EmaxL-. The ΔQMIN↓ and ΔQP2P↑ from Response type 3 was observed for the simulations EmaxL+ and AI. A negative ΔQMEAN of -0.1 L/min was observed for the simulation EmaxL+, corresponding to an exceptionally good left ventricular contractility. Furthermore for each simulation at exercise hemodynamics such as total cardiac output and its repartition, preload and afterload surrogates are shown in Fig 6.
Total cardiac output and its repartition between the LV and LVAD are shown together with mean aortic pressure and maximum left atrial pressure for each simulation at exercise. Abbreviations and overview of simulations are shown in Table 1. (QRegurg: aortic valve regurgitant flow due to valve insufficiency; QLVAD equals QMEAN; dotted line indicates levels for BL).
Discussion
This study shows diverse responses in hemodynamic LVAD flow during exercise. The currently held assumption on the response of the LVAD to exercise predicts only one of the different responses observed (Response type 1), while this study also shows for the first time different LVAD waveform responses that shed light on the remaining cardiac function and overall on the patient-LVAD interaction. These observations were possible because of a unique monitoring device that records high resolution pump flow waveforms, which is not an integral part of the currently clinical available device. To understand the mechanisms underlying the hemodynamic LVAD responses observed in patients, numerical cardiopulmonary simulations were used.
LVAD waveform responses
The currently held assumption posits, that LVAD flow rate and waveform parameters increase with preload (as a surrogate for physical activity or exercise) [30–32]. To better describe the LVAD interactions with the cardiovascular system during exercise, we introduced a classification strategy based on relevant and easily observable LVAD waveform parameters: QMIN, QMEAN, QMAX and QP2P. The analysis we conducted on LVAD waveforms at exercise, revealed that also a decrease in QP2P or a decrease in QMIN with exercise was observed in some patients.
Response type 1 (ΔQMIN↑, ΔQMEAN↑, ΔQMAX↑, ΔQP2P↑) supports the assumed LVAD response to exercise and was found in the majority of exercise tests performed by patients and in most results from the simulations. This response resulted from patients with an open AV at peak exercise, as well as patients remaining in full-support. Similar responses were replicated by BL (the average patient simulation) as well as all simulations with superimposed parameters (HR±, EmaxR±, TPR± and AS) except for EmaxL± and AI.
Response type 2 (ΔQP2P↓) and response type 3 (ΔQMIN↓) were found in a minority of exercise tests, all in patients with AV opening at peak exercise indicating associations of the LV remaining contractile reserve. Indeed the results of the simulations support the importance of the output through the AV to differentiate between the two waveform responses to exercise. The decline in QP2P occurred in the simulations with a poor LV (EmaxL-), whereas the decline in QMIN resulted from the simulations with a higher LV inotropic response or in presence of AI.
LVAD flow parameters
The mechanisms responsible for the change in beat-to-beat LVAD parameters were hypothesized to correspond to the hemodynamics from the systematic simulations. The cardiorespiratory simulator provided helpful insights in underlying mechanisms difficult to analyze in patients due the complexity with highly individual cardiac and peripheral responses to exercise, confounding factors and the sparse availability of hemodynamic assessments in patients during exercise tests.
QMAX.
The increase in QMAX observed at exercise, is due to an increase in the peak LV pressure. This might be due to a residual inotropic response of the LV and/or to an increased venous return that would generate a higher systolic LV pressure according to the Frank–Starling mechanism. Similar ΔQMAX with +0.5 L/min occurred for BL, HR± and EmaxR± (* further discussion on right heart function below). A ΔQMAX increase of +0.4 L/min was observed with exercise for the simulations EmaxL- and TPR+ and of +0.6 L/min for EmaxL+ and TPR-. Reduced ΔQMAX of +0.2 L/min occurred for AI. Increased ΔQMAX of +0.9 L/min occurred for AS, due to the high systolic pressure gradient across the AV in presence of AS.
QMIN.
The LVAD parameter during diastole (QMIN) is related to the systolic arterial pressure as well as preload (filling ventricular pressure). During exercise, these two variables are affected by the increase in cardiac output, peripheral vasodilation and increase in venous return. In the majority of the simulations (BL, HR±, TPR±, EmaxR+ and AS) ΔQMIN increased (range +0.1 to 0.6 L/min) with exercise. This behavior results from a higher increase in preload compared to afterload at timepoint of QMIN in the cardiac cycle (observable as t2 in Fig 3) [33]. Furthermore the modulation of LV contractility produced the ranges in simulated ΔQMIN responses with –1.0 L/min for EmaxL+ and +1.3 L/min for EmaxL-, mostly mediated by the effect that an increased contractility has on the systolic arterial pressure.
QP2P.
Depending on the rate of change from rest to exercise in QMIN and QMAX the resulting ΔQP2P increases, decreases or might remain constant. A decrease in QP2P resulted only in the simulation EmaxL-, due to the high ΔQMIN corresponding to the high preload levels.
QMEAN.
The parameter QMEAN is calculated over the whole cardiac cycle and therefore an indicator for overall hemodynamic changes rather than particular cardiovascular impairments. Generally, due to the instantaneous increase in preload with dynamic exercises a sudden increase of QMEAN can be observed [4,32,34–36]. However QMEAN can increase to a different extent according to the residual ventricular contractility, LVAD speed setting as well as other parameters. The simulation with an exceptionally high LV contractility showed even a minor reduction of QMEAN during exercise compared to rest.
The simulations EmaxR± indicate that the LVAD response classification with exercise, used in this study, might not be sensitive to stratify for differences in EmaxR. Indeed the clinical observations also do not show clear differences in patients with different RV function. However hemodynamic differences according to EmaxR could be noticed: For the exercise simulation EmaxR- a reduction in left atrial pressure, aortic pressure, total cardiac output and QMEAN occurred, compared to EmaxR+ (see Fig 6). These hemodynamic differences mainly affect the diastolic part of the LVAD flow waveform (see Fig 5). (Additional information on cardiac output for the simulations are published in Gross et al. [37]).
Clinical interpretation and relevance
Constant speed LVAD devices lack adequate adaption of pump output to exercise levels resulting in the hypoperfusion of patients during stress [3,34]. The observed differences in QMEAN responses from rest to exercise in patients and simulations underline that the support of the device in sustaining cardiac output is different among patients and can be influenced by several cardiovascular parameters [37]. Clinical observations in LVAD patients during exercise with increased as well as decreased LVAD speeds have demonstrated effects on exercise hemodynamics in patient subgroups, but not in the whole study cohorts [38,39]. The diverse characteristics of LVAD patients’ responses during exercise necessitate patient stratification. For this purpose, clinically validated beat-to-beat hemodynamic parameters can be helpful and easily obtained from continuous LVAD monitoring such as used in this study [6,8]. For exercise studies comparing baseline LVAD speed to increased speeds we propose the following patient stratification based on the baseline speed observations (A-D):
- AV closed at peak exercise (Response type 1 with AV closed): Due to the inability of the LV to increase total cardiac output, LVAD speed increase most probably will increase total cardiac output and increase unloading, in case of a non-failing RV.
- AV open at peak exercise (Response type 2): Patients may tolerate the highest LVAD speeds at exercise due to the impairment of the LV and high pulmonary pressures. These patients may benefit the most from LVAD speed increase during exercise in terms of total cardiac output.
- AV open at peak exercise (Response type 1 with AV open): This category will include the most patients and benefits with speed increase will most probably result in outcomes similar to previous studies [23–27,36].
- AV open at peak exercise (Response type 3): These patients may experience minor benefits of a speed increase during exercise due to a remaining LV contractility and its ability to increase total cardiac output. LVAD speed increase during exercise in this category could result in the redistribution of cardiac output between the LVAD and what it is ejected towards the aortic valve without the degree of increase of overall flow as it might be observed in others [37].
Furthermore, the proposed stratification could be helpful for physiological LVAD speed control algorithms to set the appropriate support mode to provide adequate support during exercise. Up to now most LVAD speed control algorithms are studied in-silico, ex-vivo or in-vivo in animals without heart failure. In-silico or ex-vivo studies could be carried out to reproduce the diverse responses found in patients in order to improve algorithms based on clinically observed evidence. For this purpose, the individual responses are available in the online data of this study (S1 Table). Characterizing the individual patient’s cardiac, peripheral and LVAD response during exercise could have implications in diagnostics and clinical management of LVAD patients and help to drive technological advancements of LVADs towards “smart pumping”. The potential stratification proposed in this section could be the foundation for further work, following clinical validation. This study comprises useful methods and reports possible mechanisms for the different responses helpful to design and carry out such clinical validation. The intra-individual differences in responses observed in two patientns emphasize the need for repeated tests performed, together with the proper clinical assessments for e.g.: volume status, right and left ventricular function. Further work may include more parameters for response stratification as shown in Table 3.
Limitations
For the patient analysis only 14 patients received the high-resolution LVAD data recorder and performed maximum bicycle exercise tests. Cardiopulmonary exercise tests and bicycle stress-echocardiography were combined, nevertheless due to the occurrences of all response types in both bicycle exercise test modalities no systematic error was introduced. Furthermore, variables as cardiac output and invasive pressure measurements were not available during exercise tests. Some of the occasionally measured non-invasive blood pressures were unreliable, due to the low arterial pulse pressure with an LVAD, and therefore not analyzed in this study. Only the HVAD pump was studied however we would not expect major differences at least for other centrifugal pumps (Heartmate 3) Despite these limitations, this study allowed the observation of hemodynamic responses to exercise not yet reported in the literature, which were corroborated by extensive numerical simulations, in which variables as cardiac output or arterial pressure behaved as reported in clinical observations.
In this study systematic simulations based on the average LVAD patient data from literature were performed, and the effect of single parameter changes investigated. This is different in real patients as multiple concomitant conditions might be present simultaneously. Nevertheless this approach can show that specific changes in parameter from the average patient (e.g. lower LV contractility) might explain some of the peculiar exercise responses observed (e.g. decrease in flow pulsatility). Worth to mention is that the simulation with a reduced (-20%) right ventricular contractility labeled as EmaxR-, cannot be interpreted as or correlated to hemodynamics resulting with a failing right ventricle.
Conclusion
Diverse responses in pump flow patterns were observed in LVAD patients performing bicycle exercise stress tests reflecting differences in the underlying remaining cardiac condition as one of the major determinants for the responses. In patients with poorer contractility exercise-responsive LVAD control might bring additional benefits in terms of adequate support of the failing left ventricle, the patient’s exercise capacity and resulting quality of life.
Supporting information
S1 Table. Hemodynamic LVAD responses for each exercise test.
Patients’ LVAD parameters at baseline and peak exercise and response type classification with constant LVAD speeds.
https://doi.org/10.1371/journal.pone.0229688.s001
(XLS)
References
- 1. Kormos RL, Cowger J, Pagani FD, Teuteberg JJ, Goldstein DJ, Jacobs JP, et al. The Society of Thoracic Surgeons Intermacs Database Annual Report: Evolving Indications, Outcomes, and Scientific Partnerships. The Annals of Thoracic Surgery. 2019 Feb;107(2):341–53. pmid:30691584
- 2. Loyaga-Rendon RY, Plaisance EP, Arena R, Shah K. Exercise physiology, testing, and training in patients supported by a left ventricular assist device. The Journal of Heart and Lung Transplantation. 2015 Aug;34(8):1005–16. pmid:25682553
- 3. Mezzani A, Pistono M, Agostoni P, Giordano A, Gnemmi M, Imparato A, et al. Exercise gas exchange in continuous-flow left ventricular assist device recipients. Rengo G, editor. PLoS ONE. 2018 Jun 1;13(6):e0187112. pmid:29856742
- 4. Martina J, Jonge N, Rutten M, Kirkels JH, Kl?pping C, Rodermans B, et al. Exercise Hemodynamics During Extended Continuous Flow Left Ventricular Assist Device Support: The Response of Systemic Cardiovascular Parameters and Pump Performance: Exercise Hemodynamics During CF-LVAD Support. Artificial Organs. 2013 Sep;37(9):754–62. pmid:24074245
- 5. Rich JD, Burkhoff D. HVAD Flow Waveform Morphologies: Theoretical Foundation and Implications for Clinical Practice. ASAIO Journal. 2017;63(5):526–35. pmid:28323662
- 6. Granegger M, Masetti M, Laohasurayodhin R, Schloeglhofer T, Zimpfer D, Schima H, et al. Continuous Monitoring of Aortic Valve Opening in Rotary Blood Pump Patients. IEEE Transactions on Biomedical Engineering. 2016 Jun;63(6):1201–7. pmid:26461795
- 7. Hayward C, Lim CP, Schima H, Macdonald P, Moscato F, Muthiah K, et al. Pump Speed Waveform Analysis to Detect Aortic Valve Opening in Patients on Ventricular Assist Device Support. Artificial Organs. 2015;39(8):704–9. pmid:26234450
- 8. Moscato F, Granegger M, Edelmayer M, Zimpfer D, Schima H. Continuous monitoring of cardiac rhythms in left ventricular assist device patients. Artificial Organs. 2014;38(3):191–8. pmid:23902542
- 9. Vollkron M, Schima H, Huber L, Benkowski R, Morello G, Wieselthaler G. Advanced Suction Detection for an Axial Flow Pump. Artificial Organs. 2006 Sep;30(9):665–70. pmid:16934094
- 10. Moscato F, Granegger M, Naiyanetr P, Wieselthaler G, Schima H. Evaluation of Left Ventricular Relaxation in Rotary Blood Pump Recipients Using the Pump Flow Waveform: A Simulation Study. Artificial Organs. 2012;36(5):470–8. pmid:22171892
- 11. Naiyanetr P, Moscato F, Vollkron M, Zimpfer D, Wieselthaler G, Schima H. Continuous assessment of cardiac function during rotary blood pump support: A contractility index derived from pump flow. Journal of Heart and Lung Transplantation. 2010;29(1):37–44. pmid:19782591
- 12. Ponikowski P, Voors AA, Anker SD, Bueno H, Cleland JGF, Coats AJS, et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: The Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC)Developed with the special contribution of the Heart Failure Association (HFA) of the ESC. Eur Heart J. 2016 Jul 14;37(27):2129–200. pmid:27206819
- 13. Rudski LG, Lai WW, Afilalo J, Hua L, Handschumacher MD, Chandrasekaran K, et al. Guidelines for the Echocardiographic Assessment of the Right Heart in Adults: A Report from the American Society of Echocardiography. Journal of the American Society of Echocardiography. 2010 Jul;23(7):685–713. pmid:20620859
- 14. Kossaify A. Echocardiographic Assessment of the Right Ventricle, from the Conventional Approach to Speckle Tracking and Three-Dimensional Imaging, and Insights into the “Right Way” to Explore the Forgotten Chamber. Clin Med Insights Cardiol. 2015;9:65–75. pmid:26244034
- 15. Gross C, Marko C, Mikl J, Altenberger J, Schlöglhofer T, Schima H, et al. LVAD Pump Flow Does Not Adequately Increase With Exercise. Artif Organs. 2019 Mar;43(3):222–8. pmid:30155903
- 16. Granegger M, Moscato F, Casas F, Wieselthaler G, Schima H. Development of a pump flow estimator for rotary blood pumps to enhance monitoring of ventricular function. Artificial Organs. 2012;36(8):691–9. pmid:22882439
- 17. Granegger M, Schima H, Zimpfer D, Moscato F. Assessment of aortic valve opening during rotary blood pump support using pump signals. Artificial Organs. 2014;38(4):290–7. pmid:24102321
- 18. Fresiello L, Meyns B, Di Molfetta A, Ferrari G. A Model of the Cardiorespiratory Response to Aerobic Exercise in Healthy and Heart Failure Conditions. Frontiers in Physiology [Internet]. 2016 Jun 8;7. Available from: http://journal.frontiersin.org/Article/10.3389/fphys.2016.00189/abstract
- 19. Fresiello L, Rademakers F, Claus P, Ferrari G, Di Molfetta A, Meyns B. Exercise physiology with a left ventricular assist device: Analysis of heart-pump interaction with a computational simulator. Lionetti V, editor. PLOS ONE. 2017 Jul 24;12(7):e0181879. pmid:28738087
- 20. Graefe R, Henseler A, Körfer R, Meyns B, Fresiello L. Influence of left ventricular assist device pressure-flow characteristic on exercise physiology: Assessment with a verified numerical model. In press on The International Journal of Artificial Organs (IJAO).
- 21. Moscato F, Wirrmann C, Granegger M, Eskandary F, Zimpfer D, Schima H. Use of continuous flow ventricular assist devices in patients with heart failure and a normal ejection fraction: A computer-simulation study. The Journal of Thoracic and Cardiovascular Surgery. 2013 May;145(5):1352–8. pmid:22841169
- 22. Schima H, Vollkron M, Wieselthaler G, Huber L, Wolner E. COMPUTER MODEL OF THE ASSISTED CIRCULATION ON THE INTERACTION OF THE IMPAIRED HEART WITH IMPLANTED ROTARY BLOOD PUMPS: ASAIO Journal. 2000 Mar;46(2):196.
- 23. Brassard P, Jensen AS, Nordsborg N, Gustafsson F, Moller JE, Hassager C, et al. Central and Peripheral Blood Flow During Exercise With a Continuous-Flow Left Ventricular Assist Device: Constant Versus Increasing Pump Speed: A Pilot Study. Circulation: Heart Failure. 2011 Sep 1;4(5):554–60. pmid:21765126
- 24. Jacquet L, Vancaenegem O, Pasquet A, Matte P, Poncelet A, Price J, et al. Exercise Capacity in Patients Supported With Rotary Blood Pumps Is Improved by a Spontaneous Increase of Pump Flow at Constant Pump Speed and by a Rise in Native Cardiac Output: EXERCISE CAPACITY WITH ROTARY PUMPS. Artificial Organs. 2011 Jul;35(7):682–90. pmid:21615428
- 25. Jakovljevic DG, George RS, Nunan D, Donovan G, Bougard RS, Yacoub MH, et al. The impact of acute reduction of continuous-flow left ventricular assist device support on cardiac and exercise performance. Heart. 2010 Sep 1;96(17):1390–5. pmid:20643664
- 26. Jung MH, Hansen PB, Sander K, Olsen PS, Rossing K, Boesgaard S, et al. Effect of increasing pump speed during exercise on peak oxygen uptake in heart failure patients supported with a continuous-flow left ventricular assist device. A double-blind randomized study: Effect of increasing pump speed during exercise. European Journal of Heart Failure. 2014 Apr;16(4):403–8. pmid:24464845
- 27. Muthiah K, Robson D, Prichard R, Walker R, Gupta S, Keogh AM, et al. Effect of exercise and pump speed modulation on invasive hemodynamics in patients with centrifugal continuous-flow left ventricular assist devices. The Journal of Heart and Lung Transplantation. 2015 Apr;34(4):522–9. pmid:25662859
- 28. Grinstein J, Kruse E, Sayer G, Fedson S, Kim GH, Sarswat N, et al. Novel echocardiographic parameters of aortic insufficiency in continuous-flow left ventricular assist devices and clinical outcome. The Journal of Heart and Lung Transplantation. 2016 Aug;35(8):976–85. pmid:27373822
- 29. Johnson NP, Zelis JM, Tonino PAL, Houthuizen P, Bouwman RA, Brueren GRG, et al. Pressure gradient vs. flow relationships to characterize the physiology of a severely stenotic aortic valve before and after transcatheter valve implantation. European Heart Journal. 2018 Jul 21;39(28):2646–55. pmid:29617762
- 30. Gaddum NR, Stevens M, Lim E, Fraser J, Lovell N, Mason D, et al. Starling-Like Flow Control of a Left Ventricular Assist Device: In Vitro Validation: Starling-Like Control of LVAD. Artificial Organs. 2014 Mar;38(3):E46–56. pmid:24372519
- 31. Salamonsen RF, Lim E, Gaddum N, AlOmari A-HH, Gregory SD, Stevens M, et al. Theoretical Foundations of a Starling-Like Controller for Rotary Blood Pumps: STARLING-LIKE CONTROLLER FOR AN IRBP. Artificial Organs. 2012 Sep;36(9):787–96. pmid:22626056
- 32. Schima H, Vollkron M, Jantsch U, Crevenna R, Roethy W, Benkowski R, et al. First Clinical Experience with an Automatic Control System for Rotary Blood Pumps During Ergometry and Right-Heart Catheterization. The Journal of Heart and Lung Transplantation. 2006 Feb;25(2):167–73. pmid:16446216
- 33. Lai JV, Muthiah K, Robson D, Prichard R, Walker R, Pin Lim C, et al. Impact of Pump Speed on Hemodynamics With Exercise in Continuous Flow Ventricular Assist Device Patients: ASAIO Journal. 2019 Feb;1.
- 34. Gross C, Marko C, Mikl J, Altenberger J, Schlöglhofer T, Schima H, et al. LVAD pump flow does not adequately increase with exercise. Artificial Organs [Internet]. 2018 Aug 28; Available from: http://doi.wiley.com/10.1111/aor.13349
- 35. Schmidt T, Bjarnason-Wehrens B, Mommertz S, Hannig M, Schulte-Eistrup S, Willemsen D, et al. Changes in Total Cardiac Output and Oxygen Extraction During Exercise in Patients Supported With an HVAD Left Ventricular Assist Device: CHANGES IN TOTAL CARDIAC OUTPUT AND OXYGEN EXTRACTION. Artificial Organs [Internet]. 2018 Feb 12; Available from: http://doi.wiley.com/10.1111/aor.13102
- 36. Fresiello L, Buys R, Timmermans P, Vandersmissen K, Jacobs S, Droogne W, et al. Exercise capacity in ventricular assist device patients: clinical relevance of pump speed and power. European Journal of Cardio-Thoracic Surgery. 2016 Oct;50(4):752–7. pmid:27174552
- 37. Gross C, Moscato F, Schlöglhofer T, Maw M, Meyns B, Marko C, et al. LVAD speed increase during exercise, which patients would benefit the most? A simulation study. Artif Organs. 2019 Sep 13;aor.13569.
- 38. Jung MH, Houston B, Russell SD, Gustafsson F. Pump speed modulations and sub-maximal exercise tolerance in left ventricular assist device recipients: A double-blind, randomized trial. The Journal of Heart and Lung Transplantation. 2017 Jan;36(1):36–41. pmid:27522486
- 39. Noor MR, Bowles C, Banner NR. Relationship between pump speed and exercise capacity during HeartMate II left ventricular assist device support: influence of residual left ventricular function. European Journal of Heart Failure. 2012 Jun;14(6):613–20. pmid:22505397