https://orcid.org/0000-0002-5003-700X
Abstract
The evaluation of exercise capacity and cardiac output (QC) is fundamental in the management of patients with advanced heart failure (AdHF). QC and peak oxygen uptake (VO2) have a pivotal role in the prognostic stratification and in the definition of therapeutic interventions, including medical therapies and devices, but also specific treatments such as heart transplantation and left ventricular assist device (LVAD) implantation. Due to the intertwined relationship between exercise capacity and daily activities, exercise intolerance dramatically has impact on the quality of life of patients. It is a multifactorial process that includes alterations in central and peripheral haemodynamic regulation, anaemia and iron deficiency, pulmonary congestion, pulmonary hypertension, and peripheral O2 extraction. This paper aims to review the pathophysiological background of exercise limitations in HF patients and to examine the complex physiology of exercise in LVAD recipients, analysing the interactions between the cardiopulmonary system, the musculoskeletal system, the autonomic nervous system, and the pump. We performed a literature review to highlight the current knowledge on this topic and possible interventions that can be implemented to increase exercise capacity in AdHF patients—including administration of levosimendan, rehabilitation, and the intriguing field of LVAD speed changes. The present paper confirms the role of CPET in the follow-up of this peculiar population and the impact of exercise capacity on the quality of life of AdHF patients.
Background
Advanced heart failure (AdHF) is defined by the presence of severe symptoms of heart failure (HF) and objective evidence of severe cardiac dysfunction despite optimal medical and electrical therapy. In clinical practice, a persistent New York Heart Association (NYHA) class III–IV, a reduced left ventricle ejection fraction (LVEF) < 30%, frequent hospitalizations for HF and severe impairment of exercise capacity with a peak oxygen uptake (VO2) < 12 mL/kg/min, or <50% predicted value are the most commonly encountered indicators. These parameters help the clinician in selecting the patients who may benefit from heart transplantation (HT) or long-term mechanical cardiac support (MCS) such as left ventricular assist device (LVAD). Actually, LVAD should always be considered in patients with progressive clinical decay who are waiting for HT [bridge to transplantation (BTT)], those who are non-eligible for HT due to reversible contraindications [bridge to candidacy (BTC)], or those who are categorically excluded from HT [destination therapy (DT)]. Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) profiles from 2 to 4 are the best established to give indication for durable MCS devices, while profiles 5 and 6 with high-risk characteristics may also be considered.1 In this scenario, cardiopulmonary exercise test (CPET) plays a pivotal role in risk-stratifying patients with AdHF, enabling to measure cardiac output (QC) and VO2 which are mandatory parameters for both prognosis and therapy.2,3 Exercise capacity measurement is pivotal in the management of AdHF patients, and its importance is reflected in the ability to perform daily activities. In this paper, our purpose is to analyse the pathophysiological background of exercise limitations in HF patients and to examine the complex physiology of exercise in LVAD recipients. In this subgroup of patients, we aim to analyse the entangled interactions between the cardiopulmonary system, the musculoskeletal system, the autonomic nervous system, and the pump. Highlighting the current knowledge and describing the possible interventions that can be implemented will hopefully lead to a better understanding of this topic, eventually bringing better care to HF patients.
Advanced heart failure and non-invasive haemodynamic monitoring
As HF medical and electrical therapies progress (i.e. sacubitril/valsartan, SGLT2 inhibitors, implantable defibrillators, and cardiac resynchronization), patients are followed up for longer periods. These patients may need repeated invasive haemodynamic studies, but complications and adverse events related to the invasive procedure must be considered. Moreover, the invasive haemodynamic evaluation is usually performed at rest, rarely during exercise, and almost always in the supine position which is far away from what is done during regular activities by patients. On the other hand, CPET with cycle ergometer is the method of choice for measuring VO2, it can be associated with various techniques for the study of non-invasive haemodynamics, and it is safe and easy repeatable and frequently available.3
Non-invasive cardiac output monitoring
According to Fick’s law, QC is directly proportional to VO2 and inversely proportional to arteriovenous O2 difference (ΔC(a − v)O2). Two non-invasive methods for QC monitoring during exercise are best established:
CPET-QC with inert gas rebreathing (IGR) technique
IGR is the most common and reliable method. By using an oxygen-enriched mixture of an inert soluble gas as 0.5% nitrous oxide (N2O) and an inert insoluble gas as 0.1% sulfur hexafluoride (SF6) inflated into a bag rebreathing system connected to the patient via a mouthpiece, it is possible estimate QC at rest and at the peak of effort. At the end of an exhalation, the bag valve is activated automatically so patients start rebreathing from the prefilled bag for a period usually of 10–20 s; after that, they breath ambient air again. QC evaluation is performed by a photoacoustic analyser that measures gas concentration over a three to five breath interval. SF6, as an insoluble gas, is used to determine lung volume, while N2O, as a soluble gas, decreases during rebreathing with a rate that is proportional to pulmonary blood flow (PBF). QC is equal to PBF only if arterial oxygen saturation (SpO2) is up to 98%, indicating the absence of pulmonary shunts.4
CPET-QC with bioimpedance cardiograph technique (Physioflow)
Physioflow uses a high-frequency/low-amperage electrical current going through several electrodes placed to the chest. Basal thoracic impedance represents a steady state while changes in aortic blood flow during cardiac cycle induce opposing changes in electrical impedance. Thus, QC evaluation is dependent on the thoracic flow inversion time (FIT) measured as the first derivative of the impedance signal. FIT is defined as the interval between the first peak following the QRS to the electrocardiogram and the first zero after the nadir of the ejection velocity. Measurement is performed continuously, usually with time intervals between 15 and 20 s.5
Peripheral blood flow, O2 delivery, and muscular uptake monitoring
Near-infrared spectroscopy assessment (NIRS) technique allows the real-time measurement of the oxygenation status of human tissues which can be paired with CPET thus providing the direct quantification of the changes in muscular concentrations of oxygenated and deoxygenated haemoglobin (O2Hb and HHb, respectively), oxygenated and deoxygenated myoglobin (O2Mb and HMb, respectively), and total heme (tHbMb). Hb and Mb contain an iron core within each heme, which varies its light absorption in the near-infrared range based on whether or not oxygen is bound to it. Provided Mb is constant and changes in tHb can be interpreted as modification in muscular blood flow while differences between O2Hb and HHb reflect mainly muscle oxygen uptake. Usually, during a CPET, a NIRS sensor is applied above quadriceps muscle (vastus lateralis or rectus femoris) because it is the most metabolically active muscle when cycling.6
Advanced heart failure and exercise
Effort limitation in HF due to neurohumoral activation is a multifactorial process that involves alterations in central and peripheral haemodynamic regulation, anaemia and iron deficiency, pulmonary impairment such as congestion and or pulmonary hypertension, and muscular efficiency in O2 extraction. Therefore, by acting on these factors, it is possible to improve patient performance.
Peak QC in advanced heart failure
The correlation between VO2 and peak QC has a strong linearity. A reduced QC response during exercise is one of the early and major factors limiting exercise performance in patients with mild to moderate reduction of functional capacity (VO2 ≥ 50% predicted). However, in AdHF (VO2 < 50% predicted), this correlation seems weaker, making estimation of peak QC questionable in this setting.7 A partial explanation may lie in the fact that these patients are more anaemic and have lower iron levels. Accordingly, it is essential to evaluate iron profile in AdHF and perform a correction if the values fall below the recommended standards (Figure 1).7
Linear regression between peak exercise VO2 and CO in group 1 (A) (peak VO2 < 50% of predicted), group 2 (B) (peakVO2 ≥ 50% and <65% of predicted), and group 3 (C) (peak VO2 ≥ 65% of predicted). Reproduced with permission from Del Torto et al.7
Levosimendan and exercise in advanced heart failure
It is well known that long-term intermittent administration of levosimendan in AdHF outpatients leads to haemodynamic stabilization with prevention of re-hospitalization due to HF, preservation or improvement in NYHA class, enhancement of exercise tolerance, and, in palliative settings, preservation of health-related quality of life (QoL).8,9 Levosimendan effects are a relevant tool to understand exercise physiology in AdHF. Indeed, the determinants of the above-reported levosimendan–induced effects have been recently explored: Levosimendan showed to have pleiotropic effects on the heart, muscles, and lungs. However, these positive effects last for a short time, from a few days to several weeks, according to the drug’s pharmacokinetics10 (Figure 2).
Levosimendan infusion effects on exercise performance. Green circles: beneficial effects. Red circles: irrelevant effect. Peak VO2, peak oxygen uptake; VE/VCO2, minute ventilation to carbon dioxide production relationship slope; CO, cardiac output; t-Hb, total haemoglobin; O2-Hb, oxygenated haemoglobin; H-Hb, deoxygenated haemoglobin; DLCO, diffusing capacity of the lung for carbon monoxide; DLNO, diffusing capacity of the lung for nitric oxide; Vd/Vt, dead space to tidal volume ratio; PetCO2, end-tidal carbon dioxide pressure; Dm, membrane diffusion; Vcap, capillary volume. Reproduced with permission from Apostolo et al.10
Cardiac effects
As an inotropic agent, levosimendan achieved a significant dose-dependent improvement of QC and stroke volume (SV) both at rest and on exertion. Moreover, it displayed a diastolic positive effect with a reduction of pulmonary capillary wedge pressure (PCWP). Best markers of these mechanisms are parallel increments of peak QC − VO2 at CPET and natriuretic peptide reduction (i.e. BNP and NTproBNP).8
Muscular effects
In addition to the inotropic effect, as a calcium-sensitizing agent, levosimendan acts as an inodilator in vascular smooth muscle cells in various vascular beds either at rest or during exertion, allowing a more physiological blood flow distribution and less extreme diversion of blood flow to the working muscles during exercise. These statements arise from NIRS muscular evaluation during effort, which showed that at each level of exercise, tHb is greater after levosimendan infusion, with a higher O2Hb and an unchanged HHb suggesting an increased O2 delivery along with an increased O2 uptake by the muscle.10
Lung effects
Patients with AdHF accumulate fluids along the alveolar–capillary membrane during exercise. Levosimendan positively influences intrathoracic fluid movement during exercise and improves ventilator efficiency, as shown by the VE/VCO2 slope reduction. On the other hand, it has no major effects on lung function and ventilation/perfusion mismatch.9
Exercise as therapy in advanced heart failure
Besides being a valuable tool for diagnosis and prognosis, exercise is part of AdHF therapy; regular physical exercise guided by CPET parameters improves functional capacity, ameliorates QoL, and reduces hospitalizations due to HF. In this setting, it is useful to estimate the maximum exercise effort, as reaching this heart rate allows an accurate evaluation of functional capacity, the detection of ischaemia, and the definition of a rehabilitation plan.11 Although different training protocol regimens have been studied, such as high-intensity interval training or resistance training, aerobic exercise proved to be the most suitable in AdHF. Cattadori et al.12 showed that a short-term aerobic training programme was adequate to provide an increase in functional capacity in more than 90% of patients with HF. However, different physiological adaptations were outlined: (i) about half of the patients displayed an increase of QC and a reduction in a − v O2 difference, most likely due to a lower basal QC, advocating inadequate muscle perfusion and improvement in muscle function during several weeks of training (and that these patients would probably benefit from a longer training period); (ii) 21% of patients had a decrease in QC but an increase in a − v O2 difference and VO2, suggesting that a better performance is secondary to a greater ability to extract O2 from muscles; (iii) 23% manifested a parallel increase of QC and a − v O2 difference; and (iv) only 6% did have a clear worsening of performance due to a drop in QC, a − v O2 difference and hence VO2.12
Exercise in left ventricular assist device
LVADs have become a well-established therapy in AdHF patients, either as a bridge to transplant or as long-term (destination) therapy. Improved technology, with longer durability of devices and less complications have widened the indications for destination therapy and increased the number of patients with long-term device therapy. Improving exercise tolerance and hence quality of life has become the benchmark of care for these patients.
Cardiopulmonary exercise testing is the gold standard to evaluate exercise capacity in HF. Even after LVAD implantation, peak VO2 has a prognostic value, similar to HF patients not supported by mechanical circulatory support devices. similarly, VE/VCO2 maintains a prognostic value in these patients.13
Physiology of exercise in left ventricular assist device
In AdHF, exercise intolerance is due to a variety of causes in addition to QC decrease, such as respiratory abnormalities, muscular deconditioning, endothelial disfunction, blunted chronotropic response, ventilation/perfusion (V/Q) mismatch, anaemia, and reduced right ventricular function, which are not completely reversed in LVAD recipients. On the other hand, improvement in central haemodynamics in LVAD-supported HF patients may translate in an increase in skeletal muscle perfusion, reversal of peripheral neurohormonal hyperactivation, and enhancement of general condition and nutritional status, all of which may improve exercise tolerance.14 Irrespective of clinical stability, restoration of organ function, general improvement of the quality of life, and the capacity to perform everyday activities and light exercise, patients show a nearly trivial improvement in their peak VO2 after LVAD implantation, with a peak VO2 corresponding to an average of 50% of predicted.15
There are few data comparing exercise performance pre- and post-LVAD implantation because patients with AdHF are often too sick to perform a stress test. Exercise capacity before and after LVAD has been investigated in a few reports analysing a very limited number of cases, and results are contradictory.13 Nevertheless, increase in QC and peak VO2 seemed not to correlate to left ventricular ejection fraction, filling pressures, septal position, and aortic valve opening at rest.16 In a study in which 15 patients performed CPET during pre-LVAD evaluation and at least 4 months later, VO2 peak, exercise duration, and ventilatory efficiency slightly but significantly increased. Moreover, after LVAD, anaerobic threshold (AT) was identifiable in a larger number of patients and VO2 at improved AT. AT is strictly related to everyday life, and this is in line with improvement in the quality of life reported by patients.14 Also, periodic breathing, a marker of disease severity and an important prognostic index, was detectable in less patients, while another study comparing HF patients with and without LVAD showed a similar percentage of exercise oscillatory ventilation.17,18
Cardiovascular system
In normal subjects, the physiological response to exercise is an increased heart rate, increased contractility, and subsequently an increased QC. In HF patients, these mechanisms are impaired. Chronotropic insufficiency is not reversed after LVAD.19QC in LVAD patients is determined by flow through the pump and ejection through the aortic valve, according to left ventricle residual contractility. Flow through the device depends on pressure difference across the pump and on LVAD speed, which is usually fixed and adjusted by the physician. Adjusting the speed allows the highest QC without an excessive left ventricular unloading and/or right ventricle loading.
LVAD flow is also highly dependent on afterload. Systemic vascular resistance (SVR) decreases significantly during exercise in LVAD patients, due to exercise-induced peripheral vasodilation,20 even though there are other factors contributing to afterload changes, such as aortic stiffness and blood volume. QC also depends on right ventricular function.
After LVAD, many patients show left ventricular reverse remodelling and this may improve contractility. One study on 30 patients showed that the effect of speed reduction was minimal in patients with LV ejection fraction > 40% while in patients with a lower ejection fraction, there was a significant drop in peak VO2.21
Right ventricular function
Evaluation of RV function in LVAD patients is complex, and multiple scores exist to predict post-implantation RV dysfunction. It affects up to 40% of patients over time22 and significantly impairs the quality of life and survival.23 Some degree of dysfunction is common in AdHF, and this could have impact on QC during exercise. There are small studies suggesting that RV function may not be a limiting factor for exercise capacity in LVAD patients. Some studies showed no correlation between right ventricular dysfunction evaluated with echo parameters at rest and peak VO2.24 Invasive haemodynamic studies showed an increase in right atrial pressure and wedge during exercise suggesting that right heart filling pressure increase is secondary to left ventricular filling pressure increase.25 This was confirmed by another study in which during exercise, a reduction in right ventricular volume and an increase in right ventricular SV were observed.26 In a more recent study27 conducted on 26 patients with LVAD, authors performed a complete invasive haemodynamic assessment (one group with conductance catheter and the other with traditional Swan–Ganz catheters) during cardiopulmonary exercise testing. There was a modest increase in RV contractility during submaximal exercise, but there was no consistent increase at peak effort. Moreover, there were large increases in pulmonary arterial, left-sided filling and right-sided filling pressures during submaximal and peak exercise. These findings might explain the mechanism for persistent reductions in functional capacity in these patients.
Respiratory system
In HF patients, there is an increase in minute ventilation relative to gas exchange due to early anaerobic metabolism, reduced respiratory muscle strength,28 impaired alveolar capillary diffusion,29 and restrictive physiology. The prevalent cause of hyperpnoea though is increased physiological dead space due to alveolar hypoperfusion.30 In LVAD patients, the reduction of pulmonary congestion and better perfusion of respiratory muscles should improve respiratory function. A small study showed a lower grade of restriction at spirometry over time.31 Though, another study found physiological dead space in LVAD patients to be similar to dead space in AdHF patients.32
Musculoskeletal system
There are conflicting results regarding the musculoskeletal system after LVAD placement. One study noted a three-fold increase in blood flow to exercising muscles.33 Other studies noted a lack of change in blood flow to exercising muscles and no change s in muscle endurance.15
Sympathetic nervous system
LVAD implant can restore baroreflex activity, decrease plasma norepinephrine and dopamine levels, and improve sympathetic function, thus affecting the nervous system during exercise.34
LVAD speed changes
LVAD QC might be sufficient at rest or during light exercise but could be insufficient to guarantee a near maximal effort. The effect of LVAD speed increase on exercise performance has been investigated with discrepant results, showing an increase in VO2 peak with increased pump speed24 elevated filling pressures and markedly diminished SvO2 persist at maximal safe pump speed (hayward impact of left ventricular). These last data are confirmed in a study of increase in LVAD speed with invasive haemodynamic evaluation at rest and during exercise in which increase in pump speed was associated with increases in both right- and left-sided filling pressures during exercise.35 Another study including 20 patients with LVAD confirmed an increase in QC at rest and during exercise (measured with IGR) with a higher pump speed, and this translated in an increased peak VO2, improved ventilatory efficiency (decreased VE/VCO2 slope), and postponed anaerobic threshold.36 The postponed anaerobic threshold is justified from a physiological point of view by the fact that during progressive exercise, VO2 is independent of QC up to the anaerobic threshold and becomes QC dependent above it. The anaerobic threshold is therefore strictly dependent on QC and correlates more with exercise tolerance in everyday activity (Figure 3).36 The challenge of allowing the patients to change their device speed is fascinating. Though, this might require some caution and deeper investigations on physiological changes that might be produced by LVAD speed changes.
![Graphic representation of the Fick principle at rest (white symbols) and at peak exercise (black symbols) with pump speed at 2 (squares) and at 4 (circles). Pump speed increases cardiac output (CO, Y-axis) both at rest and peak exercise, oxygen uptake (VO2) at peak exercise but not arteriovenous O2 differences [Δ(Ca − Cv)O2] (X-axis). ⁎P = 0.01 vs. speed 4 at peak exercise. §P b 0.01 vs. speed 4 at peak exercise. #P b 0.05 vs. speed 4 at rest. Reproduced with permission from Vignati et al.36](https://oup.silverchair-cdn.com/oup/backfile/Content_public/Journal/eurjpc/30/Supplement_2/10.1093_eurjpc_zwad239/2/m_zwad239f3.jpeg?Expires=1748104993&Signature=lBsg2xOtLXDRpxh0n6k5SeaZ5Galmz~tfnV7xox4c3SMa9hbWUGzNzukRUsvONlg9aonhW7dtxSulAQZyL6Vn03XNYWJ2avsfo49EIoYBxnzvTbiEQWI3CaSPMp4jh648kop7mTsMdsdxBGQsPdUUOPxrGFWqVaAKKYbaNzT4tHiMQdQwrMlLj5rRBIeFxVHKX~Wah9rKEFWUzYdyhBt8zrMgGjMED28tNuU~NtXb5QPyL3HmteYZgN0XCH6q4Kc~dzr1RXNh1DuJfHFEt5EIpGHR1~Iu1bhHEzvu8u1GP3DHgwyvuZYGlefhKGMaTjP-YDTXzt4TlQ0SpsEMa6vzw__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
Graphic representation of the Fick principle at rest (white symbols) and at peak exercise (black symbols) with pump speed at 2 (squares) and at 4 (circles). Pump speed increases cardiac output (CO, Y-axis) both at rest and peak exercise, oxygen uptake (VO2) at peak exercise but not arteriovenous O2 differences [Δ(Ca − Cv)O2] (X-axis). ⁎P = 0.01 vs. speed 4 at peak exercise. §P b 0.01 vs. speed 4 at peak exercise. #P b 0.05 vs. speed 4 at rest. Reproduced with permission from Vignati et al.36
One comprehensive study investigated in 33 LVAD patients the effects of pump speed increase on analysing CPET parameters, IGR QC, NIRS during incremental, and constant workload.37 It has also been investigated how these changes influence lung fluid balance and sleep disorders. Sleep alterations with central and obstructive apnoea are frequent in severe HF patients and correlate with a poor quality of life and prognosis. The comparison between CPET at fixed-speed and incremental LVAD-speed CPET showed an improvement of functional capacity expressed by an increase in peak VO2, higher values of the VO2/work relationship and O2 pulse, and a reduction of the VE/VCO2 slope. During constant workload CPET with two different speeds, a significantly improvement in QC and a better O2 kinetic expressed as Tau (time constant) value reduction have been observed. The increase in QC during exercise is associated with an improvement of oxygen delivery to the exercising muscle. In fact, NIRS analysis showed increased O2 HB and tHB at rest and at each phase of exercise with increased speed. HHb showed no difference, indicating an increased peripheral extraction at increased LVAD speed. Lung diffusion was very low in all patients, and increasing LVAD speed for a prolonged time was associated with further reduction of DLCO, suggesting thoracic fluid accumulation.
Sleep disorders are frequent in HF patients and are associated with poor prognosis. LVAD speed increase during sleep reduces AHI by reducing central sleep apneas which are QC dependent, but it also slightly increases obstructive sleep apnea syndrome which are likely related to an increase in right atrial pressure and to the consequent upper airway fluid content.38
Rehabilitation
Exercise training beneficial effects on functional capacity and prognosis are well known in HF patients.39 As reported above, LVAD patients show a peak VO2 of about 50% of the predicted value. Exercise training might improve functional capacity in these patients who are implanted in most cases in a severe deconditioned status. Since 2011, the feasibility of exercise training in LVAD-bearing patients40 and the consequent improvement in exercise tolerance and peak VO2 have been demonstrated. Later studies have confirmed the beneficial effect of exercise training on exercise capacity and quality of life in LVAD patients, and longer programmes41 seem more advantageous then short programmes.42 Furthermore, a trial demonstrated that patients randomized to exercise training or standard care showed improvement of 10% of peak VO2 in the trained group.43 Despite these small studies have reported encouraging results, more studies are needed to indicate the most effective training programmes, intensity of exercise, interval and number of sessions, and different modalities.
Conclusions
CPET allows a holistic and multi-parametric assessment of patients with AdHF. Specifically, the assessment of the pathophysiological changes in the body affected by a complex syndrome such as HF allows tailoring the treatment of each patient and prescribing appropriate rehabilitation therapies. The periodic repetition of CPET and possibly of complex CPET allows even in the most severe patients the early identification of clinical worsening and is fundamental in the follow-up of this peculiar HF population.
Author contributions
All authors contributed in the manuscript drafting and revision of the text.
Funding
Publication fees are supported by Fondazione IEO-Monzino.
References
Author notes
Conflict of interest: None declared.
Comments