Physiology of exercise and heart failure treatments: cardiopulmonary exercise testing as a tool for choosing the optimal therapeutic strategy

Abstract

In the last decades, the pharmacological treatment of heart failure (HF) become more complex due to the availability of new highly effective drugs. Although the cardiovascular effects of HF therapies have been extensively described, less known are their effects on cardiopulmonary function considered as a whole, both at rest and in response to exercise. This is a ‘holistic’ approach to disease treatment that can be accurately evaluated by a cardiopulmonary exercise test. The aim of this paper is to assess the main differences in the effects of different drugs [angiotensin-converting enzyme (ACE)-inhibitors, Angiotensin II receptor blockers, β-blockers, Angiotensin receptor-neprilysin inhibitors, renal sodium-glucose co-transporter 2 inhibitors, iron supplementation] on cardiopulmonary function in patients with HF, both at rest and during exercise, and to understand how these differences can be taken into account when choosing the most appropriate treatment protocol for each individual patient leading to a precision medicine approach.

In the last decades, the pharmacological treatment of heart failure (HF) become more complex due to the availability of new highly effective drugs. Currently, numerous pharmacological classes of drugs have been approved for HF treatment including βs, angiotensin-converting enzyme (ACE) inhibitors angiotensin II receptor blockers, angiotensin receptor-neprilysin inhibitors (ARNI), mineralocorticoids and the recently introduced the sodium-glucose co-transporter (SGLT2) inhibitors (dapaglifozin and empaglifozin). All provide an improvement in HF with reduced ejection fraction (HFrEF) outcome superior to the best therapy previously available. European Society of Cardiology (ESC) guidelines recommended for HFrEF the combined use of ACE-I/ARNI, β-blockers, mineralocorticoid receptor antagonist (MRA), and SGLT2.¹ Although the cardiovascular effects of HF therapies have been extensively described, less known is their effects on cardiopulmonary function considered as a whole, both at rest and in response to exercise. This is a ‘holistic’ approach to disease treatment that can be accurately evaluated by a cardiopulmonary exercise test (CPET). Thus it is important to assess the main differences in the effects of different drugs on cardiopulmonary function in patients with HF, both at rest and during exercise, and to understand how these differences can be taken into account when choosing the most appropriate treatment protocol for each individual patient leading to a precision medicine approach.

In a high percentage of patients with HF lung function, considered as a combination of lung mechanics, lung diffusion and gas exchange capacity, as well as ventilatory response to exercise, are impaired so that the greater the lung function impairment the greater the HF severity.^2–4 Usually, a lung-restrictive disease is detected, frequently associated with a reduction in lung diffusion as shown by a reduced alveolar-capillary diffusion for carbon monoxide (DLCO).⁵ Lung diffusion deterioration can be attributed mostly to impairment in the alveolar-capillary membrane resistance, whereas the diffusion resistance imparted by the capillary volume component is highly mutable. DLCO and membrane diffusion impairment correlate with the severity of HF, as was demonstrated by splitting up patients into groups with progressively lower maximal exercise oxygen uptake capacity (VO₂) Figure 1.⁶ Moreover, membrane diffusion impairment stratifies HF patients prognosis.⁷ Lung function impairment results in an exaggerated ventilatory response to exercise, represented by an increase in the steepness of the relationship between ventilation and carbon dioxide production during exercise (VE/VCO₂ slope). The connection between exercise hyperventilation and HF severity is highlighted by the lower survival observed in patients with the highest VE/VCO₂ slope.⁸ Indeed, many studies performed in the last 25 years at Centro Cardiologico Monzino highlighted how lung function impairment may become a target of HF therapy allowing a precise match between respiratory impairment of HF patients and specific drug characteristics. Below we will review the cardiorespiratory effect of each class of currently recommended HF drugs.

ACE-inhibitors

The earliest works evaluating the effects of treatment with ACE-inhibitors on pulmonary function in HF dates back to the 1990s- A study by Guazzi has pointed out how ACE inhibitor–mediated improvement in subjective well-being and exercise tolerance should at least partially be referred to as an amelioration in lung function. In fact, treatment with enalapril in stable HF patients improved lung diffusion at rest and ventilatory efficiency and peak VO₂ during exercise. Another study was performed in 1997⁹ where HF patients and normal volunteers were treated with enalapril (20 mg/d) or enalapril plus aspirin in a random cross-over design. Although no significant effects were detected in normal subjects, in HF patients a significant increase in lung diffusion (DLCO), exercise tolerance, and peak VO₂ and a decrease in peak dead space/tidal volume (VD/VT) and VE/VCO₂ slope were observed after enalapril therapy. Interestingly, the beneficial effects of ACE inhibition were lost if enalapril was associated with aspirin. Moreover, in a subgroup of patients, haemodynamic recording of cardiac output and wedge pressure was obtained after enalapril or enalapril plus aspirin treatment as well as after vasodilating therapy with hydralazine–isosorbide dinitrate. An increase in cardiac index and reduction in wedge pressure appeared more evident after acute vasodilation therapy than after ACE inhibition; however, treatment with hydralazine–isosorbide dinitrate did not exert any influence on lung diffusion or ventilation efficiency during exercise. Putting all these results together, it was hypothesized that ACE-inhibitors are able to ameliorate gas exchange impairment at the alveolar-capillary level thanks to an increase in prostaglandin concentration, with a favourable effect on both local vessel tone and microvascular permeability. This interpretation found confirmation in a further study,¹⁰ which observed: (i) lung diffusion after ACE inhibition is specifically due to an increase in the membrane diffusion subcomponent of DLCO; (ii) DLCO and membrane diffusion effects were detectable after 1 month of enalapril therapy, but not after 48 h, suggesting that haemodynamic changes are hardly involved; (iii) increase in membrane diffusion is maintained even after normalization for alveolar volume, proving that lung diffusion improvement is not merely due to an increase in the surface available for gas exchange. The counteracting effect of aspirin on the ACE inhibitor benefit in HF was further demonstrated after adding aspirin at a daily dose of 325 mg in patients affected by primary dilated cardiomyopathy. A reduction in peak VO₂, ventilatory efficiency during exercise, and lung diffusion in patients treated with an ACE inhibitor was observed, but there was no effect in patients not taking ACE-inhibitors.¹¹ Moreover, a worse prognosis was observed in ACE-inhibited patients treated with a high-dose of aspirin in comparison with low-dose or no aspirin treatment.¹² On the other hand, ACE inhibition had no effects on mechanical lung parameters, whereas effects on lung diffusion and exercise were clearly observed after 6 months of therapy and were still evident at the evaluation at a 12-month follow-up.¹³ Finally, we showed that ACE inhibition benefit on lung function in HF is modulated by genetic polymorphisms. Indeed, vulnerability to acute fluid overload appeared to be significantly higher in a double-deletion genotype than in double insertion and insertion–deletion genotypes of an ACE insertion/deletion polymorphism in patients with HF treated with enalapril at the highest tolerated dose.¹⁴

Angiotensin II receptor blockers (sartanics)

The effects of sartanics on pulmonary function are different from those of ACEI inhibitors albeit with a similar clinical impact. In fact, the effects of ACE-inhibitors on the lungs are not due primarily to a reduction in angiotensin concentration but to an increase in bradykinin and, as a consequence, in vasodilating prostaglandins, as confirmed by the counteracting effect of aspirin. In a population of HF naive to ACE-inhibitors, angiotensin II receptor blockers, and aspirin was randomized to receive placebo, enalapril, losartan, or each of the two drugs plus aspirin in a cross-over design. A significant increase in exercise tolerance and peak VO₂ was observed both after enalapril and losartan, but it was counteracted by aspirin only in the enalapril group. Moreover, an increase in lung diffusion was observed only in the enalapril group.¹⁵ A following study demonstrated a synergistic effect of enalapril and losartan on exercise performance. Enalapril and losartan alone induced a similar increase in peak VO₂, but only enalapril was able to improve ventilatory efficiency (as shown by a significant reduction in VE/VCO₂ slope and peak VD/VT). Instead, losartan treatment, but not enalapril treatment, was associated with a significant increase in the slope of the relationship between VO₂ and work, suggesting activation of exercising muscle perfusion. Likely as a consequence of their different mechanisms of action, the association of enalapril and losartan induced a significantly higher increase in peak VO₂ than that obtained by each treatment alone, probably mirroring a synergistic activity on lung function and muscle perfusion.¹⁶

β-blockers

β-blockers are a cornerstone therapy in HF, being able to improve survival, increase left ventricular (LV) contractility and ameliorate patients well-being. However, data are inconclusive about changes in exercise since most studies were unable to demonstrate any significant increase in β-blocker-induced exercise performance.

Starting from this unexpected observation, several studies were carried out at the Heart Failure Unit of Centro Cardiologico Monzino to evaluate the cardiorespiratory response to β-blockers therapy.

At the beginning of the new millennium, a few studies investigated the effect of carvedilol treatment in HF. In the first study, 15 HF patients were treated with a placebo or carvedilol for 4 months following a cross-over double-blinded design.¹⁷ Carvedilol did not affect pulmonary mechanics or exercise capacity, both peak workload and VO₂ albeit it improved Minnesota quality of life score and reduced the ventilatory response to exercise. The latter was shown by a reduction in peak ventilation, ventilation during the steady state of a constant workload exercise, and ventilation during the intermediate phase of a ramp protocol exercise as well as by a reduction of the VE/VCO₂ slope, the latter reduction being particularly evident in patients with high value when on placebo. Notably, VE/VCO₂ slope reduction correlated with Minnesota score improvement, suggesting that the amelioration in patients well-being could be somehow related to the reduction in the inappropriately elevated ventilatory response to exercise. Accordingly, this study provided the rationale for the gap between clinical and prognostic carvedilol-induced improvement and absence of exercise performance changes by showing the positive carvedilol-induced effects on ventilation regulation. In a subsequent study,¹⁸ the blunting effect of carvedilol on ventilation during exercise in HF was studied both in normal conditions and at a simulated altitude of 2000 m above sea level, obtained by making the patients breath a gas mixture with a 16% FiO₂. Ventilation was reduced by carvedilol in comparison to placebo even in hypoxic conditions, both during a ramp protocol and during a constant workload exercise. Moreover, in the latter experimental conditions, it was possible to evaluate blood gas behaviour by collecting arterial blood samples at rest and during the steady state phase of exercise. Interestingly, at simulated altitude, arterial pO₂ during exercise resulted significantly lower when patients were treated with carvedilol than when they were treated with placebo. Therefore, the reduced ventilatory response to exercise-induced by carvedilol is probably advantageous in normoxic conditions, but becomes detrimental at altitude, when an increase in ventilation is a pivotal compensatory mechanism to hypoxia. However, it must be underlined that, at present, we do not have scientific evidences that the reduction in hyperventilation is responsible for the improved survival by carvedilol. An even more interesting result of this study derived from the analysis of lung diffusion at rest. Indeed, on carvedilol, DLCO was significantly lower than on placebo due to a reduction of membrane conductance. This new observation paved the way for the hypothesis of a direct influence of β receptor blockade on alveolar-capillary membrane function.

All these studies were performed using carvedilol as a β-blocker but what happened with different β-blockers? Are these class effects or are they peculiar to carvedilol? This topic was firstly addressed in a retrospective work comparing patients treated with carvedilol (n = 304), or with bisoprolol (n = 187), or non-treated with β-blockers (n = 81).¹⁹ Patients without β-blockers and patients treated with bisoprolol shared the same ventilatory behaviour, that is equal values of VE/VCO₂ slope and of PetCO₂ during the isocapnic buffering period. On the contrary, patients in carvedilol had a significantly lower VE/VCO₂ slope and a significantly higher PetCO₂ during the isocapnic phase than the other two groups. This suggests that only carvedilol inhibits the ventilatory response to exercise in HF, probably acting on chemoreceptor sensitivity.

Carvedilol and bisoprolol were compared also in a prospective study with a double-blind cross-over design²⁰ in 53 HF patients. All patients performed pulmonary function tests, lung diffusion measurements with subcomponents determination, and a maximal CPET. Information added by this experimental setting were the following: (i) lung diffusion (in particular membrane conductance subcomponent) was much more impaired on carvedilol than on bisoprolol, (ii) exercise capacity was significantly lower during carvedilol treatment than during bisoprolol treatment, and (iii) differences in peak VO₂ was particularly evident in patients with an altered DLCO, suggesting that functional capacity reduction was mainly a consequence of lung diffusion impairment. Since the major difference between carvedilol and bisoprolol consists in β receptors selectivity (bisoprolol is 120 folds more selective towards β1 receptors than carvedilol), it has been postulated that the different behaviour of these two drugs could be related to a different degree of blockage of β2 receptors.

To better investigate this hypothesis, we conducted a further study to compare unselective and β1 selective β-blockers. In the CARNEBI trial²¹ carvedilol (unselective), nebivolol, and bisoprolol (both highly β1 selective) were compared in a cross-over, randomized trial on 61 HF patients. In agreement with the hypothesis that carvedilol detrimental effects on lung diffusion are related to the blocking of alveolar β2 receptors (that are known to be involved in alveolar oedema clearance), we observed that lung diffusion was significantly lower on carvedilol than on bisoprolol or nebivolol, once again because of a reduction in membrane conductance subcomponent. On the contrary, ventilatory efficiency was significantly better on carvedilol than on bisoprolol or nebivolol, as proved by a lower VE/VCO₂ slope. This beneficial effect on ventilatory efficiency is hardly reconcilable with the impairment in lung diffusion, which is usually associated with a worsening in VE/VCO₂ slope. In the CARNEBI trial, we, therefore, tested the hypothesis that the blunting effect of carvedilol on ventilation could be due to interference with chemoreceptor sensitivity, which is under control of α, β1, β2, and NO receptors. Accordingly, we measured CO₂ and O₂ sensitivity during different β-blockers treatments. Results were consistent with the starting hypothesis, as carvedilol patients showed a lower sensitivity both to CO₂ and O₂ compared to, while an intermediate sensitivity was observed on nebivolol (Figure 2).

To complete our knowledge of β-blockers effects on cardiorespiratory function we extended our research to healthy subjects.²² Healthy volunteers were treated for a few days with carvedilol or bisoprolol. Even in this setting, carvedilol but not bisoprolol shows a detrimental effect on lung diffusion, both in normal conditions and after fluid challenge with rapid acute saline infusion (25 mL/Kg) used as a pulmonary oedema model. Differently from what was observed in HF patients, VE/VCO₂ slope increased in carvedilol but not in bisoprolol-treated subjects, suggesting that, in normal conditions, impairment in lung diffusion translates into impairment in ventilatory efficiency as well. In HF this effect is probably counteracted by the interference with chemosensitivity.

The observation that blocking the alveolar β2 receptors with carvedilol is detrimental in HF patients because of an impairment in lung diffusion, raised the hypothesis that stimulation of the same receptors could be conversely somehow beneficial. To explore this hypothesis we investigated the effect of the treatment with a topical β2 agonist (indacaterol) in patients with HF treated with β-blockers with different β1–β2 selectivity (carvedilol or bisoprolol) in a double-blind, placebo-controlled trial.²³ The study didn’t show any significant effect of indacaterol treatment on lung diffusion regardless of the treatment with bisoprolol or carvedilol. However, we do not know if a significant concentration of the β agonist actually reached its target receptor in the alveola, as these drugs are built in order to exert their bronchodilation effect in the bronchi, and not reach the alveola, where they can be systemically absorbed causing undesired effects.

It must be emphasized that these experimental observations are not usually applied in the clinical field so we do not know whether these findings could help the individual HF patient. However, we believe that it is worth testing the clinical relevance of our findings to assess whether the most appropriate match between β-blocker and individual HF patients has true clinical benefits.

Mineralocorticoid receptor antagonists

Another class of drugs recommended for the treatment of HF patients is MRAs. It is demonstrated that these drugs improve HF status and prognosis by counteracting the effect of aldosterone which is increased in chronic HF and has several properties which are likely to be detrimental in HF. In particular, aldosterone has been related to extracellular matrix turnover increase which is associated with cardiac, kidney, and lung fibrosis. Focalizing our attention on the lung it is known that in HF patients the alveolar-capillary membrane undergoes chronic changes in HF which are associated with an increase in its fibrotic content and lung diffusion abnormalities, frequently observed in HF patients. Both are related to a reduction in exercise performance and poor prognosis.²⁴ A study performed in 2017²⁵ showed a positive effect of spironolactone on gas diffusion through the lung and exercise capacity. The suggested mechanism by which anti-aldosteronic drugs improve HF clinical condition and prognosis might be the following: (i) a more complete modulation of neurohumoral activation when spironolactone is added to standard HF treatment that exerts its effect on the heart or on the peripheral vasculature; (ii) restoring, at least in part, gas diffusion in humans improving a damaged alveolar-capillary membrane; and (iii) counteracting the interstitial fibrosis promoting by high aldosterone levels possibly through local dehydration.²⁵

Angiotensin receptor-neprilysin inhibitors

Sacubitril/valsartan, a first-in-class angiotensin receptor-neprilysin inhibitor (ARNI), is one of the novel therapies recommended by the ESC guidelines¹ for the treatment of HF patients with reduced ejection fraction (HFrEF) to reduce mortality and HF hospitalizations. In particular, sacubitril is a prodrug that, upon activation, inhibits the endopeptidase neprilysin, which cleaves vasoactive peptides such as B-type natriuretic peptide, while valsartan inhibits the angiotensin II receptor. Through these actions, the drug combination dilates blood vessels and reduces extracellular fluid. Recent studies showed a drug-induced improvement in exercise tolerance assessed by a 6-min walk test (6-MWT) in patients with HFrEF.²⁶ One of the first reports showing an amelioration of CPET parameters was published by our group in 2019,²⁷ and a previous non-randomized trial by Vitale et al.²⁶ evaluated the effect of this drug on exercise performance. Ninety-nine ambulatory patients with HFrEF underwent serial CPETs after initiation of sacubitril/valsartan on top of their recommended HF therapy. After a median follow-up of 6.2 months peak oxygen consumption (VO₂) improved, minute ventilation/carbon dioxide production relationship (VE/VCO₂ slope) decreased, VO₂ at anaerobic threshold increased, and oxygen pulse increased, and so did ΔVO₂/Δwork. This might result in a net improvement in exercise tolerance and performance. Based on these preliminary results, we speculated that sacubitril/valsartan might have a synergistically favourable effect on hemodynamics and muscle efficiency through reduced afterload and LV filling pressure and that the beneficial effect on autonomic function may translate to a more efficient and coordinated ventilation pattern. Therefore, looking beyond exercise performance, we prospectively investigated the mechanism behind the favourable effects of the drug. In a study conducted by our group on 79 HFrEF patients (86% males, age 66 ± 10 years),²⁸ it was evident that sacubitril/valsartan effects follow a double pathway (i.e. haemodynamic and non-haemodynamic) with an improvement both on values more related to a haemodynamic (diuretic? ) mechanism (i.e. Forced expiratory volume in the first second, and E/e′ values as a marker of diastolic dysfunction) and a pleiotropic effect, as demonstrated by LV reverse remodelling, and by an amelioration in lung diffusion capacity (DLCO) (Figure 3A). The drug’s effect on biomarkers also seems to confirm this hypothesis (Figure 3B), with a significant reduction both in HF biomarkers correlated with filling pressures and fluid congestion (NTproBNP, immature surfactant binding protein isoform B (proSP-B),^29–35 and in those more linked to non-haemodynamic mechanisms (surfactant binding protein isoform D (SP-D), ST-2). Further studies are ongoing to understand how the behaviour of these changes could different over time, and preliminary results (not yet published) seem to suggest a more rapid haemodynamic effect (as shown by a fast decrease in NT-proBNP also in patients treated with a low-dose of sacubitril/valsartan), and a slower pleiotropic, anti-fibrotic effect (as shown by ST-2 decrease) only evident after months of high-dose therapy. Moreover, our experience confirmed how these results were paralleled by a significant reverse LV remodelling with a decrease in both 2D and 3D LV volumes and by an improvement in diastolic dysfunction and filling pressures.³⁶ The effects were consistent regardless of aetiology, even if more pronounced in non-ischaemic HFrEF patients.

Taken together these data confirm the benefits of sacubitril/valsartan therapy, far beyond the world in randomized clinical-trials. In this regard, a recent real-file study on a population of 201 consecutive patients enrolled in Centro Cardiologico Monzino has identified that more than one-third of the patients were able to tolerate the higher dose of sacubitril/valsartan and that these patients were younger, had higher Hb, and better blood pressure and renal function.³⁷ Interestingly, MECKI score stratification was useful to discriminate patients who continued treatment from those who did not.

Renal sodium-glucose co-transporter 2 inhibitors

The renal sodium-glucose co-transporter 2 inhibitor (SGLT2i) class has recently been approved as an additional mainstay of prognostic-modifying therapy for HFrEF, in addition to the gold-standard treatment of the triple neurohormonal blockade strategy. Dapagliflozin and empagliflozin, two SGLT2 initially approved for the treatment of patients with type 2 diabetes mellitus (DM) have been shown to result in a significant improvement in outcome in patients with HF.^38–40

The DAPA-HF study, a multicentre, randomized trial, demonstrated a significant reduction in the incidence of HF-related hospitalizations, cardiovascular mortality, and all-cause mortality in patients with HFrEF, regardless of being diabetic.⁴¹ These findings were confirmed by the EMPEROR-reduced study.⁴²

Two recent studies demonstrate a significant improvement in exercise capacity in patients treated with SGLT2i: in the first⁴³ an increase of 3.1 mL/kg/min in pVO₂ in the treated patients compared to a minimal increase of 0.1 mL/kg/min in the control group (P = 0.030) and an improvement in VE/VCO₂; in the second EMPA-TROPISM,⁴⁴ pVO₂ increased significantly in the treated HFrEF non-diabetic patients by 1.1 mL/kg/min compared to a decrease of 0.5 mL/kg/min in the placebo, which is in line with the improvement in functional capacity. The VE/VCO₂ slope was found to have a non-significant trend of improvement in the empagliflozin group, whereas dapagliflozin resulted in a statistically significant improvement. These values measured by CPET contrast with the neutral results of the EMPERIAL⁴⁵ studies regarding the use of empagliflozin over the 6-MWT distance. However, the choice of the 6-MWT as the primary endpoint could be the reason for the neutral results, as it is not an optimal measure of health status improvement in patients with HF, being influenced by many HF-related comorbidities and lacking reproducibility.

Iron deficiency

Iron deficiency (ID) occurs as a comorbidity in more than 60% of HF patients either in the absence or presence of anaemia. Indeed, iron plays on top of erythropoiesis a key role in several energy production reactions of cells, regulating the enzymatic activity of the Krebs cycle and respiratory chain in mitochondria, particularly in highly metabolically active cells such as skeletal and cardiac myocytes. ID is associated with different symptoms, such as fatigue or tiredness, exercise performance reduction, and poor prognosis. The current most widely accepted definition of ID, which has been adopted by several society guidelines, including ESC, American Heart Association, HF Society of America and American College of Cardiology requires serum ferritin <100 µg/L or ferritin between 100 and 299 µg/L and transferrin saturation (TSAT) < 20%. Recently several studies have drawn attention to the clinical condition characterized by TSAT < 20% independently from ferritin levels, reported as impaired iron transport (IIT) and related to a reduction of iron availability for body functions. IIT results are associated with a worse clinical profile and increased risk of mortality compared to patients with normal iron status^46,47 in HF patients and to a reduced exercise capacity, even compared to those with absolute ID.^46–50 Specifically, IIT is associated with a reduced VO₂ peak and a VE/VCO₂ slope increase (Figure 4).⁵¹ Of note peak VO₂ and VE/VCO₂ are the most relevant tools for HF assessment and prognosis determination.^52–54 Of note VE/VCO₂ increase in HF patients with ID is likely due to the increased sympathetic tone which recognized among its causes an ID-related impairment of mitochondrial respiration of cardiomyocytes and skeletal muscle cells.

In light of this evidence, iron carboxymaltose supplementation therapy has become part of the therapeutic strategies recommended by the guidelines. Confirming these data in a recent study in patients with HF, anaemia and ID, intravenous ferric carboxymaltose improves the hypercapnic ventilatory response and sleep-related breathing disorders, peak VO₂ and VO₂/work while it did not show any relevant change on placebo.⁵⁵ These described effects, along with better peripheral oxygen delivery, can contribute to explain the global clinical improvement observed in anaemic HF patients after proper iron replenishment.

Conclusion

This pathophysiological analysis of the HF drugs effects emphasises that a complete evaluation of patients, including not only cardiovascular status, but also lung function, both at rest and during exercise, is needed for a rational choice of the most appropriate pharmacological therapy. Specifically, all HF patients should undergo complete spirometry with alveolar-capillary membrane function analysis and a symptom-limited CPET. Indeed, albeit at different levels, all the HF-recommended treatment influence, directly or indirectly, the respiratory function and its regulation. Therefore, we believe that coupling of patients and drug characteristics allows a specific drug tailoring for each individual HF patient in line with the need for a precision medicine approach.

Authors contribution

J.C., M.C., and M.M. drafted the manuscript. I.M. and P.A. critically revised the manuscript. All authors gave final approval and agree to be accountable for all aspects of the work ensuring integrity and accuracy.

References

Author notes