Abstract
Heart failure (HF) is accompanied by major neuroendocrine changes including the activation of the natriuretic peptide (NP) pathway. Using the unique model of patients undergoing implantation of the CARMAT total artificial heart and investigating regional differences in soluble neprilysin (sNEP) in patients with reduced or preserved systolic function, we studied the regulation of the NP pathway in HF.
Venous blood samples from two patients undergoing replacement of the failing ventricles with a total artificial heart were collected before implantation and weekly thereafter until post-operative week 6. The ventricular removal was associated with an immediate drop in circulating NPs, a nearly total disappearance of circulating glycosylated proBNP and furin activity and a marked decrease in sNEP. From post-operative week 1 onwards, NP concentrations remained overall unchanged. In contrast, partial recoveries in glycosylated proBNP, furin activity, and sNEP were observed. Furthermore, while in patients with preserved systolic function (n = 6), sNEP concentrations in the coronary sinus and systemic vessels were similar (all P > 0.05), in patients with reduced left-ventricular systolic function, sNEP concentration, and activity were ∼three-fold higher in coronary sinus compared to systemic vessels (n = 21, all P < 0.0001), while the trans-pulmonary gradient was neutral (n = 5, P = 1.0).
The heart plays a pivotal role as a regulator of the endocrine response in systolic dysfunction, not only by directly releasing NPs but also by contributing to circulating sNEP, which in turn determines the bioavailability of other numerous vasoactive peptides.
Introduction
Heart failure (HF) is accompanied by the activation of the natriuretic peptide (NP) pathway. Increased intra-cardiac pressures promote the release of the NPs from cardiomyocytes into the bloodstream. While atrial NP (ANP) is released mainly by the atria and brain NP (BNP) mainly by the ventricles in normal individuals, ANP and BNP are both produced at a larger extent by the ventricles in patients with reduced systolic function.1 Circulating NPs may undergo post-translational modifications such as threonine 71 (T71)-glycosylation of proBNP, cleavage by the circulating proBNP-convertase furin, and breakdown by neprilysin (NEP), amongst others.2 , 3 , 4 Neprilysin also degrades numerous other vasoactive peptides (e.g. substance P).5
Given the major role of the NP pathway in HF, several attempts have been made to potentiate their beneficial effects.6 , 7 Inhibition of NEP activity by sacubitril/valsartan was shown to improve the outcome in patients with symptomatic HF with reduced systolic function.8 However, the predominant source of soluble NEP (sNEP) in patients with HF remains unclear.
The CARMAT total artificial heart (TAH) is a biventricular pulsatile assist device, implanted via sternotomy after excision of both ventricles, leaving the entire native atria in place and is capable of fully restoring cardiac output in patients with advanced HF.9 In this study, by measuring variations in circulating mediators associated with the replacement of the failing ventricles with a TAH and by investigating potential regional differences of sNEP in patients with reduced or preserved cardiac systolic function, we aimed at a better understanding of the pathophysiology of HF, in particular the cardiac contribution to the regulation of the endocrine response.
Methods
Patients with advanced heart failure undergoing total artificial heart implantation
Patients 1 and 2 undergoing TAH implantation have been described elsewhere.9 , 10 Briefly, both were men, aged 68 and 73 years, respectively, with end-stage bi-ventricular HF with left ventricular ejection fraction (LVEF) <20% (Table 1). The implant procedures were uneventful, both patients were rehabilitated and discharged home. Based on body weight, inferior vena cava, left atrial pressure, and inflow pressures, both patients presented features of congestion until hospital discharge (Table 2).
Baseline characteristics of the patients with reduced cardiac systolic function undergoing total artificial heart implant (n = 2)
| Patient 1 | Patient 2 | |
|---|---|---|
| Age (years) | 68 | 73 |
| Left ventricular ejection fraction (%) | 15% | 17% |
| Aetiology of heart failure | Dilated cardiomyopathy | Ischaemic heart disease |
| INTERMACS (class) | 2 | 2 |
| Comorbidities | Hypertension | Atrial fibrillation |
| Dyslipidaemia | ||
| Peripheral artery disease | ||
| Discharge (post- operative day) | 150 | 77 |
| Patient 1 | Patient 2 | |
|---|---|---|
| Age (years) | 68 | 73 |
| Left ventricular ejection fraction (%) | 15% | 17% |
| Aetiology of heart failure | Dilated cardiomyopathy | Ischaemic heart disease |
| INTERMACS (class) | 2 | 2 |
| Comorbidities | Hypertension | Atrial fibrillation |
| Dyslipidaemia | ||
| Peripheral artery disease | ||
| Discharge (post- operative day) | 150 | 77 |
Baseline characteristics of the patients with reduced cardiac systolic function undergoing total artificial heart implant (n = 2)
| Patient 1 | Patient 2 | |
|---|---|---|
| Age (years) | 68 | 73 |
| Left ventricular ejection fraction (%) | 15% | 17% |
| Aetiology of heart failure | Dilated cardiomyopathy | Ischaemic heart disease |
| INTERMACS (class) | 2 | 2 |
| Comorbidities | Hypertension | Atrial fibrillation |
| Dyslipidaemia | ||
| Peripheral artery disease | ||
| Discharge (post- operative day) | 150 | 77 |
| Patient 1 | Patient 2 | |
|---|---|---|
| Age (years) | 68 | 73 |
| Left ventricular ejection fraction (%) | 15% | 17% |
| Aetiology of heart failure | Dilated cardiomyopathy | Ischaemic heart disease |
| INTERMACS (class) | 2 | 2 |
| Comorbidities | Hypertension | Atrial fibrillation |
| Dyslipidaemia | ||
| Peripheral artery disease | ||
| Discharge (post- operative day) | 150 | 77 |
Parameters at baseline and after total artificial heart implant
| Parameter | Baseline | Week 1 | Week 2 | Week 3 | Week 4 | Week 5 | Week 6 |
|---|---|---|---|---|---|---|---|
| Patient 1 | |||||||
| Weight (kg) | 69 | 78 | 75 | 79 | 75 | 76 | 78 |
| Mean arterial pressure (mmHg) | 83 | 78 | 74 | 83 | 94 | 98 | 98 |
| Cardiac output (L/min) | 2.5 | 4.4 | 4.6 | 5.2 | 5.1 | 5.1 | 5.1 |
| Left atrial pressure (mmHg) | — | 8 | — | — | — | — | — |
| Left inflow pressure (mmHg) | — | 19 | 21 | 20 | 18 | 18 | 17 |
| Inferior vena cava diameter (mm) | 31 | 28 | 26 | 29 | |||
| Right inflow pressure (mmHg) | — | 10 | 13 | 12 | 11 | 11 | 13 |
| Creatinine (µmol/L) | 90 | 281 | 97 | 113 | 93 | 84 | 75 |
| Patient 2 | |||||||
| Weight (kg) | 59 | 67 | 79 | 82 | 82 | 80 | 80 |
| Mean arterial pressure (mmHg) | 69 | 85 | 77 | 92 | 77 | 83 | 76 |
| Cardiac output (L/min) | 2.9 | 5.2 | 5.5 | 5.1 | 5.1 | 6 | 6.1 |
| Left atrial pressure (mmHg) | — | 14 | — | — | — | — | — |
| Left inflow pressure (mmHg) | — | 16 | 20 | 25 | 24 | 24 | 22 |
| Inferior vena cava diameter (mm) | 35 | — | 45 | — | — | — | 31 |
| Right inflow pressure (mmHg) | — | 10 | 16 | 19 | 16 | 19 | 16 |
| Creatinine (µmol/L) | 159 | 137 | 74 | 73 | 134 | 129 | 92 |
| Parameter | Baseline | Week 1 | Week 2 | Week 3 | Week 4 | Week 5 | Week 6 |
|---|---|---|---|---|---|---|---|
| Patient 1 | |||||||
| Weight (kg) | 69 | 78 | 75 | 79 | 75 | 76 | 78 |
| Mean arterial pressure (mmHg) | 83 | 78 | 74 | 83 | 94 | 98 | 98 |
| Cardiac output (L/min) | 2.5 | 4.4 | 4.6 | 5.2 | 5.1 | 5.1 | 5.1 |
| Left atrial pressure (mmHg) | — | 8 | — | — | — | — | — |
| Left inflow pressure (mmHg) | — | 19 | 21 | 20 | 18 | 18 | 17 |
| Inferior vena cava diameter (mm) | 31 | 28 | 26 | 29 | |||
| Right inflow pressure (mmHg) | — | 10 | 13 | 12 | 11 | 11 | 13 |
| Creatinine (µmol/L) | 90 | 281 | 97 | 113 | 93 | 84 | 75 |
| Patient 2 | |||||||
| Weight (kg) | 59 | 67 | 79 | 82 | 82 | 80 | 80 |
| Mean arterial pressure (mmHg) | 69 | 85 | 77 | 92 | 77 | 83 | 76 |
| Cardiac output (L/min) | 2.9 | 5.2 | 5.5 | 5.1 | 5.1 | 6 | 6.1 |
| Left atrial pressure (mmHg) | — | 14 | — | — | — | — | — |
| Left inflow pressure (mmHg) | — | 16 | 20 | 25 | 24 | 24 | 22 |
| Inferior vena cava diameter (mm) | 35 | — | 45 | — | — | — | 31 |
| Right inflow pressure (mmHg) | — | 10 | 16 | 19 | 16 | 19 | 16 |
| Creatinine (µmol/L) | 159 | 137 | 74 | 73 | 134 | 129 | 92 |
Parameters at baseline and after total artificial heart implant
| Parameter | Baseline | Week 1 | Week 2 | Week 3 | Week 4 | Week 5 | Week 6 |
|---|---|---|---|---|---|---|---|
| Patient 1 | |||||||
| Weight (kg) | 69 | 78 | 75 | 79 | 75 | 76 | 78 |
| Mean arterial pressure (mmHg) | 83 | 78 | 74 | 83 | 94 | 98 | 98 |
| Cardiac output (L/min) | 2.5 | 4.4 | 4.6 | 5.2 | 5.1 | 5.1 | 5.1 |
| Left atrial pressure (mmHg) | — | 8 | — | — | — | — | — |
| Left inflow pressure (mmHg) | — | 19 | 21 | 20 | 18 | 18 | 17 |
| Inferior vena cava diameter (mm) | 31 | 28 | 26 | 29 | |||
| Right inflow pressure (mmHg) | — | 10 | 13 | 12 | 11 | 11 | 13 |
| Creatinine (µmol/L) | 90 | 281 | 97 | 113 | 93 | 84 | 75 |
| Patient 2 | |||||||
| Weight (kg) | 59 | 67 | 79 | 82 | 82 | 80 | 80 |
| Mean arterial pressure (mmHg) | 69 | 85 | 77 | 92 | 77 | 83 | 76 |
| Cardiac output (L/min) | 2.9 | 5.2 | 5.5 | 5.1 | 5.1 | 6 | 6.1 |
| Left atrial pressure (mmHg) | — | 14 | — | — | — | — | — |
| Left inflow pressure (mmHg) | — | 16 | 20 | 25 | 24 | 24 | 22 |
| Inferior vena cava diameter (mm) | 35 | — | 45 | — | — | — | 31 |
| Right inflow pressure (mmHg) | — | 10 | 16 | 19 | 16 | 19 | 16 |
| Creatinine (µmol/L) | 159 | 137 | 74 | 73 | 134 | 129 | 92 |
| Parameter | Baseline | Week 1 | Week 2 | Week 3 | Week 4 | Week 5 | Week 6 |
|---|---|---|---|---|---|---|---|
| Patient 1 | |||||||
| Weight (kg) | 69 | 78 | 75 | 79 | 75 | 76 | 78 |
| Mean arterial pressure (mmHg) | 83 | 78 | 74 | 83 | 94 | 98 | 98 |
| Cardiac output (L/min) | 2.5 | 4.4 | 4.6 | 5.2 | 5.1 | 5.1 | 5.1 |
| Left atrial pressure (mmHg) | — | 8 | — | — | — | — | — |
| Left inflow pressure (mmHg) | — | 19 | 21 | 20 | 18 | 18 | 17 |
| Inferior vena cava diameter (mm) | 31 | 28 | 26 | 29 | |||
| Right inflow pressure (mmHg) | — | 10 | 13 | 12 | 11 | 11 | 13 |
| Creatinine (µmol/L) | 90 | 281 | 97 | 113 | 93 | 84 | 75 |
| Patient 2 | |||||||
| Weight (kg) | 59 | 67 | 79 | 82 | 82 | 80 | 80 |
| Mean arterial pressure (mmHg) | 69 | 85 | 77 | 92 | 77 | 83 | 76 |
| Cardiac output (L/min) | 2.9 | 5.2 | 5.5 | 5.1 | 5.1 | 6 | 6.1 |
| Left atrial pressure (mmHg) | — | 14 | — | — | — | — | — |
| Left inflow pressure (mmHg) | — | 16 | 20 | 25 | 24 | 24 | 22 |
| Inferior vena cava diameter (mm) | 35 | — | 45 | — | — | — | 31 |
| Right inflow pressure (mmHg) | — | 10 | 16 | 19 | 16 | 19 | 16 |
| Creatinine (µmol/L) | 159 | 137 | 74 | 73 | 134 | 129 | 92 |
Venous blood samples were collected on the day before the operation and weekly thereafter until post-operative week 6 in tubes containing sodium citrate. Blood samples were immediately centrifuged and stored at −80 °C. BNP, mid-regional-pro-ANP (MR-proANP), pro-BNP T71-glycosylation, furin activity, sNEP concentration and activity, and substance P concentration were measured as previously described.2 , 3 The study was approved by the competent authorities and ethics committees, and patients gave signed informed consent.
Regional differences in circulating soluble neprilysin and cardiac RNA analysis
To determine the cardiac contribution of sNEP in patients with advanced HF with reduced cardiac systolic function, we used EDTA plasma samples from three cohorts:
Samples simultaneously drawn from the coronary sinus and the cubital vein of 21 patients with advanced HF and reduced cardiac systolic function (NCT01949246).11 , 12
Samples simultaneously drawn from the pulmonary artery and the left atrium of 16 patients with reduced or preserved cardiac systolic function undergoing cardiac surgery (NCT01723930).
Samples drawn from the cardiac coronary sinus, femoral artery, femoral vein, and pulmonary artery from six patients with preserved cardiac systolic function undergoing elective cardiac catheterization remote from any acute event, as previously described.13
Patient characteristics are summarized in Supplementary material online, Tables S1–S3.
The cardiac tissue samples from advanced HF patients (n = 17) and controls (n = 8) for RNA analysis have been previously described.14 Total RNA was extracted using the Qiagen RNeasy kit as per manufacturer’s instructions.
Variables are expressed as median [interquartile range]. Groups were compared with a Wilcoxon signed-rank or rank-sum test, as appropriate. A two-sided P-value <0.05 was considered significant.
Results
Circulating mediators in two patients with advanced heart failure undergoing total artificial heart implantation
Preoperatively, the NP pathway was markedly activated in both patients (Take home figure, Panels A, B). The following values were obtained at baseline for Patient 1 and 2, respectively: BNP 474 and 695 pg/mL; MR-proANP 778 and 946 pmol/L; furin activity 2.11 and 2.74 pmol/mL/min; percentage of T71-glycosylated proBNP 43.7% and 46.1%; sNEP concentration 358 and 475 pg/mL; sNEP activity 437 and 628 pmol/mL/min.
Circulating mediators in patients with advanced heart failure with reduced cardiac systolic function undergoing TAH implantation. BNP, B-type natriuretic peptide; MR-proANP, mid-regional pro-atrial natriuretic peptide; proBNP T71-glycosylation, pro-B-type natriuretic peptide glycosylation at threonine 71; sNEP, soluble neprilysin.
The replacement of the failing ventricles with TAH was associated with an immediate drop in circulating NPs (BNP ∼90%, MR-proANP ∼67%), and a nearly total disappearance of T71-glycosylated proBNP (∼1%) and circulating furin activity (∼3%). Replacement of the ventricles was also associated with a marked decrease in both sNEP concentration and activity, and a striking reciprocal increase in circulating substance P.
From week 1 onwards, NP plasma levels remained overall unchanged. In contrast, partial recoveries in T71-glycosylated proBNP, furin activity, sNEP activity and concentrations were observed, whereas substance P concentration reciprocally decreased.
Regional differences in circulating soluble neprilysin and cardiac RNA analysis
To confirm the cardiac contribution of circulating sNEP in patients with advanced HF, we compared concentration and activity of sNEP in the coronary sinus and in other vascular beds in patients with reduced or preserved cardiac systolic function. No difference in sNEP concentration and activity between coronary sinus, femoral artery, femoral vein, and pulmonary artery was seen in six patients with preserved cardiac systolic function (all P > 0.05, see Supplementary material online, Figure S1). By contrast, sNEP concentration and activity were 2.9-fold [1.7–5.8] and 3-fold [1.9–4.7] higher in the blood collected from the coronary sinus than in the blood collected from the cubital vein, respectively, in 21 HF patients with reduced systolic function (all P < 0.0001, see Supplementary material online, Figure S2A and B). These findings were in line with a 1.5-fold increase in NEP mRNA levels in left ventricular samples from patients with reduced systolic function compared with healthy individuals (P = 0.03, see Supplementary material online, Figure S2C).
Since this analysis could not exclude a sNEP contribution from the lungs, we measured the trans-pulmonary gradient of sNEP concentration in five patients with reduced cardiac systolic function (LVEF < 45%) and found it neutral (P = 1.0, see Supplementary material online, Figure S2D). Trans-pulmonary gradient of sNEP concentration was slightly positive in 11 patients with preserved cardiac systolic function (P = 0.017, see Supplementary material online, Figure S3).
Discussion
The replacement of a failing heart by a TAH offered a unique opportunity to show the essential role of the heart as a regulator of the cardiovascular endocrine response in HF.
First, our study allowed to distinguish atrial and ventricular contributions to the NP pathway in patients with advanced HF. Our findings confirmed that both ANP and BNP are mainly produced by the ventricles and, to a smaller extent, by the atria.1 It was estimated that ∼90% of circulating BNP and ∼67% of MR-proANP (a surrogate of ANP production) originated from the ventricles, although a reduction in atrial release of NPs after TAH implantation cannot be excluded. Furthermore, this massive drop in plasma NPs was accompanied by an almost complete disappearance of T71-glycosylated proBNP and circulating furin activity. These results confirm that proBNP is glycosylated most preferentially in ventricular cardiomyocytes,15 and strongly suggest that circulating furin activity is mostly derived from ventricular cardiomyocytes. From week 1 onward, proBNP glycosylation was partially restored, suggesting that the remaining atrial cardiomyocytes acquired ventricular features.
Second, the removal of the failing ventricles induced a marked decrease in sNEP concentration and activity, both partially recovering afterwards. Of interest, the postoperative decrease in sNEP was associated with an immediate increase in circulating substance P. While the surgery itself may account in part for the increase in substance P concentration, the strong correlation between substance P and sNEP activity in patients at steady state, and the reciprocal kinetics between substance P and sNEP activity in our two patients, both strongly suggest that the variations in sNEP activity after TAH implantation onwards have affected the plasma concentrations of NEP substrates. Of note, we cannot exclude that the drop in sNEP does not also affect circulating NPs, however, BNP is a poor substrate for sNEP5 and MR-proANP is not expected to be a NEP substrate.
While the first post-operative sampling was performed 5 days after TAH implantation, the decrease in sNEP concentration may result either from the removal of a major source of sNEP (the heart) or from a global haemodynamic improvement. In the present study, we found similar concentrations of sNEP in the coronary sinus and in other vascular beds, indicating no relevant contribution from the heart in circulating sNEP in patients with preserved cardiac systolic function. By contrast, in patients with reduced cardiac systolic function, we found a three-fold higher sNEP concentration and activity in the coronary sinus compared to cubital vein and a neutral trans-pulmonary gradient, strongly suggesting that the heart is a major source of sNEP in patients with reduced cardiac systolic function. These data are supported by higher NEP mRNA in failing left ventricles, compared with healthy controls. Altogether, these data indicate that in patients with reduced cardiac systolic function the heart becomes a source of sNEP, which will in turn modulate the plasma concentrations of sNEP substrates.
Conclusion
In conclusion, this study permitted insight into the pivotal role of the heart as a regulator of the endocrine response in patients with HF with reduced cardiac systolic function. The heart not only contributes directly to the release of NPs but it also constitutes a major source of sNEP in patients with reduced cardiac systolic function.
Supplementary material
Supplementary material is available at European Heart Journal online.
Acknowledgements
We are grateful to Marie-Céline Fournier for technical support and to Wendy Gattis Stough for critical review of the manuscript.
Funding
National Institute of Health (NHLBI K23HL098370 to J.S. and Q.A.T.).
Conflict of interest: A.C.S. reports personal fees from Novartis, Servier, Vifor, outside the submitted work. A.C. reports a patent pending. A.M. reports personal fees from Novartis, Orion, Roche, Servier, Cardiorentis, Zs Pharma, grants and personal fees from Adrenomed, grants from MyCartis, Critical diagnostics, outside the submitted work. C.L. reports personal fees from Carmat, outside the submitted work. D.M.S. reports grants and personal fees from Carmat, during the conduct of the study. JPS reports personal fees from Biotronik, Boston Scientific, Medtronic, Impulse Dynamics, Respicardia, Respicardia, grants and personal fees from Abbott, outside the submitted work. J.P. reports grants and personal fees from Medtronic, grants and personal fees from LFB Biomedicaments, grants and personal fees from Pulsion Medical Systems, personal fees from Maquet, grants from Baxter, outside the submitted work. A.M.A. reports he is a recipient of speaker's honoraria, travel support, and research grants from diagnostic companies with an interest in HF markers including Roche Diagnostics, Alere, and Critical Diagnostics.
References
Author notes
This paper was guest edited by Prof. Anthony DeMaria.
Mattia Arrigo and Nicolas Vodovar contributed equally to this work.
