Prediction of adverse cardiovascular events in children using artificial intelligence-based electrocardiogram

BACKGROUND

Convolutional neural networks (CNNs) have emerged as a novel method for evaluating heart failure (HF) in adult electrocardiograms (ECGs). However, such CNNs are not applicable to pediatric HF, where abnormal anatomy of congenital heart defects plays an important role. ECG-based CNNs reflecting neurohormonal activation (NHA) may be a useful marker of pediatric HF. This study aimed to develop and validate an ECG-derived marker of pediatric HF that reflects the risk of future cardiovascular events.

METHODS

Based on 21,378 ECGs from 8324 children, a CNN was trained using B-type natriuretic peptide (BNP) and the occurrence of major adverse cardiovascular events (MACEs). The output of the model, or the electrical heart failure indicator (EHFI), was compared with the BNP regarding its ability to predict MACEs in 813 ECGs from 295 children.

RESULTS

EHFI achieved a better area under the curve than BNP in predicting MACEs within 180 days (0.826 versus 0.691, p = 0.03). On Cox univariable analyses, both EHFI and BNP were significantly associated with MACE (log10 EHFI: hazard ratio [HR] = 16.5, p < 0.005 and log10 BNP: HR = 4.4, p < 0.005). The time-dependent average precisions of EHFI in predicting MACEs were 32.4%-67.9% and 1.6-7.5-fold higher than those of BNP in the early period. Additionally, the MACE rate increased monotonically with EHFI, whereas the rate peaked at approximately 100 pg/mL of BNP and decreased in the higher range.

CONCLUSIONS

ECG-derived CNN is a novel marker of HF with different prognostic potential from BNP. CNN-based ECG analysis may provide a new guide for assessing pediatric HF.

Sodium loading in ambulatory patients with heart failure with reduced ejection fraction: Mechanistic insights into sodium handling

AIMS

Sodium restriction was not associated with improved outcomes in heart failure patients in recent trials. The skin might act as a sodium buffer, potentially explaining tolerance to fluctuations in sodium intake without volume overload, but this is insufficiently understood. Therefore, we studied the handling of an increased sodium load in patients with heart failure with reduced ejection fraction (HFrEF).

METHODS AND RESULTS

Twenty-one ambulatory, stable HFrEF patients and 10 healthy controls underwent a 2-week run-in phase, followed by a 4-week period of daily 1.2 g (51 mmol) sodium intake increment. Clinical, echocardiographic, 24-h urine collection, and bioelectrical impedance data were collected every 2 weeks. Blood volume, skin sodium content, and skin glycosaminoglycan content were assessed before and after sodium loading. Sodium loading did not significantly affect weight, blood pressure, congestion score, N-terminal pro-brain natriuretic peptide, echocardiographic indices of congestion, or total body water in HFrEF (all p > 0.09). There was no change in total blood volume (4748 ml vs. 4885 ml; p = 0.327). Natriuresis increased from 150 mmol/24 h to 173 mmol/24 h (p = 0.024), while plasma renin decreased from 286 to 88 μU/L (p = 0.002). There were no significant changes in skin sodium content, total glycosaminoglycan content, or sulfated glycosaminoglycan content (all p > 0.265). Healthy controls had no change in volume status, but a higher increase in natriuresis without any change in renin.

CONCLUSIONS

Selected HFrEF patients can tolerate sodium loading, with increased renal sodium excretion and decreased neurohormonal activation.

Updates in Cardiorenal Syndrome

Cardiorenal syndrome is a term that refers to a collection of disorders involving both the heart and kidneys, encompassing multi-directional pathways between the 2 organs mediated through low arterial perfusion, venous congestion, and neurohormonal activation. The pathophysiology is complex and includes hemodynamic and neurohormonal changes, but inconsistent findings from recent studies suggest this is very heterogenous disorder. Management for ADHF remains focused on decongestion and neurohormonal blockade to overcome the intense sodium and fluid avidity of the CRS.

Pathophysiology of Advanced Heart Failure: What Knowledge Is Needed for Clinical Management?

Understanding of heart failure (HF) has evolved from a simple hemodynamic problem through a neurohormonally and proinflammatory-driven syndrome to a complex multiorgan dysfunction accompanied by inadequate energy handling. This article discusses the most important clinical aspects of advanced HF pathophysiology. It presents the concept of neurohormonal activation and its deleterious effect on cardiovascular system and reflex control. The current theories regarding the role of inflammation, cytokine activation, and myocardial remodeling in HF progression are presented. Advanced HF is a multiorgan syndrome with interplay between cardiovascular system and other organs. The role of iron deficiency is also discussed.

Neurohormonal activation induces intracellular iron deficiency and mitochondrial dysfunction in cardiac cells

Background

Iron deficiency (ID) is common in patients with heart failure (HF) and is associated with poor outcomes, yet its role in the pathophysiology of HF is not well-defined. We sought to determine the consequences of HF neurohormonal activation in iron homeostasis and mitochondrial function in cardiac cells.

Methods

HF was induced in C57BL/6 mice by using isoproterenol osmotic pumps and embryonic rat heart-derived H9c2 cells were subsequently challenged with Angiotensin II and/or Norepinephrine. The expression of several genes and proteins related to intracellular iron metabolism were assessed by Real time-PCR and immunoblotting, respectively. The intracellular iron levels were also determined. Mitochondrial function was analyzed by studying the mitochondrial membrane potential, the accumulation of radical oxygen species (ROS) and the adenosine triphosphate (ATP) production.

Results

Hearts from isoproterenol-stimulated mice showed a decreased in both mRNA and protein levels of iron regulatory proteins, transferrin receptor 1, ferroportin 1 and hepcidin compared to control mice. Furthermore, mitoferrin 2 and mitochondrial ferritin were also downregulated in the hearts from HF mice. Similar data regarding these key iron regulatory molecules were found in the H9c2 cells challenged with neurohormonal stimuli. Accordingly, a depletion of intracellular iron levels was found in the stimulated cells compared to non-stimulated cells, as well as in the hearts from the isoproterenol-induced HF mice. Finally, neurohormonal activation impaired mitochondrial function as indicated by the accumulation of ROS, the impaired mitochondrial membrane potential and the decrease in the ATP levels in the cardiac cells.

Conclusions

HF characteristic neurohormonal activation induced changes in the regulation of key molecules involved in iron homeostasis, reduced intracellular iron levels and impaired mitochondrial function. The current results suggest that iron could be involved in the pathophysiology of HF.

A normal sodium diet preserves serum sodium levels during treatment of acute decompensated heart failure: A prospective, blind and randomized trial

BACKGROUND

We tested the hypothesis that a normal sodium diet could be associated with preservation of serum sodium during treatment of acute decompensated heart failure (ADHF).

METHODS AND RESULTS

Forty-four patients hospitalized for ADHF were blindly randomized by using block method to a low sodium diet (LS: 3 g/day of dietary sodium chloride; n = 22, 59.5 ± 11.9 y.o., 50% males. LVEF = 30.0 ± 13.6%); and a normal sodium diet (NS: 7 g/day; n = 22, 56.4 ± 10.3 y.o., 68% males; LVEF = 27.8 ± 11.7%), and both groups were submitted to fluid restriction of 1.000 mL/day. At the 7th day of intervention 16 patients of LS group and 15 patients of NS group were assessed for difference in serum sodium. Both groups had equivalent decongestion, reflected by similar percent reduction of body weight (LS: -5.0 ± 4.7% vs NS: -4.5 ± 5.2%. p = 0.41). Reduction of the N terminal fragment of type B natriuretic peptide (NT-proBNP) was significant only in the NS (-1497.0 [-18843.0 - 1191.0]. p = 0.04). The LS group showed lower levels of serum sodium (135.4 ± 3.5 mmol/L) compared to the NS group (137.5 ± 1.9 mmol/L; p = 0.04). Four cases of hyponatremia were observed only in the LS group (22%). The NS group exhibited higher mean blood pressure values (79.4 ± 2.4 mmHg vs 75.5 ± 3.0 mmHg. p = 0.03), and lower heart rate (73.2 ± 1.6 bpm vs 75.5 ± 2.1 bpm. p = 0.02).

CONCLUSIONS

These results suggest that a normal sodium diet, when compared to a low sodium diet, is associated with similar degrees of decongestion, but with higher levels of natremia, blood pressure and lower neurohormonal activation during ADHF treatment.

TRIAL REGISTRATION

clinicaltrials.gov Identifier no. NCT03722069.

Relationship between non-osmotic arginine vasopressin secretion and hemoglobin A1c levels in adult patients with congenital heart disease

Arginine vasopressin (AVP), which induces vasoconstriction and conserves solute-free water when released during high plasma osmolality, is secreted through 2 mechanisms: osmoregulation and baroregulation. This study aims to clarify the mechanisms and influencing factors for non-osmotic AVP secretion in adult patients with congenital heart disease (CHD). AVP levels were measured in 74 adults with CHD. Non-osmotic AVP secretion was defined as excessive AVP secretion relative to the AVP level inferred from plasma osmolality. Accordingly, 10 patients (13.5%) demonstrated non-osmotic AVP secretion, with AVP levels higher than those in patients without non-osmotic AVP secretion (6.4 ± 3.1 vs. 1.6 ± 0.9 pg/ml; p < 0.0001). Non-osmotic AVP secretion was significantly correlated with diuretic use [odds ratio (OR) 7.227; confidence interval (CI) 1.743-29.962; p = 0.0006], HbA1c level (OR 11.812; CI 1.732-80.548; p = 0.012), and B-type natriuretic peptide (BNP) level (OR 1.007; CI 1.001-1.012; p = 0.022). Multiple logistic regression analysis revealed that there was a significant association between non-osmotic AVP secretion and HbA1c level (OR 9.958; 1.127-87.979; p = 0.0039), and a nearly significant relationship between non-osmotic AVP secretion and BNP (OR 1.006; CI 1.000-1.012; p = 0.056). In conclusion, this study showed that 13.5% of adult patients with CHD demonstrated non-osmotic AVP secretion, which could be correlated with heart failure and insulin resistance. The AVP system might be one of the mechanisms linking heart failure and the onset of type 2 diabetes mellitus in adults with CHD.

Heart-Kidney Interactions in Cardiorenal Syndrome Type 1

The exact significance of kidney function deterioration during acute decompensated heart failure (ADHF) episodes is still under debate. Several studies reported a wide percentage of worsening renal function (WRF) in ADHF patients ranging from 20% to 40%. This is probably because of different populations enrolled with different baseline kidney and cardiac function, varying definition of acute kidney injury (AKI), etiology of kidney dysfunction (KD), and occurrence of transient or permanent KD over the observational period. Current cardiorenal syndrome classification does not distinguish among the mechanisms leading to cardiac and renal deterioration. Cardiorenal syndrome type 1 (CRS-1) is the result of a combination of neurohormonal activation, fluid imbalance, arterial underfilling, increased renal and abdominal pressure, and aggressive decongestive treatment. A more complete mechanistic approach to CRS-1 should include evaluation of baseline kidney function, timing, course and magnitude of KD, and introduction of specific biomarkers able to identify early kidney damage. Therefore, clinical and laboratory parameters may yield a different combination among predisposing, precipitating, and amplifying factors that may influence cardiorenal syndrome development. Thus, CRS-1 is a heterogeneous syndrome that needs to be better defined and categorized taking into account clinical status, renal condition, and treatment. The application of universal definitions for WRF/AKI definition would be the first step to achieve a clear classification.

Neprilysin inhibition in heart failure: mechanisms and substrates beyond modulating natriuretic peptides

The autonomic nervous system, the renin-angiotensin-aldosterone system, and the natriuretic peptide system represent critical regulatory pathways in heart failure and as such have been the major targets of pharmacological development. The introduction and approval of angiotensin receptor neprilysin inhibitors (ARNi) have broadened the available drug treatments of patients with chronic heart failure with reduced ejection fraction. Neprilysin catalyses the degradation of a number of vasodilator peptides, including the natriuretic peptides, bradykinin, substance P, and adrenomedullin, as well as vasoconstrictor peptides, including endothelin-1 and angiotensin I and II. We review the multiple, potentially competing, substrates for neprilysin inhibition, and the resultant composite clinical effects of ARNi therapy. A mechanistic understanding of this novel therapeutic class may provide important insights into the expected on-target and off-target effects when this agent is more widely prescribed.

The first-in-man randomized trial of a beta3 adrenoceptor agonist in chronic heart failure: the BEAT-HF trial

AIMS

The third isotype of beta adrenergic receptors (β3 ARs) has distinctly different effects on cardiomyocytes compared with β1 and β2 ARs. Stimulation of β3 ARs may reduce cardiomyocyte Na⁺ overload and reduce oxidative stress in heart failure (HF). We examined if treatment with the β3 AR agonist mirabegron increases LVEF in patients with HF.

METHODS AND RESULTS

In a double-blind trial we randomly assigned 70 patients with NYHA class II-III HF and LVEF <40% at screening-echocardiography to receive mirabegron or placebo for 6 months as add-on to optimized standard therapy. The primary endpoint was an increase in LVEF after 6 months as measured by computed tomography (CT). Changes in LVEF after 6 months between treatment groups were not significantly different (0.4%, -3.5 to 3.8%, P = 0.82). In an exploratory analysis, based on an expectation that the pathophysiological substrate targeted with treatment is dependent on the baseline LVEF, patients with LVEF <40% by CT given mirabegron had a significant increase in LVEF while no increase was seen in patients given placebo. The changes were significantly different between groups (5.5%, 0.6-10.4%, P < 0.03). Additionally, there was interaction between baseline LVEF and change in LVEF in the entire group of patients treated with mirabegron (R²  = 0.40, β  = -0.63, P < 0.001), but not in the placebo group (R²  = 0.00, β  = -0.01, P = 0.95). Treatment was generally well tolerated. Three patients in each group had fatal or life-threatening events.

CONCLUSIONS

The primary endpoint was not reached. Exploratory analysis indicated that β3 AR stimulation by mirabegron increased LVEF in patients with severe HF. Treatment appeared safe. Additional studies in severe HF are needed.

TRIAL REGISTRATION

NCT01876433.

Type 4 cardiorenal syndrome

The Acute Dialysis Quality Initiative consensus conference proposed a classification of cardiorenal syndrome (CRS), aiming for a better delineation of each subtype. Although the exact pathophysiology of type 4 CRS is not completely understood, the mechanisms involved are probably multifactorial. There is growing evidence that oxidative stress is a major connector in the development and progression of type 4 CRS. Giving its complexity, poor prognosis and increasing incidence, type 4 CRS is becoming a significant public health problem. Patients with chronic kidney disease are particularly predisposed to cardiac dysfunction, due to the high prevalence of traditional cardiovascular risk factors in this population, but the contribution of risk factors specific to chronic kidney disease should also be taken into account. Much remains to be elucidated about type 4 CRS: despite progress over the last decade, there are still significant questions regarding its pathophysiology and there is as yet no specific therapy. A better understanding of the mechanisms involved may provide potential targets for intervention. The present review will provide a brief description of the definition, epidemiology, diagnosis, prognosis, biomarkers and management strategies of type 4 CRS, and the pathophysiological mechanisms and risk factors presumably involved in its development will be particularly highlighted.

Copeptin in Heart Failure

Heart failure (HF) is one of the most common causes of hospitalization and mortality in the modern Western world and an increasing proportion of the population will be affected by HF in the future. Although HF management has improved quality of life and prognosis, mortality remains very high despite therapeutic options. Medical management consists of a neurohormonal blockade of an overly activated neurohormonal axis. No single marker has been able to predict or monitor HF with respect to disease progression, hospitalization, or mortality. New methods for diagnosis, monitoring therapy, and prognosis are warranted. Copeptin, a precursor of pre-provasopressin, is a new biomarker in HF with promising potential. Copeptin has been found to be elevated in both acute and chronic HF and is associated with prognosis. Copeptin, in combination with other biomarkers, could be a useful marker in the monitoring of disease severity and as a predictor of prognosis and survival in HF.