Metabolite Profiles of Heart Failure, Central Hemodynamic Derangement, and Response to Heart Transplantation

Carl Lejonberg; Tomasz Czuba; Anna Egerstedt; Nilanjana Ghosh; Selvi Celik; Charlotta Ljungman; Jakob Lundgren; Göran Rådegran; Olof Gidlöf; J Gustav Smith

doi:10.1161/JAHA.124.039248

. 2025 Jul 14;14(14):e039248. doi: 10.1161/JAHA.124.039248

Metabolite Profiles of Heart Failure, Central Hemodynamic Derangement, and Response to Heart Transplantation

Carl Lejonberg ^1,^2,^✉, Tomasz Czuba ^1,^2,³, Anna Egerstedt ^1,^2,⁴, Nilanjana Ghosh ³, Selvi Celik ^1,², Charlotta Ljungman ⁵, Jakob Lundgren ^1,⁴, Göran Rådegran ^1,⁵, Olof Gidlöf ^1,², J Gustav Smith ^1,^2,^3,^4,⁵

PMCID: PMC12533607 PMID: 40654253

Abstract

Background

Heart failure (HF) is characterized by hemodynamic derangements that are likely to mediate systemic metabolic perturbations but limited data are available. Plasma metabolite profiling provides opportunities for comprehensive investigation of such perturbations. Here, we aimed to characterize plasma profiles that associate with HF and their relationship with central hemodynamics, symptom burden, and response to restoration of cardiac function by heart transplantation.

Methods

Untargeted metabolite profiling was conducted with mass spectrometry in 2 independent case–control samples.

Results

In total, 89 of 797 studied metabolites were significantly associated with HF in both cohorts with concordant directionality. Amino acid, carbohydrate, and nucleotide metabolites were enriched for association with HF and were consistently increased in HF cases. A subset of patients with advanced HF subsequently underwent heart transplantation, after which 17 of the 89 metabolites returned significantly toward healthy control levels. These 17 metabolites represent increased catecholamine and heme metabolism, conjugated bile acids, kynurenine pathway mediators, spermidine metabolism, and allantoin, as well as tricarboxylic acid cycle and glycolysis intermediates. Most of these metabolites associated with symptom burden and at least 1 of 12 central hemodynamic parameters, primarily relating to either increased systemic or pulmonary venous congestion, lower cardiac output, or lower left ventricular stroke work.

Conclusions

We comprehensively identified metabolite profiles associated with HF and central hemodynamics that reverse by cardiac transplantation. Increased levels of most metabolites also associated with higher symptom burden. Our findings provide perspectives on the metabolic consequences of HF with potential implications for noninvasive monitoring and tailored therapy.

Keywords: heart failure, hemodynamics, metabolomics, transplantation

Subject Categories: Heart Failure, Metabolism

Nonstandard Abbreviations and Acronyms

HF‐OPMP: hospital‐based heart failure outpatient management program
HF‐RHC: patients with heart failure undergoing right‐heart catheterization
PC: principal component
UPLC‐MS/MS: ultrahigh performance liquid chromatography‐tandem mass spectroscopy

Clinical Perspective.

What Is New?

We identified plasma metabolites associated with heart failure that reverse by transplantation.
Identified metabolites further associate with symptom burden and central hemodynamic parameters.

What Are the Clinical Implications?

The identified metabolites inform on metabolic consequences of heart failure and may help guide disease monitoring and therapy.

Heart failure (HF) is the end stage of heart disease and a leading cause of mortality and morbidity. ¹ HF is characterized by reduced cardiac output and increased filling pressures, resulting in limitations of perfusion as well as venous congestion of blood and a hypokinetic circulation ² that may mediate metabolic effects on many tissues including lung, kidney, liver, and gut. Metabolites that are heavily influenced by central hemodynamic consequences of HF, such as increased levels of catecholamines and aldosterone, are associated with increased mortality in patients with HF and constitute central therapeutic targets for HF. ³ These metabolites were first implicated in HF pathophysiology more than 60 years ago in targeted experiments and are the targets of beta blockers and mineralcorticoid receptor antagonists (MRAs). ⁴ , ⁵ Sodium‐glucose cotransporter 2 inhibitors are also central in HF treatment and are thought to act both by modulating central hemodynamic parameters as well as having direct effects on cardiac energetics. ⁶ , ⁷ Other foundational therapies include inhibitors of angiotensin converting enzymes, angiotensin II receptors, and neprilysin, which target circulating proteins that also influence central hemodynamics and are likely to confer beneficial downstream metabolic effects. ⁸ , ⁹ A range of additional therapies that influence central hemodynamics are available in HF today, including myosin activators, guanylate cyclase stimulators, I_f channel inhibitors, hydralazine/isosorbide dinitrate, digoxin, and diuretics, which results in a relatively complicated therapeutic landscape. ² There is thus an increasing need for objective markers informative on pathobiological processes to guide therapeutic selections. ¹⁰ We propose that systematic profiling of metabolic perturbations in HF may both identify potential novel therapeutic targets and plasma profiles that may serve as biomarkers to guide tailoring of existing therapies ¹¹ , ¹² to the individual (precision medicine).

Systematic profiling of protein‐depleted plasma using mass spectrometry results in identification of a large number of small molecules involved in metabolic reactions in the human body, often referred to as the metabolome. ¹³ Such plasma profiles constitute a representation of the overall metabolic state of human tissues. ¹⁴ A handful of plasma metabolomics studies of HF have been reported with inconsistent findings, largely reflecting very small sample sizes and differences in assayed metabolites, disease stage, and comorbid conditions. ¹⁵ , ¹⁶ No study has comprehensively explored metabolic correlates of central hemodynamic derangement, and clinical symptoms and signs are not reliable indicators of such hemodynamic measures. ¹⁷ , ¹⁸

Therefore, we aimed to apply an untargeted metabolite profiling platform to 2 independent HF case–control samples with detailed information on disease stage based on central hemodynamic parameters from cardiac catheterization. For metabolite profiling, we employed a core laboratory with evidence of high reproducibility and the highest metabolite coverage of any commercially available platform. ¹⁹ , ²⁰ Furthermore, we obtained serial samples before and after restoration of cardiac function and central hemodynamics with heart transplantation in a subset of patients with HF to facilitate study of the consequences of HF without confounding by comorbid conditions as these are unlikely to improve by transplantation.

METHODS

Detailed aggregated study data that support the findings of this study are available in Data S1. Remaining data are available from the corresponding author upon reasonable request.

Study Samples

Samples from male and female participants of 2 case–control studies underwent metabolite profiling: a discovery study with 247 patients with HF undergoing right heart catheterization (HF‐RHC), compared with 30 controls, and a validation study including 83 patients with HF attending a hospital‐based HF outpatient management program (HF‐OPMP), with controls representing a random subset of individuals without HF from the population‐based MDCS (Malmö Diet and Cancer Study). Clinical data were collected retrospectively from patient records in HF cases, and prospectively in controls. All studies were approved by the local ethics committee at Lund University, Sweden, and all participants provided written informed consent.

Patients With HF Undergoing Right Heart Catheterization

A total of 247 consecutive patients undergoing RHC for hemodynamic evaluation of HF as part of routine clinical care at Skåne University Hospital in Lund, Sweden, between August 2011 and April 2017 provided plasma samples as part of a local biobanking project at Skåne University Hospital (Lund Cardiopulmonary Register). ²¹

RHC was performed as described previously ²² and in Data S1. Blood was drawn from the central venous catheter, immediately centrifuged at 2000g for 10 minutes, whereafter plasma was extracted. Samples were then stored at −80 °C without thawing until metabolite analysis was performed.

A subset of 28 patients who had HF‐RHC subsequently underwent heart transplantation after which a second plasma sample was obtained at a routine 6‐month follow‐up with RHC for endomyocardial biopsy and were processed as with other RHC samples. Among these, no signs of transplant rejection were detected at biopsy. Baseline samples were obtained at a median of 232 days (range 7–1854) before heart transplantation and consecutive sampling at 6‐month follow‐up with a median of 167 days (range 148–191 days). Data on exercise capacity from clinical cycle ergometry exams from within 1 month of the RHC date were collected from electronic health records, as were functional class at the time of RHC according to the New York Heart Association classification and drug therapies used.

For controls, 30 subjects matched by age and sex to participants with HF‐RHC but without HF were recruited by advertising in the local newspaper, of whom 9 individuals had well‐controlled type 2 diabetes without complications (for the purpose of a parallel study) and others were free from any known condition and without medications. Peripheral blood samples were obtained for metabolite analysis and were immediately frozen to −80 °C until analysis. In the HF‐RHC and controls, creatinine measures were obtained from the metabolite profiling platform in all subjects. This measure was strongly correlated with creatinine as obtained from routine clinical chemistry (Spearman's rho 0.88, P<0.001) at the time of blood sampling from patient records for all HF cases and as measured at baseline in the MDCS.

HF Patients From Hospital‐Based Outpatient Management Program

Peripheral venous blood samples were obtained from 83 patients with HF attending an outpatient‐based, nurse‐led HF management program at Skåne University Hospital in Lund ²³ between 2013 and 2014, through median cubital vein venipuncture. Samples were collected in EDTA‐coated tubes and processed as HF‐RHC samples.

For controls, a random subset of 583 individuals without a diagnosis of HF at baseline were selected from the MDCS as described previously. ²¹ The MDCS included 30 447 men (born between 1923 and 1945) and women (born between 1923 and 1950) from the city of Malmö in southern Sweden. Participants underwent baseline examinations between 1991 and 1996, including anthropometric measures, blood pressure measurement using a mercury‐column sphygmomanometer after 10 minutes of rest in the supine position, filled out a questionnaire, and underwent blood sampling from the median cubital vein in EDTA‐coated tubes under standardized fasting conditions. ²⁴ Samples were immediately centrifuged and plasma extracted, aliquoted and stored at −80 °C.

Metabolite Profiling

Metabolite profiling was performed at a core laboratory (Metabolon, Morrisville, NC, USA) in 2 batches, 1 for HF‐RHC and 1 for HF‐OPMP. Standardized protocols for Ultrahigh Performance Liquid Chromatography‐Tandem Mass Spectroscopy (UPLC‐MS/MS) were applied as described in detail previously. ²⁰ , ²⁵ , ²⁶

Raw data were extracted, peak identified, and quality control processed using Metabolon hardware and software. ²⁷ We have previously described that a similar set of metabolites is identified by this untargeted platform when applied across multiple population‐based cohorts of variable sample size, demographic characteristics, and storage times. ²⁰

Samples were prepared using an automated liquid handling platform (MicroLab STAR, Hamilton Company, Reno, NV, USA). For quality control purposes, several recovery standards were added before the first step in the extraction process.

Briefly, methanol precipitation was performed for 2 minutes under vigorous shaking (Glen Mills Genogrinder 2000) to remove proteins and dissociate small molecules bound to proteins or trapped in the precipitated protein matrix, followed by centrifugation. The resulting extract was divided into fractions for analysis by 4 protocols: 2 separate reverse phase UPLC‐MS/MS methods with positive ion electrospray ionization mode, a reverse phase UPLC‐MS/MS method with negative ion electrospray ionization mode, and a hydrophilic interaction liquid chromatography UPLC‐MS/MS method with negative ion electrospray ionization mode. Experimental samples and controls were randomized across the platform run. The UPLC‐MS/MS platform used an Acquity UPLC (Waters Corp., Milford, MA, USA) and a Q‐Exactive Orbitrap mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) interfaced with a heated electrospray ionization source and Orbitrap mass analyzer operating at 35 000 mass resolution.

Instrument variability was determined by calculating the median relative SD for the standards that were added to each sample before injection into the mass spectrometers (median relative SD=6%). Overall variability was determined by calculating the median relative SD for all endogenous metabolites (ie, noninstrument standards) present in 100% of the pooled human plasma samples (median relative SD=8%).

The MS analysis alternated between MS and data‐dependent MS scans using dynamic exclusion. The scan range varied slightly between methods and covered a range from 70 to 1000 m/z.

Compounds were identified by comparing them to library entries of purified standards or recurrent unknown entities. Metabolon maintains a library based on authenticated standards that contains the retention time/index, mass‐to‐charge ratio, and chromatographic data (including MS/MS spectral data) on all molecules present in the library. Furthermore, biochemical identifications were carried out based on 3 criteria: retention time/index within a narrow window of the retention time/index of the proposed identification, accurate mass match to the library (±10 ppm), and the MS/MS forward and reverse scores between the experimental data and authentic standards. The MS/MS scores were obtained based on a comparison of the ions present in the experimental spectrum to the ions present in the library spectrum. Peaks were quantified using the area under the curve method. The raw area counts for each metabolite in each sample were normalized to correct for variation resulting from instrument day‐to‐day tuning differences by the median value for each run day; the median was set to 1.0 for each run. This preserved variations among samples allowed metabolites of widely different raw peak areas to be compared on a similar graphical scale.

Statistical Analysis

Metabolite values were normalized to sample volumes and standardized to facilitate comparability of risk estimates. We have previously reported that modest missingness and positive skewness of most nonxenobiotic plasma metabolites is a consistent finding in both large and small population‐based cohort studies. ²⁰ In the 2 case–control collections studied here, we also observed modest missingness (Figure S1) and a marked positive skew (Figure S2) for a large proportion of metabolites. As in previous studies we imputed missing metabolites for our main analysis with the lowest value detected for each metabolite, under the assumption that missingness indicates a value below the lower limit of detection for nonxenobiotic metabolites. After excluding metabolites with >20% missing data as previously recommended, ²⁸ , ²⁹ we conducted a secondary analysis to explore the impact of a nonparametric random forest approach using the R package MissForest.

Finally, we performed natural log transformation of values, which largely removed skewed distributions (Figure S2). We confirmed approximate log‐normal distributions for the 89 significant metabolites by visual inspection of histograms. We also confirmed normal distributions for potential confounders adjusted for in regression models.

Underlying patterns of metabolite variation was first explored by principal components analysis of a covariance matrix for the overall plasma metabolome, including both HF‐RHC cases and controls. The plasma metabolome variance explained by individual principal components was derived from eigenvalues and visualized by a scree plot. The contribution of HF case–control status was tested by logistic regression models with each principal component as independent predictor. The contribution of each metabolite to each principal component was explored by correlation coefficients and loading.

For primary discovery analysis, we next tested the association of each metabolite with HF in logistic regression models using four separate adjustment models. For primary analysis, we adjusted for age, sex, and diabetes (model 1). For metabolites reproducibly associated with HF in both samples we explored additional adjustment for potential confounders including creatinine as a measure of renal function (model 2), and further in the validation cohort with systolic and diastolic blood pressure, antihypertensive treatment, body mass index, smoking, and history of myocardial infarction (model 3), and a full model that included both covariates from model 3 and creatinine (model 4).

In the discovery stage, metabolites meeting a P value threshold of 6×10⁻⁵ were considered significant based on Bonferroni adjustment for 797 individual tests. In the validation stage, metabolites meeting a threshold of 3×10⁻⁴ were considered significant, based on Bonferroni adjustment for 165 individual tests. Enrichment of specific metabolite classes and pathways was explored based on the metabolite groupings supplied by Metabolon using 2‐sided Fisher's exact tests.

Metabolites associated with HF in both stages with concordant directionality were tested for reversal toward control levels in serial samples before and after heart transplantation, using repeated‐measures linear regression models incorporating mixed effects, with adjustment for age and sex. P values <0.05 were considered statistically significant.

Finally, the 17 metabolites associated with HF in both stages that reversed toward controls after heart transplantation were tested for correlation using Spearman's rank correlation coefficients, explored for central hemodynamic correlates in the HF‐RHC and with measures of symptom burden. Association with each of 12 hemodynamic measures, which were also standardized, was tested for each metabolite using general linear regression models with adjustment for age and sex. For these analyses, P values <2×10⁻⁴ were considered significant, based on Bonferroni‐adjustment for 17×12 different tests. Association with symptom burden was explored in cases from the HF‐RHC using logistic regression models for NYHA functional class I and II versus III and IV and linear regression models for exercise capacity during cycle ergometry (proportion of observed exercise capacity in relation to that expected based on population‐specific nomograms used in clinical practice in Sweden). The association with use of specific HF therapies was also explored in patients with HF with reduced ejection fraction from the HF‐RHC cohort with age and sex‐adjusted linear regression models for each class of foundational therapy. Finally, we explored sex interactions for identified metabolites using interaction terms.

Analyses were performed in STATA v15.1 (StataCorp, College Station, TX USA), R v4.1.2 (R Foundation for Statistical Computing, Vienna, Austria) and RStudio v2021.09.2 build 382 (RStudio, Boston, MA, USA).

RESULTS

Descriptive characteristics at baseline of both case–control samples are shown in Table 1. As expected, NT‐proBNP (N‐terminal pro‐B‐type natriuretic peptide) was modestly elevated in the outpatient cohort (median 1113 pg/mL [482–2937], HF‐OPMP), more markedly elevated in the patients undergoing RHC (2362 pg/mL [1099–4837], HF‐RHC) but normal in controls (60 pg/mL [32–101]). Most patients with HF were in NYHA class II or III, had a reduced ejection fraction (86% in HF‐OPMP and 64% in HF‐RHC) and were on evidence‐based therapies. Heart transplant recipients were younger at baseline and had lower burden of comorbidities and higher NT‐proBNP (3609 pg/mL [2368–4792]) than other patient groups with HF.

Table 1.

Descriptive Characteristics of Study Samples

	HF‐RHC cohort		HF‐OPMP		Transplant recipients (n=28)
	HF (n=247)	Controls (n=30)	HF (n=83)	Controls (n=583)	Transplant recipients (n=28)
Baseline age, y	57.7 (14.0)	61.9 (5.6)	66.0 (11.9)	57.0 (5.8)	49.0 (11.6)
Male sex, n (%)	166 (67)	19 (63.3)	64 (77.1)	237 (40.7)	20 (71.4)
Body mass index (kg/m²)	27.0 (4.8)	…	27.1 (4.7)	25.4 (3.7)	26.4 (3.2)
Systolic BP (mm Hg)	117.2 (24.2)	…	124.0 (20.9)	138.5 (18.1)	105.3 (15.1)
Diastolic BP (mm Hg)	74.4 (10.7)	…	74.8 (12.0)	85.5 (9.0)	72.8 (9.1)
Current smoking, n (%)	28 (11.3)	0 (0)	7 (8.4)	159 (28.2)	1 (3.6)
Hypertension at inclusion, n (%)	83 (33.6)	0 (0)	…	…	5 (17.9)
Atrial fibrillation at inclusion, n (%)	114 (46.0)	0 (0)	33 (39.8)	3 (0.5)	13 (46.4)
Diabetes at inclusion, n (%)	49 (19.8)	9 (30)	11 (13.3)	45 (7.7)	2 (7.1)
N‐terminal pro‐B‐type natriuretic peptide (pg/mL)	2362 (1099–4837)	…	1113 (482–2937)	60 (32.1–100.5)	3609 (2368–4792)
Creatinine (μmol/L)	107.0 (48.4)	…	107.5 (34.5)	84.4 (13.1)	103.9 (29.2)
Estimated glomerular filtration rate (mL/min per 1.73 m²)	69.2 (25.7)	…	55.6 (21.4)	79.6 (18.5)	70.8 (23.4)
Antihypertensive treatment, n (%)	240 (96.8)	…	83 (100)	69 (11.8)	…
Angiotensin‐converting enzyme inhibitor/angiotensin‐II receptor blocker, n (%)	196 (79.0)		81 (97.6)	11 (1.9)
Beta blocker, n (%)	210 (85.0)		75 (90.4)	50 (8.6)
Aldosterone antagonist, n (%)	149 (60.6)		47 (56.6)	…
Loop diuretic, n (%)	205 (82.7)		64 (77.1)	…
History of myocardial infarction at inclusion, n (%)	42 (16.9)	…	15 (18.1)	9 (1.5)	1 (3.6)
HF subtype, n (%)
HF with preserved EF	65 (26.2)		8 (9.6)		2 (7.1)
HF with mildly reduced EF	24 (9.7)	…	4 (4.8)	…	2 (7.1)
HF with reduced EF	159 (64.1)		71 (85.6)		24 (85.7)
New York Heart Association class, n (%)		…		…
I	17 (6.9)		1 (1.2)		1 (3.6)
II	65 (26.5)		22 (26.5)		5 (17.9)
III	143 (58.4)		36 (43.4)		21 (75.0)
IV	20 (8.2)		24 (29.0)		1 (3.6)

Open in a new tab

Categorical variables are presented as count and proportion, continuous variables as mean±SD or median and interquartile range as appropriate. No patients used sodium‐glucose cotransporter‐2 inhibitors or angiotensin receptor‐neprilysin inhibitors. BP indicates blood pressure; EF, ejection fraction; HF, heart failure; HF‐RHC, patients with heart failure undergoing right‐heart catheterization; and HF‐OPMP, hospital‐based heart failure outpatient management program.

Metabolite Profiles in HF

In total, 1370 metabolites were identified in the discovery cohort HF‐RHC of which 330 (24%) were of unknown identity, 240 were xenobiotic substances (18%) and 3 were without variance, leaving 797 metabolites for analysis. Of those, 725 (91%) were also present among the 1372 metabolites identified in the validation cohort HF‐OPMP.

In principal component analysis exploring underlying patterns in the plasma metabolome in HF‐RHC, the first principal component (PC1) was found to explain 15% of variability and other components explained ≤6% (Figure S3). PC1 was strongly associated with HF case–control status (P<0.001) whereas other PCs were not (P>0.50 for all). A total of 184 metabolites correlated with PC1 with r >0.5 (Table S1), all of which had loadings of 4% to 8% indicating small contributions to HF and PC1 of many metabolites.

In regression analysis of individual metabolites, a total of 165 metabolites (Figure 1) were associated with HF in HF‐RHC after adjustment for age, sex, and diabetes, of which 89 were reproducibly associated with HF in HF‐OPMP with concordant effect direction. Of these metabolites, none were missing in >20% in the HF‐RHC and 5 were missing in >20% from the HF‐OPMP (S‐adenosylhomocysteine, galactonate, heme, 3‐hydroxyadipate, taurochenodeoxycholic acid 3‐sulfate). These metabolites should therefore be viewed more critically. ³⁰ However, none were missing in >50% (S1). Analysis with nonparametric random forest imputation resulted in concordant results (Table S2). Nearly half of the 89 metabolites were amino acids (43 metabolites, 48%), which were all increased in patients with HF and nearly one fifth were lipids (16 metabolites, 18%) of which several were lower in HF (Figure 2). Amino acids were markedly enriched (1.8‐fold, P=0.005) among HF‐associated metabolites, as were carbohydrates (3‐fold, P=0.009) and nucleotides (2.9‐fold, P=0.006) whereas lipids were underrepresented (0.33, P=1×10⁻⁵). Metabolites were further grouped across 103 subpathways, of which 3 were enriched among HF‐associated metabolites: polyamine metabolism (P=0.007), pentose metabolism (P=0.01), and histidine metabolism (P=0.03) whereas amino sugar metabolism was nominally significant (P=0.05).

Manhattan plot depicting metabolite associations between all 797 metabolites (y axis) and heart failure in the HF‐RHC cohort, categorized by metabolite class (x axis). The plot is based on regression analysis with adjustment for age, sex, and diabetes. The dashed red line represents the negative logarithm of the Bonferroni correction level (0.05/797 metabolites), with metabolites above this level being associated with heart failure. HF‐RHC indicates patients with heart failure undergoing right‐heart catheterization.

Enrichment of metabolite classes amongst 89 heart failure associated metabolites from 725 metabolites present in both cohorts. A, Metabolite classes. B, Metabolite pathways. Asterisk indicates statistically significant enrichment according to Fisher's exact P<0.05. ***P<0.001, **P<0.01, *P<0.05. SAM indicates S‐adenosylmethionine; and TCA, tricarboxylic acid.

To further account for potential confounders, additional models were examined adjusted for creatinine (model 2). One third (26 metabolites) of the 89 reproducible metabolites remained associated with HF at strict Bonferroni‐adjusted thresholds in both stages. The independence of metabolites from conventional risk factors was further tested in the replication cohort, using models 3 and 4 as described in the methods section. In model 3 and 4, two thirds of metabolites (65 and 65 metabolites) remained associated with HF in HF‐OPMP. All 4 models are presented in Table S2.

Metabolite Responses to Heart Transplantation

We next examined whether the 89 metabolites associated with HF were reversible upon heart transplantation. We found that 17 metabolites changed significantly toward levels observed in controls after transplantation, representing multiple biochemical pathways (Table 2). Metabolites within the same biochemical pathways were correlated as shown in Figure 3, including sulfated bile acids, kynurenine metabolism (kynurenine, quinolinate), tricarboxylic acid (TCA) cycle intermediates (malate, alpha‐ketoglutarate), and spermidine metabolism (acisoga, acetylspermidine). Only 3 of the 89 metabolites (biliverdin, N‐carbamoylalanine, gamma‐glutamylalanine) were significantly different in plasma metabolite concentration between patients with HF from HF‐RHC who underwent transplantation compared with other patients with HF. One metabolite, allantoin, displayed significant interaction with sex, with higher levels (0.9 SDs) in male patients with HF (P<0.05 in both cohorts).

Table 2.

Metabolites Associated With Heart Failure That Reverse by Heart Transplantation

Metabolite	HF‐RHC		HF‐OPMP		Post‐Tx
Metabolite	OR	P value	OR	P value	Beta	P value
Amino acids
Methionine, cysteine, and taurine metabolism
N‐acetylmethionine	2.87	7.8×10⁻⁷	2.05	1.7×10⁻⁴	−0.25	0.02
Glutathione metabolism
Cysteinylglycine disulfide	2.83	9.9×10⁻⁶	4.10	6.5×10⁻¹¹	−0.47	0.04
Histidine metabolism
Formiminoglutamate	3.38	3.3×10⁻⁶	5.23	1.1×10⁻¹⁴	−0.69	1.6×10⁻⁵
Tyrosine metabolism
Vanillylmandelate	7.77	1.1×10⁻⁸	3.24	2.2×10⁻¹⁰	−0.42	0.01
Tryptophan metabolism
Kynurenine	3.65	4.5×10⁻⁶	3.25	9.6×10⁻¹¹	−1.46	6.2×10⁻²¹
Polyamine metabolism
Acisoga	4.29	1.9×10⁻⁸	2.12	7.5×10⁻⁶	−0.58	3.1×10⁻⁴
N(1)+N(8)‐acetylspermidine	6.00	7.0×10⁻⁸	3.21	5.1×10⁻¹¹	−0.81	6.3×10⁻⁶
Carbohydrates
Glycolysis, gluconeogenesis, and pyruvate metabolism
3‐phosphoglycerate	4.79	4.0×10⁻⁹	2.68	8.1×10⁻¹⁰	−0.72	0.002
Cofactors and vitamins
Nicotinate and nicotinamide metabolism
Quinolinate	16.82	9.3×10⁻⁹	14.53 3.7×10⁻¹⁸	−0.64	3.9×10⁻⁴
Hemoglobin and porphyrin metabolism
Heme	4.28	2.8×10⁻⁸	6.59	1.8×10⁻¹⁶	−0.56	0.02
Biliverdin	4.31	2.1×10⁻⁸	2.06	2.4×10⁻⁶	−1.32	2.0×10⁻¹⁸
Energy
Tricarboxylic acid cycle
Alpha‐ketoglutarate	4.86	1.7×10⁻⁷	2.14	1.9×10⁻⁸	−0.42	0.04
Malate	5.97	2.6×10⁻⁶	2.05	1.3×10⁻⁶	−0.52	6.3×10⁻⁴
Lipids
Primary bile acid metabolism
Glycochenodeoxycholate 3‐sulfate	3.71	2.9×10⁻⁶	2.37	5.3×10⁻⁶	−0.63	1.1×10⁻⁴
Secondary bile acid metabolism
Taurochenodeoxycholic acid 3‐sulfate	3.81	1.3×10⁻⁶	1.69	8.4×10⁻⁵	−0.41	0.04
Sphingomyelins
Sphingomyelin^*	0.30	9.2×10⁻⁶	0.52	2.9×10⁻⁶	0.62	2.4×10⁻⁵
Nucleotide
Purine metabolism
Allantoin	4.77	1.1×10⁻⁹	6.14	3.9×10⁻¹⁵	−0.52	0.049

Open in a new tab

Metabolites that are significantly associated with heart failure in both cohorts, and 6 months after heart transplantation in the HF‐RHC subcohort. Beta indicates beta coefficient; HF‐RHC, patients with heart failure undergoing right‐heart catheterization; HF‐OPMP, hospital‐based heart failure outpatient management program, and OR, odds ratio.

Sphingomyelin (d18:1/25:0, d19:0/24:1, d20:1/23:0, d19:1/24:0).

Pairwise Spearman correlation coefficients for 17 plasma metabolites reproducibly associated with heart failure that reverse after transplantation. Asterisk indicates a compound that has not been confirmed based on a metabolite standard, but Metabolon is confident in its identity.

Relation of Plasma Metabolites With Hemodynamic Parameters

Information on 12 hemodynamic parameters was obtained from patients with HF‐RHC, as described in Table S3. Association was tested with each measure for each of the 17 metabolites associated with HF and reversal after transplantation. Of these, all except heme and 3‐phosphoglycerate were associated with at least 1 of the 12 parameters. The largest number of associations were observed with mixed venous oxygen saturation (14 metabolites), left‐ventricular stroke work‐index (12 metabolites) and right atrial pressure (11 metabolites). Associations with hemodynamic measures are illustrated in Figure 4 and detailed in Table S4.

Association with 12 hemodynamic measures for 17 plasma metabolites reproducibly associated with heart failure that reverse after heart transplantation. CI indicates cardiac index; LVSWI, left ventricular stroke work index; MAP, mean arterial pressure; MPAP, mean pulmonary arterial pressure; PAWP, pulmonary artery wedge pressure; PVRI, pulmonary vascular resistance index; RAP, right atrial pressure; RVSWI, right ventricular stroke work index; SaO₂, arterial oxygen saturation; SVI, stroke volume index; SVO₂, mixed venous oxygen saturation; and SVRI, systemic vascular resistance index.

Relation of Plasma Metabolites With Measures of Symptom Burden

We next explored the association of the 17 metabolites with measures of symptom burden, including functional class and exercise capacity (Table 3). Most metabolites (11 of 17, 65%) were associated with lower exercise capacity of which all except 1 also associated with worse NYHA functional class. One metabolite (a sphingomyelin) associated with higher exercise capacity and better NYHA class. These associations were also independent of NT‐proBNP for metabolites strongly related to lower cardiac output/stroke work (catecholamine metabolite vanillylmandelate and TCA intermediates alpha‐ketoglutarate, malate) and higher pulmonary venous congestion (kynurenine and spermidine metabolites: acisoga, kynurenine, N(1)+N(8)‐acetylspermidine) (Table S5).

Table 3.

Association of Metabolites With Measures of Symptom Burden

Metabolite	Functional class		Exercise capacity
Metabolite	OR	P value	Beta	P value
Amino acids
Methionine, cysteine, and taurine metabolism
N‐acetylmethionine	2.89	0.68	−8.89	1.0 × 10⁻³
Glutathione metabolism
Cysteinylglycine disulfide	3.21	0.28	−6.20	0.01
Histidine metabolism
Formiminoglutamate	5.17	1.4 × 10⁻³	−7.12	4.1 × 10⁻³
Tyrosine metabolism
Vanillylmandelate	6.67	1.9 × 10⁻⁴	−12.94	2.2 × 10⁻⁸
Tryptophan metabolism
Kynurenine	5.04	2.5 × 10⁻³	−10.14	3.6 × 10⁻⁵
Polyamine metabolism
Acisoga	5.40	1.3 × 10⁻³	−9.61	1.3 × 10⁻⁵
N(1)+N(8)‐acetylspermidine	6.56	1.0 × 10⁻⁴	−8.91	4.5 × 10⁻⁵
Carbohydrates
Glycolysis, gluconeogenesis, and pyruvate metabolism
3‐phosphoglycerate	3.06	0.44	−1.90	0.42
Cofactors and vitamins
Nicotinate and nicotinamide metabolism
Quinolinate	3.83	0.07	−4.84	0.06
Hemoglobin and porphyrin metabolism
Heme	3.31	0.23	0.77	0.74
Biliverdin	4.04	0.02	−4.06	0.10
Energy
Tricarboxylic acid cycle
Alpha‐ketoglutarate	7.92	1.8 × 10⁻⁵	−9.52	3.3 × 10⁻⁵
Malate	6.78	1.1 × 10⁻⁴	−8.78	6.0 × 10⁻⁵
Lipids
Primary bile acid metabolism
Glycochenodeoxycholate 3‐sulfate	6.91	3.5 × 10⁻⁵	−7.78	3.5 × 10⁻⁴
Secondary bile acid metabolism
Taurochenodeoxycholic acid 3‐sulfate	7.21	4.7 × 10⁻⁵	−7.51	9.9 × 10⁻⁴
Sphingomyelins
Sphingomyelin^*	1.97	0.01	9.82	1.5 × 10⁻⁵
Nucleotide
Purine metabolism
Allantoin	3.69	0.07	−3.15	0.16

Open in a new tab

Association of metabolites reproducibly associated with HF that reverse after transplantation with measures of symptom burden, including functional class according to the New York Heart Association (I–II vs III–IV) and exercise capacity as proportion of observed in Watts during cycle ergometry to that expected based on an age and sex specific reference cohort used in clinical practice in Sweden, in patients with HF from the cohort of patients with HF undergoing right‐heart catheterization. Beta indicates beta coefficient; HF, heart failure; and OR, odds ratio.

Sphingomyelin (d18:1/25:0, d19:0/24:1, d20:1/23:0, d19:1/24:0).

Relation of Plasma Metabolites With Drug Therapy

Finally, we explored whether the 17 metabolites were related to the use of specific foundational HF therapies in patients with HF with reduced ejection fraction from the HF‐RHC cohort. No patient from this cohort used angiotensin receptor‐neprilysin inhibitors or SGLT2 (sodium‐glucose cotransporter 2) inhibitors, which were not reimbursed for HF use in Sweden at that time. Most patients with HF with reduced ejection fraction (62%) used both an ACE inhibitor (ACEi)/angiotensin receptor blocker, a beta blocker, and an MRA. Two drugs were used in 26% of whom most (18%) used an ACEi/angiotensin receptor blocker and a beta blocker, whereas 1% to 3% used a single drug. Only 4% used none of these drugs.

Interestingly, use of ACEi/angiotensin receptor blocker correlated with lower levels of metabolites associated with lower cardiac output/stroke work (alpha‐ketoglutarate, malate) and high pulmonary venous congestion (acisoga, kynurenine, N(1)+N(8)‐acetylspermidine). In contrast, beta blockers and MRAs were associated with higher levels of acisoga and kynurenine. MRAs were also associated with higher levels of vanillylmandelate, N(1)+N(8)‐acetylspermidine, and malate (Table S6).

DISCUSSION

We here conducted the largest study to date of untargeted metabolite profiling in HF, leveraging a comprehensive and well‐validated platform, identifying 89 metabolites that were reproducibly associated with HF. Our study is the first metabolite profiling study to also provide a repeated‐sampling perspective after heart transplantation, in which 17 metabolites reversed toward levels in healthy controls, indicating that these represent consequences of HF rather than of comorbid conditions. Furthermore, our study is also the first to relate metabolite findings to central hemodynamic measures, finding that most of the identified metabolites relate to cardiac output, stroke work, venous congestion, or hypokinetic circulation, and to describe differences in hemodynamic profile between metabolites.

We propose that the identified metabolite profile may serve as a noninvasive biomarker to guide individual tailoring of therapy in HF and to monitor treatment response. We argue that metabolites that primarily correlate with cardiac stroke work (such as vanillylmandelate and TCA intermediates) may inform benefit and dose selection for therapies with large impact on stroke work such as beta blockers ³¹ and angiotensin receptor‐neprilysin inhibitors ³² and metabolites that correlate with venous congestion (such as kynurenine or spermidine metabolites) may point to benefit of therapies with large effect on congestion such as SGLT2 inhibitors, ³³ MRA, ³⁴ or other diuretics. As current therapy is mainly guided by symptoms, it seems likely that therapy targeting sensitive measures of intravascular congestion and cardiac workload even in the absence of symptoms could confer benefits as these processes can drive disease progression and confer adverse outcomes. ³⁵ However, such hypotheses need to be examined in future prospective studies, as outlined in the American Heart Association roadmap for metabolomics‐driven precision medicine. ³⁶ , ³⁷ In addition to their potential use as biomarkers, several interesting pathophysiological observations can be made regarding the metabolite profile associated with HF and its hemodynamic consequences.

First, a salient feature of our results was the increased levels of many amino acids in patients with HF. This observation is consistent with a study of cardiac blood samples, which found increased release of most essential amino acids from the failing heart that was interpreted as representing increased myocardial proteolysis. ³⁸ Among the amino acid‐derived metabolites that also declined after heart transplantation in our study, 3 pathways particularly stand out, all of which have important signaling functions of potential importance in HF: catecholamine metabolism (vanillylmandelate), the kynurenine pathway (kynurenine, quinolinate), and spermidine metabolism (acetylspermidine, acisoga). The finding of increased vanillylmandelate in HF is not unexpected as vanillylmandelate represents an important downstream metabolite of catecholamines, used as a diagnostic marker for catecholamine‐producing endocrine tumors, ³⁹ and given the central role of catecholamines in the pathophysiology of HF and as target for one of the foundational therapies—beta blockers. ² The finding that higher vanillylmandelate mainly associates with low cardiac index and left‐ventricular stroke work‐index is consistent with the current view of catecholamines in HF as resulting from impaired cardiac output. ³ Metabolites of the kynurenine pathway have also previously been linked to HF and are related to both cardiac energetics and inflammation. ⁴⁰ , ⁴¹ Kynurenine is the main metabolite of the amino acid tryptophan and mediates inflammation ⁴⁰ and quinolinate, a byproduct of kynurenine metabolism, is the biosynthetic precursor to nicotinamide adenine dinucleotide, a central coenzyme in mitochondrial metabolism. It has been described in mice ⁴² that circulating kynurenine can be absorbed, converted to quinolinate, and further metabolized to nicotinamide adenine dinucleotide⁺. ⁴³ Increasing nicotinamide adenine dinucleotide⁺ levels by supplementation with another precursor, nicotinamide riboside, has been reported to improve cardiac function in mice. ⁴⁴ , ⁴⁵ Interestingly, kynurenine and quinolinate were most strongly related to measures of congestion and may represent mediators of inflammation in response to congestion, which has been associated with adverse outcomes. ⁴⁶ The spermidine pathway metabolites are derived from the amino acid arginine and have been associated with reduced ejection fraction, ⁴⁷ ischemic HF, ⁴⁸ and mortality in cardiovascular disease. ⁴⁹ Spermidine has also previously been associated with worse symptoms and adverse outcomes in HF ⁵⁰ as well as with longevity potentially through association with cardiomyocyte authophagy. ⁵¹ Among central hemodynamic measures, these metabolites particularly associated with measures of a hypokinetic circulation and congestion. These three pathways were all associated with higher measures of symptom burden and levels were lower in patients using ACEi/angiotensin receptor blocker, consistent with the reduced afterload and resulting lower stroke work conferred by these drugs. The higher levels observed in MRA users may reflect higher disease severity in this subset.

Second, in addition to amino acids, increased levels of several lipids were also noted in patients with HF although lipids as a group were underrepresented among HF‐associated metabolites. Free fatty acids are the primary energy substrate in both failing and healthy hearts, ³⁸ and although we identified several fatty acids to be reproducibly higher in patients with HF, these did not have evidence of reversibility upon transplantation. It therefore seems likely that the lipid associations represent confounding by the increased cardiometabolic burden of cases with HF as compared with healthy controls, whereas increased cardiac fatty acid metabolism may not be detectable in peripheral blood. Instead, bile acids, being recognized as signaling molecules, were identified as reversible in response to transplantation and have also previously been associated with HF. ⁵² In the present study, one primary (glycochenodeoxycholate 3‐sulfate) and one secondary (taurochenodeoxycholic acid 3‐sulfate) sulfated bile acid was higher in patients with HF than in controls. These metabolites were correlated with right atrial pressure and could represent HF‐induced cholestasis. During cholestasis, bile acid homeostasis is upheld by different hepato‐protective mechanisms. ⁵³ Sulfonation is one such mechanism, facilitating increased solubility and excretion. Despite emerging evidence of the major bile acids ability to regulate cardiovascular function, ⁵⁴ little has previously been reported about the sulfonated conjugated forms found in plasma and their potential role as signaling molecules.

Third, instead of lipids, nucleotide and carbohydrate metabolites were enriched among HF‐associated metabolites. However, only 1 nucleotide (allantoin) and 1 carbohydrate (3‐phosphoglycerate) metabolite reversed significantly after transplantation. Allantoin, a metabolite of uric acid, is widely regarded as a marker of increased oxidative stress and has previously been described to be elevated in individuals with HF. ⁵⁵ Interestingly, allantoin displayed a sex interaction with higher levels in male patients with HF. 3‐phosphoglycerate is an intermediate in both glycolysis and the Calvin–Benson cycle. A shift from fatty acid beta‐oxidation toward glycolysis is a well‐established hallmark of the failing heart that has been detected in previous studies both based on heart tissue and in blood samples obtained from the coronary sinus together with arterial blood. However, a clear representation of this shift has not been seen in untargeted metabolite profiling studies of peripheral blood, ¹² , ¹⁵ , ¹⁶ likely reflecting that metabolite patterns in peripheral venous blood represent the end result of metabolic processes in many organs. The need for strict Bonferroni thresholds to account for multiple testing in untargeted studies may further contribute to obscure detection of significant effects in these systems. Indeed, a potential circulating biomarker of metabolic shifts in the heart is long‐chain acylcarnitines, key transporters of fatty acids for subsequent mitochondrial beta‐oxidation that have previously been associated with HF ⁵⁶ , ⁵⁷ and were significantly associated with HF in our minimally adjusted models, but this effect was lost upon additional adjustments including for renal function. We note that this finding is consistent with a previous study ³⁸ in which no increased secretion of acylcarnitines was observed by the failing heart.

Finally, among the metabolites that reversed after transplantation were also 2 metabolites implicating energy turnover in the TCA cycle, which particularly correlated with cardiac index and stroke work. Release of TCA intermediates from the heart, especially the failing heart, has been described previously and interpreted to represent clearance of excess TCA cycle 4‐carbon units produced by amino acid breakdown (excess anaplerosis). ³⁸

Our study design has several strengths. We employed a broad and well‐validated metabolomics platform, a replication cohort for external validation of our primary results, a well‐characterized patient cohort with invasive hemodynamic data, and serial blood sampling that allowed investigation of metabolic reversal after heart transplantation. To our knowledge, our study also represents the largest study to date of untargeted metabolite profiles in HF. However, several limitations need to be addressed. The sample size was limited, particularly for the transplantation analysis. Although the transplantation cohort also benefited from serial sampling, in which cases represent their own controls, it may be due to limitations in sample size that metabolites with signaling functions, in which plasma differences are likely to be more marked, stood out most importantly in the transplantation analysis. We also note that the control groups for the 2 stages were relatively different, which may affect results; for the discovery stage we employed a healthy control sample (some of whom had type 2 diabetes) whereas for the replication stage controls were from a population‐based cohort study where comorbidities were more common. These differences could introduce confounding at baseline, affect the comparability of the cohorts, and potentially reduce the generalizability of our findings. However, we were able to adjust for comorbidities such as diabetes and renal function in both cohorts, and in the transplantation analysis cases serve as their own controls. We were also unable to adjust for any differences in fasting times and dietary patterns. Cases with HF in our discovery cohort were all fasting, as is customary before RHC. Blood samples drawn from controls in our replication cohort were similarly obtained after fasting but we did not have data on specific fasting times and dietary conditions before blood sampling. Nevertheless, we believe that the metabolites that were significantly associated with HF in both cohorts are likely to represent robust associations. The limited sample size in the transplant cohort prevented adjustment for several potential confounders such as immunosuppressive therapy and posttransplant comorbidities. Larger studies are warranted to confirm our findings, ideally with the use of alternate methods for reconstitution of cardiac function, such as left ventricular assist devices, which are not subject to the potential confounders present in transplant recipients including the requirement for immunosuppressants, which may confer metabolic side effects.

Conclusions

In summary, we have applied untargeted metabolomics to describe plasma profiles reproducibly associated with HF, their hemodynamic correlates, symptom burden, and reversal after transplantation. Our findings implicate several pathways of potential importance in HF and support metabolite profiling as a potential readout of cardiac function, central hemodynamic parameters, and peripheral metabolic consequences that may act to guide therapeutic selection and monitor treatment response. A prospective trial to further develop and evaluate the utility of a metabolomics‐driven strategy for precision medicine in HF may be warranted.

Sources of Funding

J. Gustav Smith was supported by grants from the Swedish Heart‐Lung Foundation (2022‐0344), the Swedish Research Council (2021‐02273), the European Research Council (ERC‐STG‐2015‐679242), Gothenburg University, Skåne University Hospital, governmental funding of clinical research within the Swedish National Health Service, a generous donation from the Knut and Alice Wallenberg foundation to the Wallenberg Center for Molecular Medicine in Lund, and funding from the Swedish Research Council (Linnaeus grant Dnr 349‐2006‐237, Strategic Research Area Exodiab Dnr 2009‐1039) and Swedish Foundation for Strategic Research (Dnr IRC15‐0067) to the Lund University Diabetes Center. Carl Lejonberg was supported by grants from the Stig and Ragna Gorthon Foundation (2018‐1480) and the Thorsten‐Birger Segerfalk Foundation. The funding agencies did not have any role in the design and conduct of the study, collection, management, and interpretation of the data; or preparation, review, or approval of the article.

Disclosures

None.

Supporting information

Data S1

Tables S1–S6

Figures S1–S3

JAH3-14-e039248-s001.pdf^{(1.3MB, pdf)}

Acknowledgments

We acknowledge the contributions of the Region Skåne Biobank (Lund, Sweden) for storage and retrieval of all samples.

This article was sent to Sula Mazimba, MD, MPH, Associate Editor, for review by expert referees, editorial decision, and final disposition.

Supplemental Material is available at https://www.ahajournals.org/doi/suppl/10.1161/JAHA.124.039248

For Sources of Funding and Disclosures, see page 14.

REFERENCES

1. Tsao CW, Aday AW, Almarzooq ZI, Alonso A, Beaton AZ, Bittencourt MS, Boehme AK, Buxton AE, Carson AP, Commodore‐Mensah Y, et al. Heart disease and stroke statistics‐2022 update: a report from the American Heart Association. Circulation. 2022;145:e153–e639. doi: 10.1161/CIR.0000000000001052 [DOI] [PubMed] [Google Scholar]
2. McDonagh TA, Metra M, Adamo M, Gardner RS, Baumbach A, Bohm M, Burri H, Butler J, Celutkiene J, Chioncel O, et al. 2021 esc guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur Heart J. 2021;42:3599–3726. doi: 10.1093/eurheartj/ehab368 [DOI] [PubMed] [Google Scholar]
3. Mann DL, Felker GM. Mechanisms and models in heart failure: a translational approach. Circ Res. 2021;128:1435–1450. doi: 10.1161/CIRCRESAHA.121.318158 [DOI] [PMC free article] [PubMed] [Google Scholar]
4. Chidsey CA, Harrison DC, Braunwald E. Augmentation of the plasma nor‐epinephrine response to exercise in patients with congestive heart failure. N Engl J Med. 1962;267:650–654. doi: 10.1056/NEJM196209272671305 [DOI] [PubMed] [Google Scholar]
5. Deming QB, Luetscher JA Jr. Increased sodium‐retaining corticoid excretion in edema, with some observations on the effects of cortisone in nephrosis. J Clin Invest. 1950;29:808. doi: 10.1172/JCI102317 [DOI] [PubMed] [Google Scholar]
6. Joshi SS, Singh T, Newby DE, Singh J. Sodium‐glucose co‐transporter 2 inhibitor therapy: mechanisms of action in heart failure. Heart. 2021;107:1032–1038. doi: 10.1136/heartjnl-2020-318060 [DOI] [PMC free article] [PubMed] [Google Scholar]
7. Selvaraj S, Fu Z, Jones P, Kwee LC, Windsor SL, Ilkayeva O, Newgard CB, Margulies KB, Husain M, Inzucchi SE, et al. Metabolomic profiling of the effects of dapagliflozin in heart failure with reduced ejection fraction: DEFINE‐HF. Circulation. 2022;146:808–818. doi: 10.1161/CIRCULATIONAHA.122.060402 [DOI] [PMC free article] [PubMed] [Google Scholar]
8. Brown NJ, Vaughan DE. Angiotensin‐converting enzyme inhibitors. Circulation. 1998;97:1411–1420. doi: 10.1161/01.CIR.97.14.1411 [DOI] [PubMed] [Google Scholar]
9. Bozkurt B, Nair AP, Misra A, Scott CZ, Mahar JH, Fedson S. Neprilysin inhibitors in heart failure: the science, mechanism of action, clinical studies, and unanswered questions. JACC Basic Transl Sci. 2023;8:88–105. doi: 10.1016/j.jacbts.2022.05.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
10. Bhatt AS, Abraham WT, Lindenfeld J, Bristow M, Carson PE, Felker GM, Fonarow GC, Greene SJ, Psotka MA, Solomon SD, et al. Treatment of HF in an era of multiple therapies: statement from the HF Collaboratory. JACC Heart Fail. 2021;9:1–12. doi: 10.1016/j.jchf.2020.10.014 [DOI] [PubMed] [Google Scholar]
11. Muller J, Bertsch T, Volke J, Schmid A, Klingbeil R, Metodiev Y, Karaca B, Kim SH, Lindner S, Schupp T, et al. Narrative review of metabolomics in cardiovascular disease. J Thorac Dis. 2021;13:2532–2550. doi: 10.21037/jtd-21-22 [DOI] [PMC free article] [PubMed] [Google Scholar]
12. Ikegami R, Shimizu I, Yoshida Y, Minamino T. Metabolomic analysis in heart failure. Circ J. 2017;82:10–16. doi: 10.1253/circj.CJ-17-1184 [DOI] [PubMed] [Google Scholar]
13. Guo L, Milburn MV, Ryals JA, Lonergan SC, Mitchell MW, Wulff JE, Alexander DC, Evans AM, Bridgewater B, Miller L, et al. Plasma metabolomic profiles enhance precision medicine for volunteers of normal health. Proc Natl Acad Sci USA. 2015;112:E4901–E4910. doi: 10.1073/pnas.1508425112 [DOI] [PMC free article] [PubMed] [Google Scholar]
14. German JB, Hammock BD, Watkins SM. Metabolomics: building on a century of biochemistry to guide human health. Metabolomics. 2005;1:3–9. doi: 10.1007/s11306-005-1102-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
15. Wang TJ, Gupta DK. Metabolite profiles in heart failure: looking for unique signatures in a heterogeneous syndrome. J Am Coll Cardiol. 2015;65:1521–1524. doi: 10.1016/j.jacc.2015.02.019 [DOI] [PubMed] [Google Scholar]
16. Zhou J, Chen X, Chen W, Zhong L, Cui M. Comprehensive plasma metabolomic and lipidomic analyses reveal potential biomarkers for heart failure. Mol Cell Biochem. 2021;476:3449–3460. doi: 10.1007/s11010-021-04159-5 [DOI] [PubMed] [Google Scholar]
17. Stevenson LW, Perloff JK. The limited reliability of physical signs for estimating hemodynamics in chronic heart failure. JAMA. 1989;261:884–888. doi: 10.1001/jama.1989.03420060100040 [DOI] [PubMed] [Google Scholar]
18. Shah MR, Hasselblad V, Stinnett SS, Gheorghiade M, Swedberg K, Califf RM, O'Connor CM. Hemodynamic profiles of advanced heart failure: association with clinical characteristics and long‐term outcomes. J Card Fail. 2001;7:105–113. doi: 10.1054/jcaf.2001.24131 [DOI] [PubMed] [Google Scholar]
19. Yu B, Zanetti KA, Temprosa M, Albanes D, Appel N, Barrera CB, Ben‐Shlomo Y, Boerwinkle E, Casas JP, Clish C, et al. The Consortium of Metabolomics Studies (Comets): metabolomics in 47 prospective cohort studies. Am J Epidemiol. 2019;188:991–1012. doi: 10.1093/aje/kwz028 [DOI] [PMC free article] [PubMed] [Google Scholar]
20. Ghosh N, Lejonberg C, Czuba T, Dekkers K, Robinson R, Arnlov J, Melander O, Smith ML, Evans AM, Gidlof O, et al. Analysis of plasma metabolomes from 11 309 subjects in five population‐based cohorts. Sci Rep. 2024;14:8933. doi: 10.1038/s41598-024-59388-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
21. Egerstedt A, Berntsson J, Smith ML, Gidlof O, Nilsson R, Benson M, Wells QS, Celik S, Lejonberg C, Farrell L, et al. Profiling of the plasma proteome across different stages of human heart failure. Nat Commun. 2019;10:5830. doi: 10.1038/s41467-019-13306-y [DOI] [PMC free article] [PubMed] [Google Scholar]
22. Egerstedt A, Czuba T, Bronton K, Lejonberg C, Ruge T, Wessman T, Radegran G, Schulte J, Hartmann O, Melander O, et al. Bioactive adrenomedullin for assessment of venous congestion in heart failure. ESC Heart Fail. 2022;9:3543–3555. doi: 10.1002/ehf2.14018 [DOI] [PMC free article] [PubMed] [Google Scholar]
23. Martinsson A, Oest P, Wiborg MB, Reitan O, Smith JG. Longitudinal evaluation of ventricular ejection fraction and NT‐proBNP across heart failure subgroups. Scand Cardiovasc J. 2018;52:205–210. doi: 10.1080/14017431.2018.1461920 [DOI] [PubMed] [Google Scholar]
24. Smith JG, Newton‐Cheh C, Almgren P, Struck J, Morgenthaler NG, Bergmann A, Platonov PG, Hedblad B, Engstrom G, Wang TJ, et al. Assessment of conventional cardiovascular risk factors and multiple biomarkers for the prediction of incident heart failure and atrial fibrillation. J Am Coll Cardiol. 2010;56:1712–1719. doi: 10.1016/j.jacc.2010.05.049 [DOI] [PMC free article] [PubMed] [Google Scholar]
25. Evans AM, DeHaven CD, Barrett T, Mitchell M, Milgram E. Integrated, nontargeted ultrahigh performance liquid chromatography/electrospray ionization tandem mass spectrometry platform for the identification and relative quantification of the small‐molecule complement of biological systems. Anal Chem. 2009;81:6656–6667. doi: 10.1021/ac901536h [DOI] [PubMed] [Google Scholar]
26. Evans AM, Bridgewater BR, Liu Q, Mitchell MW, Robinson RJ, Dai H, Stewart SJ, DeHaven CD, Miller LA. High resolution mass spectrometry improves data quantity and quality as compared to unit mass resolution mass spectrometry in high‐throughput profiling metabolomics. J Postgenom Drug Biomarker Dev. 2014;4:1. doi: 10.4172/2153-0769.1000132 [DOI] [Google Scholar]
27. Dehaven CD, Evans AM, Dai H, Lawton KA. Organization of GC/MS and LC/MS metabolomics data into chemical libraries. J Cheminform. 2010;2:9. doi: 10.1186/1758-2946-2-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
28. Stekhoven DJ, Bühlmann P. Missforest—non‐parametric missing value imputation for mixed‐type data. Bioinformatics. 2012;28:112–118. doi: 10.1093/bioinformatics/btr597 [DOI] [PubMed] [Google Scholar]
29. Gromski PS, Xu Y, Kotze HL, Correa E, Ellis DI, Armitage EG, Turner ML, Goodacre R. Influence of missing values substitutes on multivariate analysis of metabolomics data. Metabolites. 2014;4:433–452. doi: 10.3390/metabo4020433 [DOI] [PMC free article] [PubMed] [Google Scholar]
30. Bijlsma S, Bobeldijk I, Verheij ER, Ramaker R, Kochhar S, Macdonald IA, van Ommen B, Smilde AK. Large‐scale human metabolomics studies: a strategy for data (pre‐) processing and validation. Anal Chem. 2006;78:567–574. doi: 10.1021/ac051495j [DOI] [PubMed] [Google Scholar]
31. Andersson B, Blomstrom‐Lundqvist C, Hedner T, Waagstein F. Exercise hemodynamics and myocardial metabolism during long‐term beta‐adrenergic blockade in severe heart failure. J Am Coll Cardiol. 1991;18:1059–1066. doi: 10.1016/0735-1097(91)90767-4 [DOI] [PubMed] [Google Scholar]
32. Verbrugge FH, Borlaug BA. Getting the “right” perspective on angiotensin receptor‐neprilysin inhibition in heart failure. J Am Heart Assoc. 2020;9:e017292. doi: 10.1161/JAHA.120.017292 [DOI] [PMC free article] [PubMed] [Google Scholar]
33. Omar M, Jensen J, Frederiksen PH, Kistorp C, Videbaek L, Poulsen MK, Moller S, Ali M, Gustafsson F, Kober L, et al. Effect of empagliflozin on hemodynamics in patients with heart failure and reduced ejection fraction. J Am Coll Cardiol. 2020;76:2740–2751. doi: 10.1016/j.jacc.2020.10.005 [DOI] [PubMed] [Google Scholar]
34. Schohn D, Jahn HA, Pelletier BC. Dose‐related cardiovascular effects of spironolactone. Am J Cardiol. 1993;71:40A–45A. doi: 10.1016/0002-9149(93)90244-7 [DOI] [PubMed] [Google Scholar]
35. Boorsma EM, Ter Maaten JM, Damman K, Dinh W, Gustafsson F, Goldsmith S, Burkhoff D, Zannad F, Udelson JE, Voors AA. Congestion in heart failure: a contemporary look at physiology, diagnosis and treatment. Nat Rev Cardiol. 2020;17:641–655. doi: 10.1038/s41569-020-0379-7 [DOI] [PubMed] [Google Scholar]
36. Cresci S, Pereira NL, Ahmad F, Byku M, de Las Fuentes L, Lanfear DE, Reilly CM, Owens AT, Wolf MJ. Heart failure in the era of precision medicine: a scientific statement from the American Heart Association. Circ Genom Precis Med. 2019;12:458–485. doi: 10.1161/HCG.0000000000000058 [DOI] [PubMed] [Google Scholar]
37. Cheng S, Shah SH, Corwin EJ, Fiehn O, Fitzgerald RL, Gerszten RE, Illig T, Rhee EP, Srinivas PR, Wang TJ, et al. Potential impact and study considerations of metabolomics in cardiovascular health and disease: a scientific statement from the American Heart Association. Circ Cardiovasc Genet. 2017;10:10. doi: 10.1161/HCG.0000000000000032 [DOI] [PMC free article] [PubMed] [Google Scholar]
38. Murashige D, Jang C, Neinast M, Edwards JJ, Cowan A, Hyman MC, Rabinowitz JD, Frankel DS, Arany Z. Comprehensive quantification of fuel use by the failing and nonfailing human heart. Science. 2020;370:364–368. doi: 10.1126/science.abc8861 [DOI] [PMC free article] [PubMed] [Google Scholar]
39. Boyle JG, Davidson DF, Perry CG, Connell JM. Comparison of diagnostic accuracy of urinary free metanephrines, vanillyl mandelic acid, and catecholamines and plasma catecholamines for diagnosis of pheochromocytoma. J Clin Endocrinol Metab. 2007;92:4602–4608. doi: 10.1210/jc.2005-2668 [DOI] [PubMed] [Google Scholar]
40. Lund A, Nordrehaug JE, Slettom G, Solvang SH, Pedersen EK, Midttun O, Ulvik A, Ueland PM, Nygard O, Giil LM. Plasma kynurenines and prognosis in patients with heart failure. PLoS One. 2020;15:e0227365. doi: 10.1371/journal.pone.0227365 [DOI] [PMC free article] [PubMed] [Google Scholar]
41. Konishi M, Ebner N, Springer J, Schefold JC, Doehner W, Dschietzig TB, Anker SD, von Haehling S. Impact of plasma kynurenine level on functional capacity and outcome in heart failure‐results from Studies Investigating Co‐Morbidities Aggravating Heart Failure (SICA‐HF). Circ J. 2016;81:52–61. doi: 10.1253/circj.CJ-16-0791 [DOI] [PubMed] [Google Scholar]
42. Terakata M, Fukuwatari T, Kadota E, Sano M, Kanai M, Nakamura T, Funakoshi H, Shibata K. The niacin required for optimum growth can be synthesized from l‐tryptophan in growing mice lacking tryptophan‐2,3‐dioxygenase. J Nutr. 2013;143:1046–1051. doi: 10.3945/jn.113.176875 [DOI] [PubMed] [Google Scholar]
43. Moffett JR, Arun P, Puthillathu N, Vengilote R, Ives JA, Badawy AA, Namboodiri AM. Quinolinate as a marker for kynurenine metabolite formation and the unresolved question of NAD⁺ synthesis during inflammation and infection. Front Immunol. 2020;11:31. doi: 10.3389/fimmu.2020.00031 [DOI] [PMC free article] [PubMed] [Google Scholar]
44. Walker MA, Tian R. Raising NAD in heart failure: time to translate? Circulation. 2018;137:2274–2277. doi: 10.1161/CIRCULATIONAHA.117.032626 [DOI] [PMC free article] [PubMed] [Google Scholar]
45. Diguet N, Trammell SAJ, Tannous C, Deloux R, Piquereau J, Mougenot N, Gouge A, Gressette M, Manoury B, Blanc J, et al. Nicotinamide riboside preserves cardiac function in a mouse model of dilated cardiomyopathy. Circulation. 2018;137:2256–2273. doi: 10.1161/CIRCULATIONAHA.116.026099 [DOI] [PMC free article] [PubMed] [Google Scholar]
46. Núñez J, de la Espriella R, Rossignol P, Voors AA, Mullens W, Metra M, Chioncel O, Januzzi JL, Mueller C, Richards AM, et al. Congestion in heart failure: a circulating biomarker‐based perspective. A review from the biomarkers working group of the Heart Failure Association, European Society of Cardiology. Eur J Heart Fail. 2022;24:1751–1766. doi: 10.1002/ejhf.2664 [DOI] [PubMed] [Google Scholar]
47. Puetz A, Artati A, Adamski J, Schuett K, Romeo F, Stoehr R, Marx N, Federici M, Lehrke M, Kappel BA. Non‐targeted metabolomics identify polyamine metabolite acisoga as novel biomarker for reduced left ventricular function. ESC Heart Fail. 2022;9:564–573. doi: 10.1002/ehf2.13713 [DOI] [PMC free article] [PubMed] [Google Scholar]
48. Nayak A, Liu C, Mehta A, Ko YA, Tahhan AS, Dhindsa DS, Uppal K, Jones DP, Butler J, Morris AA, et al. N8‐acetylspermidine: a polyamine biomarker in ischemic cardiomyopathy with reduced ejection fraction. J Am Heart Assoc. 2020;9:e016055. doi: 10.1161/JAHA.120.016055 [DOI] [PMC free article] [PubMed] [Google Scholar]
49. Ottosson F, Smith E, Fernandez C, Melander O. Plasma metabolites associate with all‐cause mortality in individuals with type 2 diabetes. Metabolites. 2020;10:315. doi: 10.3390/metabo10080315 [DOI] [PMC free article] [PubMed] [Google Scholar]
50. Cheng ML, Wang CH, Shiao MS, Liu MH, Huang YY, Huang CY, Mao CT, Lin JF, Ho HY, Yang NI. Metabolic disturbances identified in plasma are associated with outcomes in patients with heart failure: diagnostic and prognostic value of metabolomics. J Am Coll Cardiol. 2015;65:1509–1520. doi: 10.1016/j.jacc.2015.02.018 [DOI] [PubMed] [Google Scholar]
51. Eisenberg T, Abdellatif M, Schroeder S, Primessnig U, Stekovic S, Pendl T, Harger A, Schipke J, Zimmermann A, Schmidt A, et al. Cardioprotection and lifespan extension by the natural polyamine spermidine. Nat Med. 2016;22:1428–1438. doi: 10.1038/nm.4222 [DOI] [PMC free article] [PubMed] [Google Scholar]
52. Mayerhofer CCK, Ueland T, Broch K, Vincent RP, Cross GF, Dahl CP, Aukrust P, Gullestad L, Hov JR, Troseid M. Increased secondary/primary bile acid ratio in chronic heart failure. J Card Fail. 2017;23:666–671. doi: 10.1016/j.cardfail.2017.06.007 [DOI] [PubMed] [Google Scholar]
53. Alnouti Y. Bile acid sulfation: a pathway of bile acid elimination and detoxification. Toxicol Sci. 2009;108:225–246. doi: 10.1093/toxsci/kfn268 [DOI] [PubMed] [Google Scholar]
54. Khurana S, Raufman JP, Pallone TL. Bile acids regulate cardiovascular function. Clin Transl Sci. 2011;4:210–218. doi: 10.1111/j.1752-8062.2011.00272.x [DOI] [PMC free article] [PubMed] [Google Scholar]
55. Causse E, Fournier P, Roncalli J, Salvayre R, Galinier M. Serum allantoin and aminothiols as biomarkers of chronic heart failure. Acta Cardiol. 2017;72:397–403. doi: 10.1080/00015385.2017.1335104 [DOI] [PubMed] [Google Scholar]
56. Ahmad T, Kelly JP, McGarrah RW, Hellkamp AS, Fiuzat M, Testani JM, Wang TS, Verma A, Samsky MD, Donahue MP, et al. Prognostic implications of long‐chain acylcarnitines in heart failure and reversibility with mechanical circulatory support. J Am Coll Cardiol. 2016;67:291–299. doi: 10.1016/j.jacc.2015.10.079 [DOI] [PMC free article] [PubMed] [Google Scholar]
57. Hunter WG, Kelly JP, McGarrah RW 3rd, Khouri MG, Craig D, Haynes C, Ilkayeva O, Stevens RD, Bain JR, Muehlbauer MJ, et al. Metabolomic profiling identifies novel circulating biomarkers of mitochondrial dysfunction differentially elevated in heart failure with preserved versus reduced ejection fraction: evidence for shared metabolic impairments in clinical heart failure. J Am Heart Assoc. 2016;5:e003190. doi: 10.1161/JAHA.115.003190 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Data S1

Tables S1–S6

Figures S1–S3

JAH3-14-e039248-s001.pdf^{(1.3MB, pdf)}

[jah311128-bib-0001] 1. Tsao CW, Aday AW, Almarzooq ZI, Alonso A, Beaton AZ, Bittencourt MS, Boehme AK, Buxton AE, Carson AP, Commodore‐Mensah Y, et al. Heart disease and stroke statistics‐2022 update: a report from the American Heart Association. Circulation. 2022;145:e153–e639. doi: 10.1161/CIR.0000000000001052 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0002] 2. McDonagh TA, Metra M, Adamo M, Gardner RS, Baumbach A, Bohm M, Burri H, Butler J, Celutkiene J, Chioncel O, et al. 2021 esc guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur Heart J. 2021;42:3599–3726. doi: 10.1093/eurheartj/ehab368 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0003] 3. Mann DL, Felker GM. Mechanisms and models in heart failure: a translational approach. Circ Res. 2021;128:1435–1450. doi: 10.1161/CIRCRESAHA.121.318158 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0004] 4. Chidsey CA, Harrison DC, Braunwald E. Augmentation of the plasma nor‐epinephrine response to exercise in patients with congestive heart failure. N Engl J Med. 1962;267:650–654. doi: 10.1056/NEJM196209272671305 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0005] 5. Deming QB, Luetscher JA Jr. Increased sodium‐retaining corticoid excretion in edema, with some observations on the effects of cortisone in nephrosis. J Clin Invest. 1950;29:808. doi: 10.1172/JCI102317 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0006] 6. Joshi SS, Singh T, Newby DE, Singh J. Sodium‐glucose co‐transporter 2 inhibitor therapy: mechanisms of action in heart failure. Heart. 2021;107:1032–1038. doi: 10.1136/heartjnl-2020-318060 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0007] 7. Selvaraj S, Fu Z, Jones P, Kwee LC, Windsor SL, Ilkayeva O, Newgard CB, Margulies KB, Husain M, Inzucchi SE, et al. Metabolomic profiling of the effects of dapagliflozin in heart failure with reduced ejection fraction: DEFINE‐HF. Circulation. 2022;146:808–818. doi: 10.1161/CIRCULATIONAHA.122.060402 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0008] 8. Brown NJ, Vaughan DE. Angiotensin‐converting enzyme inhibitors. Circulation. 1998;97:1411–1420. doi: 10.1161/01.CIR.97.14.1411 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0009] 9. Bozkurt B, Nair AP, Misra A, Scott CZ, Mahar JH, Fedson S. Neprilysin inhibitors in heart failure: the science, mechanism of action, clinical studies, and unanswered questions. JACC Basic Transl Sci. 2023;8:88–105. doi: 10.1016/j.jacbts.2022.05.010 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0010] 10. Bhatt AS, Abraham WT, Lindenfeld J, Bristow M, Carson PE, Felker GM, Fonarow GC, Greene SJ, Psotka MA, Solomon SD, et al. Treatment of HF in an era of multiple therapies: statement from the HF Collaboratory. JACC Heart Fail. 2021;9:1–12. doi: 10.1016/j.jchf.2020.10.014 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0011] 11. Muller J, Bertsch T, Volke J, Schmid A, Klingbeil R, Metodiev Y, Karaca B, Kim SH, Lindner S, Schupp T, et al. Narrative review of metabolomics in cardiovascular disease. J Thorac Dis. 2021;13:2532–2550. doi: 10.21037/jtd-21-22 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0012] 12. Ikegami R, Shimizu I, Yoshida Y, Minamino T. Metabolomic analysis in heart failure. Circ J. 2017;82:10–16. doi: 10.1253/circj.CJ-17-1184 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0013] 13. Guo L, Milburn MV, Ryals JA, Lonergan SC, Mitchell MW, Wulff JE, Alexander DC, Evans AM, Bridgewater B, Miller L, et al. Plasma metabolomic profiles enhance precision medicine for volunteers of normal health. Proc Natl Acad Sci USA. 2015;112:E4901–E4910. doi: 10.1073/pnas.1508425112 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0014] 14. German JB, Hammock BD, Watkins SM. Metabolomics: building on a century of biochemistry to guide human health. Metabolomics. 2005;1:3–9. doi: 10.1007/s11306-005-1102-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0015] 15. Wang TJ, Gupta DK. Metabolite profiles in heart failure: looking for unique signatures in a heterogeneous syndrome. J Am Coll Cardiol. 2015;65:1521–1524. doi: 10.1016/j.jacc.2015.02.019 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0016] 16. Zhou J, Chen X, Chen W, Zhong L, Cui M. Comprehensive plasma metabolomic and lipidomic analyses reveal potential biomarkers for heart failure. Mol Cell Biochem. 2021;476:3449–3460. doi: 10.1007/s11010-021-04159-5 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0017] 17. Stevenson LW, Perloff JK. The limited reliability of physical signs for estimating hemodynamics in chronic heart failure. JAMA. 1989;261:884–888. doi: 10.1001/jama.1989.03420060100040 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0018] 18. Shah MR, Hasselblad V, Stinnett SS, Gheorghiade M, Swedberg K, Califf RM, O'Connor CM. Hemodynamic profiles of advanced heart failure: association with clinical characteristics and long‐term outcomes. J Card Fail. 2001;7:105–113. doi: 10.1054/jcaf.2001.24131 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0019] 19. Yu B, Zanetti KA, Temprosa M, Albanes D, Appel N, Barrera CB, Ben‐Shlomo Y, Boerwinkle E, Casas JP, Clish C, et al. The Consortium of Metabolomics Studies (Comets): metabolomics in 47 prospective cohort studies. Am J Epidemiol. 2019;188:991–1012. doi: 10.1093/aje/kwz028 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0020] 20. Ghosh N, Lejonberg C, Czuba T, Dekkers K, Robinson R, Arnlov J, Melander O, Smith ML, Evans AM, Gidlof O, et al. Analysis of plasma metabolomes from 11 309 subjects in five population‐based cohorts. Sci Rep. 2024;14:8933. doi: 10.1038/s41598-024-59388-7 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0021] 21. Egerstedt A, Berntsson J, Smith ML, Gidlof O, Nilsson R, Benson M, Wells QS, Celik S, Lejonberg C, Farrell L, et al. Profiling of the plasma proteome across different stages of human heart failure. Nat Commun. 2019;10:5830. doi: 10.1038/s41467-019-13306-y [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0022] 22. Egerstedt A, Czuba T, Bronton K, Lejonberg C, Ruge T, Wessman T, Radegran G, Schulte J, Hartmann O, Melander O, et al. Bioactive adrenomedullin for assessment of venous congestion in heart failure. ESC Heart Fail. 2022;9:3543–3555. doi: 10.1002/ehf2.14018 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0023] 23. Martinsson A, Oest P, Wiborg MB, Reitan O, Smith JG. Longitudinal evaluation of ventricular ejection fraction and NT‐proBNP across heart failure subgroups. Scand Cardiovasc J. 2018;52:205–210. doi: 10.1080/14017431.2018.1461920 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0024] 24. Smith JG, Newton‐Cheh C, Almgren P, Struck J, Morgenthaler NG, Bergmann A, Platonov PG, Hedblad B, Engstrom G, Wang TJ, et al. Assessment of conventional cardiovascular risk factors and multiple biomarkers for the prediction of incident heart failure and atrial fibrillation. J Am Coll Cardiol. 2010;56:1712–1719. doi: 10.1016/j.jacc.2010.05.049 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0025] 25. Evans AM, DeHaven CD, Barrett T, Mitchell M, Milgram E. Integrated, nontargeted ultrahigh performance liquid chromatography/electrospray ionization tandem mass spectrometry platform for the identification and relative quantification of the small‐molecule complement of biological systems. Anal Chem. 2009;81:6656–6667. doi: 10.1021/ac901536h [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0026] 26. Evans AM, Bridgewater BR, Liu Q, Mitchell MW, Robinson RJ, Dai H, Stewart SJ, DeHaven CD, Miller LA. High resolution mass spectrometry improves data quantity and quality as compared to unit mass resolution mass spectrometry in high‐throughput profiling metabolomics. J Postgenom Drug Biomarker Dev. 2014;4:1. doi: 10.4172/2153-0769.1000132 [DOI] [Google Scholar]

[jah311128-bib-0027] 27. Dehaven CD, Evans AM, Dai H, Lawton KA. Organization of GC/MS and LC/MS metabolomics data into chemical libraries. J Cheminform. 2010;2:9. doi: 10.1186/1758-2946-2-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0028] 28. Stekhoven DJ, Bühlmann P. Missforest—non‐parametric missing value imputation for mixed‐type data. Bioinformatics. 2012;28:112–118. doi: 10.1093/bioinformatics/btr597 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0029] 29. Gromski PS, Xu Y, Kotze HL, Correa E, Ellis DI, Armitage EG, Turner ML, Goodacre R. Influence of missing values substitutes on multivariate analysis of metabolomics data. Metabolites. 2014;4:433–452. doi: 10.3390/metabo4020433 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0030] 30. Bijlsma S, Bobeldijk I, Verheij ER, Ramaker R, Kochhar S, Macdonald IA, van Ommen B, Smilde AK. Large‐scale human metabolomics studies: a strategy for data (pre‐) processing and validation. Anal Chem. 2006;78:567–574. doi: 10.1021/ac051495j [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0031] 31. Andersson B, Blomstrom‐Lundqvist C, Hedner T, Waagstein F. Exercise hemodynamics and myocardial metabolism during long‐term beta‐adrenergic blockade in severe heart failure. J Am Coll Cardiol. 1991;18:1059–1066. doi: 10.1016/0735-1097(91)90767-4 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0032] 32. Verbrugge FH, Borlaug BA. Getting the “right” perspective on angiotensin receptor‐neprilysin inhibition in heart failure. J Am Heart Assoc. 2020;9:e017292. doi: 10.1161/JAHA.120.017292 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0033] 33. Omar M, Jensen J, Frederiksen PH, Kistorp C, Videbaek L, Poulsen MK, Moller S, Ali M, Gustafsson F, Kober L, et al. Effect of empagliflozin on hemodynamics in patients with heart failure and reduced ejection fraction. J Am Coll Cardiol. 2020;76:2740–2751. doi: 10.1016/j.jacc.2020.10.005 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0034] 34. Schohn D, Jahn HA, Pelletier BC. Dose‐related cardiovascular effects of spironolactone. Am J Cardiol. 1993;71:40A–45A. doi: 10.1016/0002-9149(93)90244-7 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0035] 35. Boorsma EM, Ter Maaten JM, Damman K, Dinh W, Gustafsson F, Goldsmith S, Burkhoff D, Zannad F, Udelson JE, Voors AA. Congestion in heart failure: a contemporary look at physiology, diagnosis and treatment. Nat Rev Cardiol. 2020;17:641–655. doi: 10.1038/s41569-020-0379-7 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0036] 36. Cresci S, Pereira NL, Ahmad F, Byku M, de Las Fuentes L, Lanfear DE, Reilly CM, Owens AT, Wolf MJ. Heart failure in the era of precision medicine: a scientific statement from the American Heart Association. Circ Genom Precis Med. 2019;12:458–485. doi: 10.1161/HCG.0000000000000058 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0037] 37. Cheng S, Shah SH, Corwin EJ, Fiehn O, Fitzgerald RL, Gerszten RE, Illig T, Rhee EP, Srinivas PR, Wang TJ, et al. Potential impact and study considerations of metabolomics in cardiovascular health and disease: a scientific statement from the American Heart Association. Circ Cardiovasc Genet. 2017;10:10. doi: 10.1161/HCG.0000000000000032 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0038] 38. Murashige D, Jang C, Neinast M, Edwards JJ, Cowan A, Hyman MC, Rabinowitz JD, Frankel DS, Arany Z. Comprehensive quantification of fuel use by the failing and nonfailing human heart. Science. 2020;370:364–368. doi: 10.1126/science.abc8861 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0039] 39. Boyle JG, Davidson DF, Perry CG, Connell JM. Comparison of diagnostic accuracy of urinary free metanephrines, vanillyl mandelic acid, and catecholamines and plasma catecholamines for diagnosis of pheochromocytoma. J Clin Endocrinol Metab. 2007;92:4602–4608. doi: 10.1210/jc.2005-2668 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0040] 40. Lund A, Nordrehaug JE, Slettom G, Solvang SH, Pedersen EK, Midttun O, Ulvik A, Ueland PM, Nygard O, Giil LM. Plasma kynurenines and prognosis in patients with heart failure. PLoS One. 2020;15:e0227365. doi: 10.1371/journal.pone.0227365 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0041] 41. Konishi M, Ebner N, Springer J, Schefold JC, Doehner W, Dschietzig TB, Anker SD, von Haehling S. Impact of plasma kynurenine level on functional capacity and outcome in heart failure‐results from Studies Investigating Co‐Morbidities Aggravating Heart Failure (SICA‐HF). Circ J. 2016;81:52–61. doi: 10.1253/circj.CJ-16-0791 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0042] 42. Terakata M, Fukuwatari T, Kadota E, Sano M, Kanai M, Nakamura T, Funakoshi H, Shibata K. The niacin required for optimum growth can be synthesized from l‐tryptophan in growing mice lacking tryptophan‐2,3‐dioxygenase. J Nutr. 2013;143:1046–1051. doi: 10.3945/jn.113.176875 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0043] 43. Moffett JR, Arun P, Puthillathu N, Vengilote R, Ives JA, Badawy AA, Namboodiri AM. Quinolinate as a marker for kynurenine metabolite formation and the unresolved question of NAD⁺ synthesis during inflammation and infection. Front Immunol. 2020;11:31. doi: 10.3389/fimmu.2020.00031 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0044] 44. Walker MA, Tian R. Raising NAD in heart failure: time to translate? Circulation. 2018;137:2274–2277. doi: 10.1161/CIRCULATIONAHA.117.032626 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0045] 45. Diguet N, Trammell SAJ, Tannous C, Deloux R, Piquereau J, Mougenot N, Gouge A, Gressette M, Manoury B, Blanc J, et al. Nicotinamide riboside preserves cardiac function in a mouse model of dilated cardiomyopathy. Circulation. 2018;137:2256–2273. doi: 10.1161/CIRCULATIONAHA.116.026099 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0046] 46. Núñez J, de la Espriella R, Rossignol P, Voors AA, Mullens W, Metra M, Chioncel O, Januzzi JL, Mueller C, Richards AM, et al. Congestion in heart failure: a circulating biomarker‐based perspective. A review from the biomarkers working group of the Heart Failure Association, European Society of Cardiology. Eur J Heart Fail. 2022;24:1751–1766. doi: 10.1002/ejhf.2664 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0047] 47. Puetz A, Artati A, Adamski J, Schuett K, Romeo F, Stoehr R, Marx N, Federici M, Lehrke M, Kappel BA. Non‐targeted metabolomics identify polyamine metabolite acisoga as novel biomarker for reduced left ventricular function. ESC Heart Fail. 2022;9:564–573. doi: 10.1002/ehf2.13713 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0048] 48. Nayak A, Liu C, Mehta A, Ko YA, Tahhan AS, Dhindsa DS, Uppal K, Jones DP, Butler J, Morris AA, et al. N8‐acetylspermidine: a polyamine biomarker in ischemic cardiomyopathy with reduced ejection fraction. J Am Heart Assoc. 2020;9:e016055. doi: 10.1161/JAHA.120.016055 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0049] 49. Ottosson F, Smith E, Fernandez C, Melander O. Plasma metabolites associate with all‐cause mortality in individuals with type 2 diabetes. Metabolites. 2020;10:315. doi: 10.3390/metabo10080315 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0050] 50. Cheng ML, Wang CH, Shiao MS, Liu MH, Huang YY, Huang CY, Mao CT, Lin JF, Ho HY, Yang NI. Metabolic disturbances identified in plasma are associated with outcomes in patients with heart failure: diagnostic and prognostic value of metabolomics. J Am Coll Cardiol. 2015;65:1509–1520. doi: 10.1016/j.jacc.2015.02.018 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0051] 51. Eisenberg T, Abdellatif M, Schroeder S, Primessnig U, Stekovic S, Pendl T, Harger A, Schipke J, Zimmermann A, Schmidt A, et al. Cardioprotection and lifespan extension by the natural polyamine spermidine. Nat Med. 2016;22:1428–1438. doi: 10.1038/nm.4222 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0052] 52. Mayerhofer CCK, Ueland T, Broch K, Vincent RP, Cross GF, Dahl CP, Aukrust P, Gullestad L, Hov JR, Troseid M. Increased secondary/primary bile acid ratio in chronic heart failure. J Card Fail. 2017;23:666–671. doi: 10.1016/j.cardfail.2017.06.007 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0053] 53. Alnouti Y. Bile acid sulfation: a pathway of bile acid elimination and detoxification. Toxicol Sci. 2009;108:225–246. doi: 10.1093/toxsci/kfn268 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0054] 54. Khurana S, Raufman JP, Pallone TL. Bile acids regulate cardiovascular function. Clin Transl Sci. 2011;4:210–218. doi: 10.1111/j.1752-8062.2011.00272.x [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0055] 55. Causse E, Fournier P, Roncalli J, Salvayre R, Galinier M. Serum allantoin and aminothiols as biomarkers of chronic heart failure. Acta Cardiol. 2017;72:397–403. doi: 10.1080/00015385.2017.1335104 [DOI] [PubMed] [Google Scholar]

[jah311128-bib-0056] 56. Ahmad T, Kelly JP, McGarrah RW, Hellkamp AS, Fiuzat M, Testani JM, Wang TS, Verma A, Samsky MD, Donahue MP, et al. Prognostic implications of long‐chain acylcarnitines in heart failure and reversibility with mechanical circulatory support. J Am Coll Cardiol. 2016;67:291–299. doi: 10.1016/j.jacc.2015.10.079 [DOI] [PMC free article] [PubMed] [Google Scholar]

[jah311128-bib-0057] 57. Hunter WG, Kelly JP, McGarrah RW 3rd, Khouri MG, Craig D, Haynes C, Ilkayeva O, Stevens RD, Bain JR, Muehlbauer MJ, et al. Metabolomic profiling identifies novel circulating biomarkers of mitochondrial dysfunction differentially elevated in heart failure with preserved versus reduced ejection fraction: evidence for shared metabolic impairments in clinical heart failure. J Am Heart Assoc. 2016;5:e003190. doi: 10.1161/JAHA.115.003190 [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Metabolite Profiles of Heart Failure, Central Hemodynamic Derangement, and Response to Heart Transplantation

Carl Lejonberg, MD

Tomasz Czuba, MSc

Anna Egerstedt, MD

Nilanjana Ghosh, PhD

Selvi Celik, PhD

Charlotta Ljungman, MD, PhD

Jakob Lundgren, MD, PhD

Göran Rådegran, MD, PhD

Olof Gidlöf, PhD

J Gustav Smith, MD, PhD

Abstract

Background

Methods

Results

Conclusions

Nonstandard Abbreviations and Acronyms

Clinical Perspective.

What Is New?

What Are the Clinical Implications?

METHODS

Study Samples

Patients With HF Undergoing Right Heart Catheterization

HF Patients From Hospital‐Based Outpatient Management Program

Metabolite Profiling

Statistical Analysis

RESULTS

Table 1.

Metabolite Profiles in HF

Figure 1. Manhattan plot of metabolite associations with heart failure.

Figure 2. Metabolite classes associated with heart failure.

Metabolite Responses to Heart Transplantation

Table 2.

Figure 3. Correlation of heart failure associated plasma metabolites.

Relation of Plasma Metabolites With Hemodynamic Parameters

Figure 4. Association of heart failure‐associated plasma metabolites with central hemodynamic measures.

Relation of Plasma Metabolites With Measures of Symptom Burden

Table 3.

Relation of Plasma Metabolites With Drug Therapy

DISCUSSION

Conclusions

Sources of Funding

Disclosures

Supporting information

Acknowledgments

REFERENCES

Associated Data

Supplementary Materials

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases