Author + information
- Received December 16, 2016
- Revision received April 10, 2017
- Accepted April 14, 2017
- Published online June 14, 2017.
- Jasper Tromp, MDa,
- Mohsin A.F. Khan, PhDa,b,
- Robert J. Mentz, MDc,
- Christopher M. O’Connor, MDd,
- Marco Metra, MDe,
- Howard C. Dittrich, MDf,
- Piotr Ponikowski, MDg,
- John R. Teerlink, MDh,
- Gad Cotter, MDi,
- Beth Davison, PhDi,
- John G.F. Cleland, MDj,
- Michael M. Givertz, MDk,
- Daniel M. Bloomfield, MDl,
- Dirk J. van Veldhuisen, MD, PhDa,
- Hans L. Hillege, PhDa,m,
- Adriaan A. Voors, MD, PhDa,∗ ( and )
- Peter van der Meer, MD, PhDa
- aDepartment of Cardiology, University of Groningen, University Medical Center Groningen, Groningen, the Netherlands
- bHeart Failure Research Centre, Academic Medical Centre, Amsterdam, the Netherlands
- cDuke University Medical Center, Durham, North Carolina
- dInova Heart and Vascular Institute, Falls Church, Virginia
- eUniversity of Brescia, Brescia, Italy
- fCardiovascular Research Center, University of Iowa Carver College of Medicine, Iowa City, Iowa
- gMedical University, Clinical Military Hospital, Wroclaw, Poland
- hUniversity of California at San Francisco and San Francisco Veterans Affairs Medical Center, San Francisco, California
- iMomentum Research, Durham, North Carolina
- jUniversity of Hull, Kingston upon Hull, United Kingdom
- kBrigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts
- lMerck & Co., Inc., Kenilworth, New Jersey
- mDepartment of Epidemiology, University of Groningen, University Medical Center Groningen, Groningen, the Netherlands
- ↵∗Address for correspondence:
Dr. Adriaan A. Voors, Department of Cardiology, University Medical Center Groningen, Hanzeplein 1, 9713GZ, Groningen, the Netherlands.
Objectives In this study, the authors used biomarker profiles to characterize differences between patients with acute heart failure with a midrange ejection fraction (HFmrEF) and compare them with patients with a reduced (heart failure with a reduced ejection fraction [HFrEF]) and preserved (heart failure with a preserved ejection fraction [HFpEF]) ejection fraction.
Background Limited data are available on biomarker profiles in acute HFmrEF.
Methods A panel of 37 biomarkers from different pathophysiological domains (e.g., myocardial stretch, inflammation, angiogenesis, oxidative stress, hematopoiesis) were measured at admission and after 24 h in 843 acute heart failure patients from the PROTECT trial. HFpEF was defined as left ventricular ejection fraction (LVEF) of ≥50% (n = 108), HFrEF as LVEF of <40% (n = 607), and HFmrEF as LVEF of 40% to 49% (n = 128).
Results Hemoglobin and brain natriuretic peptide levels (300 pg/ml [HFpEF]; 397 pg/ml [HFmrEF] 521 pg/ml [HFrEF]; ptrend <0.001) showed an upward trend with decreasing LVEF. Network analysis showed that in HFrEF interactions between biomarkers were mostly related to cardiac stretch, whereas in HFpEF, biomarker interactions were mostly related to inflammation. In HFmrEF, biomarker interactions were both related to inflammation and cardiac stretch. In HFpEF and HFmrEF (but not in HFrEF), remodeling markers at admission and changes in levels of inflammatory markers across the first 24 h were predictive for all-cause mortality and rehospitalization at 60 days (pinteraction <0.05).
Conclusions Biomarker profiles in patients with acute HFrEF were mainly related to cardiac stretch and in HFpEF related to inflammation. Patients with HFmrEF showed an intermediate biomarker profile with biomarker interactions between both cardiac stretch and inflammation markers. (PROTECT-1: A Study of the Selective A1 Adenosine Receptor Antagonist KW-3902 for Patients Hospitalized With Acute HF and Volume Overload to Assess Treatment Effect on Congestion and Renal Function; NCT00328692)
Heart failure with a midrange ejection fraction (HFmrEF) has recently been recognized as a new entity within the heart failure (HF) syndrome (1,2). There is a limited understanding of the differences in pathophysiological mechanisms behind HFmrEF, and how these relate to HF with a reduced (heart failure with a reduced ejection fraction [HFrEF]) and with a preserved (heart failure with a preserved ejection fraction [HFpEF]) ejection fraction. Previous attempts to understand potential differences in HFrEF and HFpEF have used biomarker-based approaches (3–7). In these conventional biomarker-based studies, baseline biomarker levels and the prognostic value of different biomarkers have been observed between HFrEF and HFpEF (5,6). However, these approaches were restricted to a limited number of biomarkers measured at a single time point using conventional statistical methods with limited power to uncover underlying pathophysiological differences. Additionally, biomarker profiles of HFmrEF have not been investigated (8–10).
Recently, novel approaches have been useful in increasing the understanding of the pathophysiology of chronic HF by uncovering biomarker associations, previously overlooked by conventional methods (10,11). In the current study, we aimed to characterize biomarker profiles of patients with HFmrEF and compared these with biomarker profiles of HFrEF and HFpEF (1).
Study design and population
This study was performed in a subcohort of the PROTECT (Placebo-Controlled Randomized Study of the Selective Adenosine A1 Receptor Antagonist Rolofylline for Patients Hospitalized with Acute Decompensated Heart Failure and Volume Overload to Assess Treatment Effect on Congestion and Renal Function) trial. The results and methodology of PROTECT have been published previously (12–14). In short, the PROTECT trial was a multicenter, randomized, double-blinded, placebo-controlled trial assessing the effect of the selective A1 adenosine receptor antagonist rolofylline in 2,033 patients with a history of HF, who were admitted with acute HF (AHF) and mild to moderate renal dysfunction. Patients eligible for inclusion had N-terminal pro-brain natriuretic peptide (NT-proBNP) levels of >2,000 pg/ml with dyspnea at rest or at mild exertion. Patients with severe renal dysfunction or potassium levels of <3.1 mmol/l were excluded (12). The overall results of this trial were neutral (14). Biomarker measurements were performed in 1,266 patients. This study assessed a subcohort of 843 patients with available measurements of left ventricular ejection fraction (LVEF) and biomarkers at admission, which were similar in characteristics to the original study population (Online Table 1). Subsequent biomarker samples after 24 hours were available in 790 patients.
Study measurements and laboratory tests
Blood sampling was performed at admission before the administration of the study drug and after 24 h. Echocardiographic assessment of LVEF was performed at admission or within 6 months before admission. A total of 435 of the echocardiograms (52%) were performed at or around admission. HFpEF was defined as having an LVEF of ≥50%, and HFrEF was defined as an LVEF of <40%. Patients with a LVEF between 40% and 49% were considered to have HFmrEF (1). A panel of 27 novel and established biomarkers were measured by Alere Inc. (San Diego, California) in all available samples. Table 1 summarizes the biomarkers according to pathophysiological domain. A literature summary for each biomarker was performed previously (11). The classification of biomarkers is based on current literature; however, the pathophysiological mechanism behind each biomarker should be judged for each biomarker individually. Galectin-3, myeloperoxidase (MPO) and neutrophil gelatinase-associated lipocalin were measured using sandwich enzyme-linked immunosorbent assays on a microtiter plate; angiogenin and C-reactive protein were measured using competitive enzyme-linked immunosorbent assays on a Luminex platform; D-dimer, endothelial cell-selective adhesion molecule, growth differentiation factor 15, lymphotoxin beta receptor (LTBR), mesothelin, neuropilin, N-terminal pro C-type natriuretic peptide, osteopontin, procalcitonin, pentraxin-3, periostin, polymeric immunoglobulin receptor, pro-adrenomedullin, prosaposin B, receptor for advanced glycation end product (RAGE), soluble ST2, syndecan-1, tumor necrosis factor-receptor 1α (TNF-R1α), TROY, vascular endothelial growth receptor-1, and WAP 4-disulphide core domain protein HE4 were measured using sandwich enzyme-linked immunosorbent assays on a Luminex platform. A panel of 4 biomarkers—endothelin-1, IL-6, kidney injury molecule (KIM)-1, and cardiac-specific troponin I—was measured in frozen plasma samples collected at baseline using high-sensitive single molecule counting technology (RUO, Erenna Immunoassay System, Singulex Inc., Alameda, California). Research assays of MR-pro-adrenomedullin, galectin-3, and ST2 were developed by Alere, and have not been standardized to the commercialized assays used in research or in clinical use. The extent to which each Alere assay correlates with the commercial assay is not fully characterized. Assay information included interassay coefficient of variation are provided in Online Table 2. The estimated glomerular filtration rate was based on the simplified Modification of Diet in Renal Disease (15).
The primary outcome of this study was all-cause mortality and/or rehospitalization at 60 days after admission. This outcome was chosen because of the relatively high event rate in comparison with the other outcomes in the PROTECT trial. A blinded clinical events committee adjudicated the outcome.
Continuous variables are presented as mean ± SD or medians with interquartile ranges. Categorical variables are presented as numbers or percentages. Intergroup differences were analyzed using Student t test, Mann-Whitney U test, Kruskal-Wallis test, analysis of variance, or chi-square test where appropriate.
To correct for multiple comparisons, principal component analysis (PCA) was performed with HFrEF and HFpEF as categorical variables, using an established method described elsewhere (16). A total of 27 principal components cumulatively explained >95% of the variation observed in the dataset when comparing HFrEF and HFpEF (Online Figures 1 and 2). The corrected significance level for multiple testing was thus set at p < 0.05/27. After this, a Spearman’s rank correlation coefficient was calculated for each possible biomarker pair in the HFrEF cohort of patients and the procedure was repeated for HFpEF and HFmrEF. This resulted in 3 sets of R values with associated p values for HFrEF, HFmrEF, and HFpEF. To adjust for multiple testing, only those correlations passing the adjusted p value cutoff calculated from the principal component analysis were deemed statistically significant and subsequently retained. These significant correlation coefficients for HFrEF, HFmrEF, and HFpEF were then displayed graphically as heatmaps with associated disease domains for all biomarkers. Network analysis was performed to analyze associations between biomarkers in HFrEF, HFmrEF, and HFpEF. Subsequently, all significant associations found within HFrEF, HFmrEF, and HFpEF were depicted separately as circular networks, consisting of nodes (biomarkers) and edges (associations). In each network, the size and color of the nodes reflect the clustering coefficient of each biomarker, and the thickness of the lines (edges) represent the strength of the interbiomarker associations (determined by Spearman's rank coefficient R values).
To study the possible differential relationship with outcome of biomarkers, a univariable interaction test was performed between LVEF and the biomarker levels at admission or a change in biomarker levels between admission and the first 24 h. After this, a multivariable interaction test was performed correcting for a risk engine containing 8 variables, specifically designed for this cohort (17). These variables include age, previous HF hospitalizations, peripheral edema, systolic blood pressure, serum sodium, urea, creatinine, and albumin levels at admission. Univariable and multivariable associations of biomarkers with outcome were tested using Cox regression analysis; due to the exploratory nature of these analyses, a p value of <0.05 was deemed statistically significant for the interaction test.
All tests were performed 2-sided and p < 0.05 was considered significant. All statistical analyses were performed using STATA version 11.0 (StataCorp LP, College Station, Texas) and R version 3.2.4.
Baseline characteristics are presented in Table 2. Patients with HFmrEF were older than HFrEF patients, but younger than HFpEF (age 71 years vs. 68 and 74, years respectively; p value for trend <0.001). With increasing LVEF, the percentage of female patients, body mass index, systolic blood pressure, and diastolic blood pressure was higher (ptrend <0.05). We observed less mitral regurgitation, fewer previous HF hospitalizations during the past year, and less ischemic heart disease and myocardial infarction with increasing LVEF (all ptrend 0.001). Median time since the previous HF hospitalization was 52 days and did not differ between HFrEF, HFmrEF, and HFpEF (p = 0.776). In contrast, a history of hypertension (ptrend <0.001) and atrial fibrillation (ptrend = 0.014) was found more often with increasing LVEF. A direct comparison between HFmrEF–HFrEF and HFmrEF–HFpEF confirms these results (Online Tables 3 and 4).
Biomarker levels on admission are presented in Table 3. With increasing LVEF, we found increasing levels of C-reactive protein, neutrophil gelatinase-associated lipocalin, KIM-1, and platelet count, and decreasing levels of growth differentiation factor-15, BNP, troponin-I, red blood cells, hemoglobin, and endothelin-1. After correction for multiple comparisons, the up- or down-sloping trend remained significant for BNP, KIM-1, red blood cells, and hemoglobin. When examining a change of biomarkers from admission to 24 h, troponin-I increased more in patients with HFrEF than in patients with HFmrEF and HFpEF; however, significance was lost after correction for multiple comparisons (Online Table 5). No significant interaction was found between the study drug and LVEF for biomarkers that significantly differed between HFrEF; HFmrEF and HFpEF also no significant interactions were observed between timing of echocardiography and LVEF for biomarker levels (all pinteraction >0.10).
Heatmaps of biomarker associations are available in Online Figures 3 to 5. The results of network analysis are shown in Figures 1 to 3⇓⇓⇓. On admission, network analysis in HFrEF showed troponin-I, BNP, and prosaposin B to be a hub. A biomarker that is a hub has a high clustering coefficient. A high clustering coefficient suggests a certain centrality of the biomarker within the network, where a large number of the biomarker interactions are mediated through the hub. In HFpEF, angiogenin, hemoglobin, galectin-3 as well as d-dimer were hubs. Compared with HFrEF, BNP is only moderately associated with other biomarkers in HFpEF on admission. Interestingly, in HFmrEF, hemoglobin, red blood cells, and endothelin-1 as well as BNP and galectin-3 were clear hubs at admission. After 24 h, interactions of biomarkers in patients with HFrEF were mainly associated with BNP and endothelin-1. In comparison, after 24 h, biomarkers in HFpEF were mainly associated with inflammation markers pentraxin-3 and RAGE, as well as with the remodeling marker osteopontin, the angiogenesis marker angiogenin, and the hematopoiesis markers hemoglobin and red blood cell count, as well as the renal function marker neutrophil gelatinase-associated lipocalin. Interestingly, BNP remains a small hub in HFpEF. In HFmrEF, after 24 h, the association between BNP and other biomarkers became very limited. Furthermore, the remodeling marker galectin-3 and inflammation marker RAGE were continuous hubs at admission through the first 24 h.
Biomarker levels and outcome
Associations of biomarkers levels at admission with outcome are shown in Online Table 6. Remodeling markers syndecan-1 (p = 0.047) and galectin-3 (p = 0.024) showed a significant interaction for the primary outcome. Here, syndecan-1 showed a significant association with outcome in HFmrEF and HFpEF, but not in HFrEF. Also, galectin-3 showed significant predictive value in HFpEF, but not in HFmrEF and HFrEF.
The associations with outcome of a change of biomarker levels within the first 24 hours is show in Online Table 7. A significant multivariable interaction was found for the inflammation biomarkers pentraxin-3 (p = 0.025), RAGE (p = 0.037), TNF-R1α (p = 0.004), oxidative stress marker MPO (p = 0.017), and the endothelial function marker pro-adrenomedullin (p = 0.016) as well as the arteriosclerosis marker LTBR (p = 0.009). After multivariable correction, pentraxin-3 was more predictive in HFmrEF and HFpEF, but not in HFrEF. A change in levels of TNF-R1α, MPO, and LTBR were related to outcome in HFpEF, but not in HFrEF and HFmrEF. Interestingly, a change of endothelial function marker pro-ADM only had predictive power in HFmrEF, but not in HFrEF and HFpEF (Online Table 7).
This study demonstrates differential biomarker profiles between AHF patients with HFrEF, HFmrEF, and HFpEF. Network analysis showed that, in HFmrEF, interaction between biomarkers were associated with BNP, galectin-3, and endothelin-1. In contrast, interactions between biomarkers in HFrEF were mostly associated with BNP, KIM-1, and troponin-I, whereas in HFpEF, biomarkers associated with inflammation and endothelial function played a central role. Both in terms of clinical characteristics and biomarker profiles, patients with HFmrEF were in between HFpEF and HFrEF. Biomarkers profiles of HFmrEF, HFpEF, and HFrEF remained relatively stable throughout the first 24 h after hospital admission. With regard to outcome, markers of inflammation showed independent predictive value in HFmrEF and HFpEF, but not in HFrEF. Levels of remodeling markers syndecan-1 and galectin-3 showed predictive value in HFmrEF and HFpEF, but not in HFrEF. Of note, pro-ADM showed predictive value in HFmrEF, but not in HFrEF and HFpEF.
Biomarker levels of patients with HFmrEF were between HFrEF and HFpEF. HFrEF patients had higher levels of biomarkers related to cardiac stretch and hematopoiesis. Network analysis showed an interassociation between biomarkers related to inflammation and cardiac stretch in HFmrEF. In HFpEF, associations related to inflammation and BNP only played a very marginal role in associations between biomarkers. In HFrEF, BNP had a more prominent role in network analyses both at admission and after 24 h. In HFmrEF, a mix of associations between cardiac stretch and inflammation was observed. In an earlier publication in a chronic HF setting, associations between inflammation markers were seen in HFpEF, whereas in HFrEF associations were found between cardiac stretch markers (10). Indeed, also in this study, network analysis revealed patterns, which were previously unknown in HFrEF and HFpEF. Biomarkers in the intermediate group were more related to HFpEF than to HFrEF in this subanalysis of the TIME-CHF trial (10). This could potentially be explained by the difference in inclusion criteria, where for the PROTECT trial a minimum NT-proBNP of >2,000 pg/ml had to be present at admission, although this was not required for the TIME-CHF trial (18). HFpEF patients are known to have lower BNP and NT-proBNP levels compared with HFrEF, which could explain why the proportion of HFpEF patients in the PROTECT trial is lower (19).
Remodeling marker syndecan-1 had predictive value in HFmrEF and HFpEF, but not in HFrEF. This was previously shown in a stable HF setting, where syndecan-1 had predictive value in HFpEF but not in HFrEF (5). In an earlier publication about syndecan-1, HFpEF was defined at an LVEF of >40%, suggesting that syndecan-1 also in a chronic setting provides predictive value in both HFmrEF and HFpEF. Galectin-3 only showed predictive value in HFpEF, but not in HFrEF and HFmrEF, in line with an earlier publication (20). Furthermore, a change in levels of inflammation markers pentraxin-3 and TNF-R1α were predictive in HFpEF, but not in HFrEF. The role of pentraxin-3 in HFpEF is readily known (21). In earlier reports, circulating TNF-R1a levels predicted incident cardiovascular disease, including HF (22). In a particular study addressing chronic HF, TNF-R1α was the strongest predictor of long-term mortality (23). Higher levels of TNF-R1α were previously reported in HFpEF patients (24). Levels of MPO were previously correlated with New York Heart Association functional class and diastolic HF, and is considered to be both a marker of inflammation and oxidative stress (25,26). A change in levels of MPO was predictive in HFpEF, but not in HFmrEF and HFrEF. LTBR is a member of the tumor necrosis factor family (27,28). Activation of LTBR results in lymphocyte recruitment and is associated with inflammatory responses in atherosclerosis (27,29). No data are available on predictive value in HF, and this is the first study reporting the differential involvement in predicting outcome in AHF patients with HFrEF, HFmrEF, and HFpEF. Of note, TNF-R1α and LTBR are members of the TNF-α super family of cytokines, suggesting a possible involvement of this family of proteins. Members of the TNF-α super family are involved in nitric oxide handling, which is considered a key mechanism in HFpEF. Whether other members of the TNF-α super family have a significant role in the pathophysiology of HFpEF needs to be explored further.
The clinical implications of this study are 4-fold. First, both the clinical and biomarker profiles of patients with HFmrEF were in between of HFrEF and HFpEF. This suggests that HFmrEF is a mix of patients similar to both HFrEF and HFpEF. There could be a considerable number of patients among HFmrEF who are closer to HFrEF and might benefit from existing HF-guideline directed therapy. Previously, large HF trials had either excluded or embedded HFmrEF within the HFpEF group (1). Future studies should distinguish which HFmrEF patients are closer to HFrEF and which are closer to HFpEF. Biomarkers could aid in recognizing patients with HFmrEF that are closer to HFrEF. These patients are likely characterized by high NT-proBNP and high cardiac damage markers, whereas having lower levels of inflammation markers compared with HFpEF patients. These patients could subsequently benefit from guideline-directed therapy and can possibly be included in future HF trials with HFrEF patients. Second, patients with HFpEF have a distinct biomarker profile from those with HFrEF, with patients with HFpEF having lower levels of cardiac stretch markers. Also, inflammation-related biomarkers had more predictive value in HFpEF and HFmrEF than in HFrEF. Third, overall biomarker profiles stay relatively stable in HFrEF, HFmrEF, and HFpEF during hospitalization; biomarker associations are more angiogenesis and inflammation related in HFpEF, cardiac stretch related in HFrEF, and both cardiac stretch and inflammation related in HFmrEF.
This study is a retrospective post hoc analysis, which is accompanied by a possible selection bias. Not all patients had complete biomarker data available at admission and after 24 h, creating a potential selection bias. Also, despite the large number of biomarker available, the choice for biomarkers was restricted by limited sample availability. It also needs to be emphasized that this is a data-driven approach and causality cannot be proven. Results of this study need to be validated in a different population. Additionally, some echocardiographic measurements were performed 6 months before admission. This did not seem to influence biomarker levels in HFrEF, HFmrEF, and HFpEF; however, we could not correct for this in network analysis. Differences with regard to outcome prediction should only be interpreted in the context of pathophysiological differences between HFrEF, HFmrEF, and HFpEF, and not with respect to possible clinical usefulness (10). For the latter, the relatively low number of events confounds the results with regard to predictive value. This was especially true for other outcomes (e.g., 30-day mortality) in the PROTECT trial, for which the number of events was even lower than the outcome used, making useful statistics on these outcomes not possible. Confirmation of the differential predictive value found is needed in more inclusive independent trials with larger number of events and HFmrEF and HFpEF patients.
Clinical characteristics and biomarker profiles of patients with HFmrEF are between patients with HFrEF and HFpEF, suggesting HFmrEF to be a heterogeneous group. Biomarker associations in HFpEF were mostly inflammation based, whilst being more cardiac stretch based in HFrEF. Biomarkers related to inflammation and cardiac remodeling had predictive value in HFmrEF and HFpEF, but not in HFrEF. These data suggest that patients with HFmrEF are a mix of HFrEF and HFpEF patients. Distinguishing HFmrEF patients closer to HFrEF could have important therapeutic consequences for this group.
COMPETENCY IN MEDICAL KNOWLEDGE: Differences between AHF patients with HFmrEF, HFrEF, and HFpEF have not been well-characterized. Results from this study suggest that AHF patients with HFpEF have a significantly different biomarker profile from patients with HFrEF. Herein, we found that inflammation plays a larger role in patients with HFpEF compared with HFrEF. Second, patients with HFmrEF are in between patients with HFpEF and HFrEF. This finding suggests that these patients should be carefully considered when treating according to guidelines, because some of them might be closer to HFrEF and some might be closer to HFpEF. Last, a change in inflammation biomarker levels might hold prognostic value for patients with HFpEF and HFmrEF.
TRANSLATIONAL OUTLOOK: Biomarker-based characterization of patient populations might help to identify novel treatment targets as well as decipher disease heterogeneity and underlying differences in pathophysiology. Although biomarker-based clinical studies can be considered a crude tool, it can be the first step in identifying novel disease entities and pathophysiological targets. Findings from biomarker-based studies, including this one, should be validated in an experimental setting.
Alere and Singulex kindly provided assays and performed biomarker measurements.
For supplemental tables and figures, please see the online version of this article.
The PROTECT trial was supported by NovaCardia, a subsidiary of Merck & Co. Dr. Cleland was on the Steering Committee (and received payment) for the PROTECT trial; served on the advisory board (and received payment) for MSD. Dr. O’Connor is a consultant to Merck & Co., Inc. Dr. Ponikowski has received honoraria from Merck & Co., Inc. Drs. Davison and Cotter are employees of Momentum Research Inc., which was contracted to perform work on the project by Merck & Co., Inc. Dr. Metra has received honoraria and reimbursements from NovaCardia (sponsor of the study) and Merck & Co., Inc. Dr. Givertz has received institutional research support and served on a scientific advisory board for Merck & Co., Inc. Dr. Teerlink has received research funds and consulting fees from Merck & Co., Inc. Dr. Bloomfield is an employee of Merck & Co., Inc. Dr. Dittrich served as a consultant to Merck & Co., Inc. Dr. Voors has received speaker and consultancy fees from Merck & Co., Inc.; was on the Steering Committee for the PROTECT trial; and received research support from Alere, Singulex, and Sphingotec. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
Barry H. Greenberg, MD, served as Guest Editor for this paper.
- Abbreviations and Acronyms
- acute heart failure
- heart failure
- heart failure with a midrange ejection fraction
- heart failure with a preserved ejection fraction
- heart failure with a reduced ejection fraction
- kidney injury molecule
- lymphotoxin beta receptor
- left ventricular ejection fraction
- N-terminal pro-brain natriuretic peptide
- receptor for advanced glycation end product
- tumor necrosis factor-receptor 1α
- Received December 16, 2016.
- Revision received April 10, 2017.
- Accepted April 14, 2017.
- 2017 American College of Cardiology Foundation
- Ponikowski P.,
- Voors A.A.,
- Anker S.D.,
- et al.
- Maisel A.S.,
- Choudhary R.
- Tromp J.,
- van der Pol A.,
- Klip I.T.,
- et al.
- Van Veldhuisen D.J.,
- Linssen G.C.M.,
- Jaarsma T.,
- et al.
- Braunwald E.
- Schmitter D.,
- Cotter G.,
- Voors A.A.
- Demissei B.G.,
- Cleland J.G.,
- O’Connor C.M.,
- et al.
- Weatherley B.D.,
- Cotter G.,
- Dittrich H.C.,
- et al.
- Voors A.A.,
- Dittrich H.C.,
- Massie B.M.,
- et al.
- Smilde T.D.J.,
- van Veldhuisen D.J.,
- Navis G.,
- Voors A.A.,
- Hillege H.L.
- Cleland J.G.,
- Chiswell K.,
- Teerlink J.R.,
- et al.
- van Veldhuisen D.J.,
- Linssen G.C.M.,
- Jaarsma T.,
- et al.
- Paulus W.J.,
- Tschöpe C.
- Vasan R.S.,
- Sullivan L.M.,
- Roubenoff R.,
- et al.
- Deswal A.,
- Petersen N.J.,
- Feldman A.M.,
- Young J.B.,
- White B.G.,
- Mann D.L.
- Putko B.N.,
- Wang Z.,
- Lo J.,
- et al.
- Carr A.C.,
- McCall M.R.,
- Frei B.
- Tang W.H.,
- Tong W.,
- Troughton R.W.,
- et al.
- Lötzer K.,
- Döpping S.,
- Connert S.,
- et al.