![ORCID Icon]({"type":"image","src":"/templates/jsp/_style2/_tandf/images/orcid.svg"} "ORCID Icon"){kind=small}
http://orcid.org/0000-0001-8066-2423
Abstract
Abstract
Objectives. Soluble suppression of tumorigenecity 2 (sST2) is prognostic in acute and chronic heart failure with reduced ejection fraction (HFrEF) but less studied in HF with preserved EF (HFpEF). We evaluated sST2 concentrations, correlations with biomarkers and echocardiographic measures of diastolic and systolic function, and associations with outcomes in HFpEF and HFrEF. Design and results. A total of 193 subjects from three different cohorts were included. Eighty-six HFpEF patients were obtained from the Karolinska Rennes (KaRen) study, 86 patients with HFrEF were recruited from referrals to Karolinska University Hospital for advanced assessment of HF, and 21 controls were included (ClinicalTrials.gov Identifier for KaRen: NCT01091467). HFrEF and controls cohorts did not have ClinicalTrials.gov registrations. sST2 was lower in HFpEF, median (interquartile range); 23 (17–31) compared to HFrEF; 35 (23–52) µg/L, p < .001. In both HFpEF and HFrEF, sST2 correlated positively with NT-proBNP (HFpEF rs=0.392, p < .001 and HFrEF rs=0.466, p < .001). In HFpEF, sST2 correlated to left atrial volume index (rs=0.276, p = .019) but not to E/E´, nor to left ventricular mass index. sST2 was in HFpEF associated with the composite endpoint of death or HF hospitalization, adjusted hazard ratio (HR) per log increase in sST2 6.62, 95% confidence interval (CI) 1.04–42.28, p = .046, and in HFrEF death, heart transplant or left ventricular assist systems; 3.51, 95% CI 1.05–11.69, p = .041. Conclusions. In patients with HFpEF compared to HFrEF, crude levels of sST2 were lower but potentially more strongly associated with outcomes. The lower levels of sST2 in HFpEF than in HFrEF may reflect lower degrees of fibrosis, but the potentially stronger association with outcomes may reflect a greater prognostic importance of progressive fibrosis and as such a greater potential for intervention. In conclusion; this study adds to the evidence of sST2 as prognostic marker in both HFpEF and HFrEF.
Trial registration:
Introduction
Heart failure with preserved ejection fraction (HFpEF) represents half of all heart failure (HF) and is associated with mortality and morbidity as high as in HF with reduced EF (HFrEF) [1]. There is no evidence based treatment for HFpEF and there is an increasing interest in identifying biomarkers reflecting the multiple mechanisms implicated in the pathogenesis of the syndrome [2], both as a diagnostic tool and to identify targets for future therapy.
A new paradigm has recently been proposed suggesting non-cardiac comorbidities as drivers of HFpEF inducing coronary microvascular endothelial inflammation causing decreased nitric oxide bioavailability leading to interstitial fibrosis, and stiffening of cardiomyocytes, highlighting the importance of inflammatory biomarkers [3,4].
Suppression of tumorigenicity 2 (ST2) is a member of the interleukin (IL) 1 receptor family and was first described in 1989 in inflammatory and autoimmune disease [5,6]. ST2 has two main isoforms: a transmembrane receptor (ST2L) and a soluble receptor that can be detected in plasma (sST2) [3]. The ligand for ST2, Interleukin 33 (IL-33) [7], is secreted by many cells in response to damage [8] and is part of a cardioprotective signaling system. When bound to ST2L, IL-33 exerts anti-hypertrophic and anti-fibrotic effects while the presence of sST2 suppresses these beneficial effects by neutralizing the beneficial activity of circulating IL-33 and thus leading to fibrosis [9,10]. Interestingly, ST2 is markedly induced in mechanically overloaded cardiac myocytes and reflects myocardial stress, ventricular remodeling and fibrosis [11].
IL-33/ST2 is well known to be involved in immune responses and elevated plasma levels of sST2 have been previously reported in asthma, septic shock, trauma and systemic lupus erythematosus [12–14]. More recently, the IL-33/ST2 pathway has emerged as a novel area of interest in HF and may be a therapeutic target in the prevention and treatment of cardiovascular diseases including HF [15–17].
The prognostic value of sST2 has been confirmed in acute dyspnea, acute coronary syndrome, acute and chronic HFrEF but has been less studied in HFpEF, especially in patients with chronic HFpEF [18–25].
Therefore, we compared sST2 concentrations in HFpEF and HFrEF, and assessed correlations with biomarkers and echocardiographic measures, and associations with outcome.
Materials and methods
Study population
A total of 193 subjects from three different cohorts were included. Patients with HFpEF (n = 86) were obtained from the Karolinska Rennes (KaRen) study, which was a prospective, bi-national, observational, multicentre study [26]. The KaRen Biomarker Study was a pre-specified sub-study including Swedish sites only. In summary, patients presenting to hospital with symptoms and signs of acute HF, N-terminal pro-brain natriuretic peptide (NT-proBNP) >300 ng/L and left ventricular ejection fraction (LVEF) ≥45% were included between May 2007 and December 2011. The patients returned to the hospital in a stable condition 4–8 weeks after enrolment for a follow-up visit including blood samples, clinical assessment and echocardiography. The patients were followed until September 2012 when patient’s status was assessed by chart review, telephone contact or by the Swedish National Patient and Population Registers. All HF hospitalizations were adjudicated and defined according to clinical assessment by the local specialist investigator. The primary outcome was a composite of death from any cause or hospitalization for HF.
Patients with HFrEF (n = 86), LVEF < 40%, were recruited from referrals to Karolinska University Hospital for advanced assessment of HF between January 2009 and September 2014. Blood samples were collected and clinical assessment including echocardiography was performed at enrolment and outcome data were obtained from chart review and the Swedish National Patient and Population Registers. The primary composite endpoint in this group was death from any cause, implantation of left ventricular assist systems (LVAS) or heart transplantation (HTx).
A control group of 21 healthy individuals (9 males, 12 females), above the age of 50 years, free from cardiovascular disease, hypertension, and a systolic blood pressure ≤140 mmHg were included.
The n = 21 controls was determined based on previous studies from our group, where we have detected statistically significant differences from our HFpEF and HFrEF cohorts [27,28].
Echocardiography
Echocardiographic examinations in HFpEF were performed according to a checklist using the same machine (ViVid Seven, GE Healthcare, Horten, Norway). In patients with atrial fibrillation, at least 10 beats were required for each recording.
Left ventricular mass, dimensions, volumes, and function were assessed. Assessment of diastolic function was performed based on the following variables: deceleration time (DT); mitral valve A-wave duration (MV A-dur); pulmonary vein A-wave duration (PV A-dur). Right ventricular (RV) function was also explored by tricuspid annulus excursion (TAPSE), pulse Doppler tissue imaging (DTI), estimated pulmonary pressures, inferior vena cava dimensions, and pulmonary pre-ejection time delay.
On the other hand, echocardiographic data in HFrEF were based upon echocardiographic examinations performed in our hospital with no specific study protocol (Only clinical protocol).
Biomarkers and laboratory analysis
All participants provided a morning, fasting blood sample that was collected in EDTA tubes for measurement of sST2 and NT-proBNP. sST2 was analyzed by using Presage® ST2 Assay kit available at Karolinska University laboratory (Critical Diagnostics, CA, USA) that quantitatively measures sST2 by enzyme-linked immunosorbant assay (ELISA) in a microtiter plate format Presage® ST2. NT-proBNP was analysed by fully automated quantitative STAT assay using Cobas® (Roche Diagnostics, Bromma, Sweden). The total coefficient of variation (CV) for sST2 was 22.9% and 19.4% at a mean sST2 of 28.7 ± 6.6 µg/L and 54.1 ± 10.5 µg/L respectively. CV for NT-proBNP was 6.2% and 6.5% at a mean NT-proBNP of 236 ± 15 ng/L and 2192 ± 142 ng/L respectively.The glomerular filtration rate (eGFR) was estimated using the CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration) [29].
Ethics
The study complies with the Declaration of Helsinki and was approved by the regional ethical review board. Written informed consent was obtained from all participants.
Statistical analysis
Baseline continuous variables are presented as median and interquartile range (IQR) and compared using Mann-Whitney and Kruskal-Wallis test while categorical variables are presented as numbers and percentages and compared using Fisher’s exact test.
Spearman’s correlations were assessed between sST2, NTproBNP, and eGFR in HFpEF and HFrEF as were echocardiographic parameters of systolic function (LVEF %) and diastolic function/structural heart disease [Left atrial volume index (LAVI), left ventricular (LV) transmitral early diastolic filling velocity/LV early diastolic myocardial velocity (E/E)’ and LV mass index (LVMI)].
The associations between sST2 and the composite outcomes in HFpEF and HFrEF respectively, were analyzed by Cox proportional hazards models and presented as hazard ratio (HR) and 95% confidence interval (CI) per log increase in sST2, crude and adjusted for age, sex, and New York Heart Association classification (NYHA class).
The post-hoc pair-wise comparisons are for descriptive purposes only; no adjustment for multiple testing was performed
All p-values were 2-sided and statistical significance was set to .05. Statistical analysis was performed using SPSS version 23.0 (SPSS Inc., Chicago, Ill. USA).
Results
Baseline characteristics of the three groups are shown in Table 1. Patients with HFpEF were older, more often female, had lower NYHA class, a higher mean arterial blood pressure, and body mass index (BMI) compared to HFrEF. Patients with HFrEF had a higher prevalence of ischemic heart diseases and more often a previous percutaneous coronary intervention (PCI) or coronary artery bypass grafting (CABG), as well as a worse renal function. Prevalence of co-morbidities such as atrial fibrillation and diabetes was similar between HFpEF and HFrEF. In HFpEF, LAVI was median (IQR), 43 (37–53) mL/m2 and LVMI was 115 (95–143) g/m2. In HFrEF, LAVI was median (IQR), 48 (40–73) mL/m2 and LVMI was 146 (127–184) g/m2.
Table 1. Baseline characteristics; median and lower and upper quartiles (Q1; Q3) and n (%).
As depicted in Figure 1 crude sST2 levels were higher in HFrEF compared to HFpEF and controls, median (IQR), 35 (23–52), 23 (17–31) and 25 (21–32) µg/L respectively (overall p < .001). Levels of sST2 increased with worsening HF severity assessed as NYHA class in both LVEF categories (Figure 2). The difference between the HF groups persisted after adjustment for age and sex (p < .001) but not after adding NYHA class to the model to adjust for heart failure severity, p = .4.
Published online:
08 March 2019Figure 1. Crude sST2 concentrations in HFpEF, HFrEF and controls, where p values represent the significant difference in sST2 between HFpEF, HFrEF and controls. Boxplots displaying interquartile range (IQR) and median (-), whiskers indicating variability outside the upper and lower quartiles.
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/icdv20/2019/icdv20.v053.i01/14017431.2019.1583363/20200214/images/medium/icdv_a_1583363_f0001_c.jpg"})
Figure 1. Crude sST2 concentrations in HFpEF, HFrEF and controls, where p values represent the significant difference in sST2 between HFpEF, HFrEF and controls. Boxplots displaying interquartile range (IQR) and median (-), whiskers indicating variability outside the upper and lower quartiles.
Published online:
08 March 2019Figure 2. sST2 concentrations by NYHA class in HFpEF and HFrEF, p-values for difference in sST2 between NYHA class in HFpEF (I vs. III) and HFrEF (II vs. IV) Boxplots displaying interquartile range (IQR) and median (-), whiskers) indicating variability outside the upper and lower quartiles.
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/icdv20/2019/icdv20.v053.i01/14017431.2019.1583363/20200214/images/medium/icdv_a_1583363_f0002_c.jpg"})
Figure 2. sST2 concentrations by NYHA class in HFpEF and HFrEF, p-values for difference in sST2 between NYHA class in HFpEF (I vs. III) and HFrEF (II vs. IV) Boxplots displaying interquartile range (IQR) and median (-), whiskers) indicating variability outside the upper and lower quartiles.
Table 2 shows the correlations between sST2 and demographic/echocardiographic data in HFpEF, HFrEF, and controls.
Table 2. Correlations between sST2 and variables in HFpEF, HFrEF and controls.
Association with outcome in HFpEF and HFrEF
In the HFpEF group, 11 patients (13%) died and the composite outcome of all-cause death or HF hospitalization occurred in 36 patients (42%) during a median follow-up time of (IQR) 522 (232–1089) days. In the HFrEF group, 28 patients (33%) died and the composite outcome of death from any cause, implantation of LVAS or HTx occurred in 56 patients (65%) during a median follow-up time of 204 (55–421) days.
In both HFpEF and HFrEF (Figure 3(a,b)), sST2 was significantly associated with the composite outcome in crude analyses (HR per log increase 10.04 [95% CI 1.89–53.44], p = .007) and (HR 3.28 [95% CI 1.06–10.16], p = .039) respectively. The association persisted after adjustment for age, sex, and NYHA class, in HFpEF, (HR 6.62 [95% CI 1.04–42.28], p = .046) and in HFrEF, (HR 3.51 [95% CI 1.05–11.69], p = .041). In HFpEF, sST2 was numerically associated also with mortality alone both in crude analyses (HR per log increase 12.39 [95% CI 0.70–218.55], p = .086) and after adjustment for age, sex, and NYHA class (HR per log increase 7.32 [95% CI 0.35–154.27], p = .200), but due to the small number of deaths in HFpEF, this association was not statistically significant.
Published online:
08 March 2019Figure 3. a and b Kaplan-Meier estimates of survival free from HF hospitalization in HFpEF and survival free from heart transplant or left ventricular assist system (LVAS) in HFrEF comparing sST2 above and below median in HFpEF and HFrEF respectively. Association with the composite endpoint in HFpEF and HFrEF analyzed with Cox Regression, HR per log increase in sST2, 95% conficence interval, and p-value depicted in each graph.
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/icdv20/2019/icdv20.v053.i01/14017431.2019.1583363/20200214/images/medium/icdv_a_1583363_f0003_c.jpg"})
Figure 3. a and b Kaplan-Meier estimates of survival free from HF hospitalization in HFpEF and survival free from heart transplant or left ventricular assist system (LVAS) in HFrEF comparing sST2 above and below median in HFpEF and HFrEF respectively. Association with the composite endpoint in HFpEF and HFrEF analyzed with Cox Regression, HR per log increase in sST2, 95% conficence interval, and p-value depicted in each graph.
Discussion
We demonstrate lower crude levels of sST2 in patients with HFpEF compared to HFrEF but no difference between the HF groups when adjusted for NYHA class. Notably, there are common genetic factors that may account for up to 40% interindividual variability in sST2 levels (34) and these factors together with different pathophysiological pathways in HFpEF and HFrEF may to some extent explain the diversity of sST2 levels in HFpEF and HFrEF (35). Concentrations of sST2 in HFpEF were associated with left atrial volumes, and in both HFpEF and HFrEF, with HF severity measured by NT-proBNP and NYHA class. sST2 was potentially more strongly associated with outcomes in HFpEF compared to in HFrEF.
Serum levels of sST2 in patients with HFpEF in this analysis were lower than in other studies including patients with HFpEF, where the median values were 25–30 µg/L [30,31] as compared to 23 µg/L in the present study. The current approved partition value to predict morbidity and mortality is 35 µg/L, however this is mainly established from HFrEF studies or studies in which HFrEF patients were dominant [32,33].
The involvement of sST2 in the fibrotic response to tissue injury has been discussed in previous studies where early experimental animal data suggested sST2 resides in the extracellular matrix of integumental tissue which may mean involvement in the growth and homeostatis of matrix [34]. Moreover, the expression levels of sST2 in cardiac fibroblasts and cardiomyocytes are increased in response to stimuli of cardiac fibrosis: biomechanical strain and Angiotensin II [11]. Furthermore, there was a transiently increased ST2 expression in animal models (after coronary artery ligation) [35].
Correlation with diastolic function/structural heart disease
Our results demonstrate that in HFpEF sST2 levels are weakly associated with LAVI but not with LV filling pressure measured as E/E’ or with LVMI. Our findings are inconsistent with previous studies that suggested an association with E/E’, albeit weak [36,37]. However, another recent study failed to demonstrate any association between ST2 levels and left ventricular diastolic or systolic function parameters or LV geometry as assessed by echocardiography [21]. LAVI is an important prognostic predictor in all HF, independent of LVEF, and in HFpEF, prognosis is associated with the degree of atrial dilation [38,39]. However; In the PARAMOUNT trial, LA strain was decreased independent of LA size or history of AF, suggesting that LA fibrosis is not solely or even primarily determined by the degree of LA enlargement and that LA dysfunction may be a marker of severity and play a pathophysiologic role in HFpEF [40].
Correlation with systolic function/structural heart disease
There was no association between sST2 and LVEF in neither HFpEF nor HFrEF. These findings are consistent with the results from the study that included Framingham offspring participants where sST2 was neither associated with LV hypertrophy nor abnormal LVEF [41] but contrasting to other studies suggesting an association between sST2 and LV function after acute myocardial infarction [42] and in patients with acute dyspnea [25]. In HFpEF patients, sST2 has recently been linked to LV functional impairment assessed as global longitudinal strain [43].
Association with outcomes
This study adds to the evidence of sST2 as an independent prognostic marker not only in HFrEF but also in HFpEF. These data in chronic HFpEF are consistent with previous studies showing that sST2 is a prognostic marker in acute HFpEF [36]and in acute HFrEF [44,45] as well as in chronic HFrEF [32,46]. Of particular importance is the concept of HFpEF as a gradual and progressive syndrome, where earlier intervention may be beneficial. This is suggested by two studies where treatment was more effective in patients with lower NT-proBNP [47,48]. Analogously, the lower levels of sST2 in HFpEF than in HFrEF may reflect lower degrees of fibrosis, but the stronger association with outcomes may reflect a greater prognostic importance of progressive fibrosis and as such a greater potential for intervention
Limitations
The small sample size of three groups especially controls was one of the limitations in this study. Moreover, the 3 cohorts were recruited in different settings and there was no specific echocardiography protocol in HFrEF patients (the needed data were available from clinical echocardiographic examinations). Additionally, as a reflection of the different heart failure phenotypes, the outcome definitions were different in HFpEF and HFrEF. Some patients with HFrEF went on to LVAS or HTx, reflecting deterioration; wherefore these were included in the composite endpoint for HFrEF. Hospitalizations were not included in the composite endpoint for HFrEF because HFrEF patients were more severely ill and frequently hospitalized during follow-up, in part due to deterioration but also for further investigations to examine eligibility for LVAS/HTx, making interpretation of this outcome difficult. Including LVAS and HTx as part of a composite endpoint is standard practice in HFrEF. The alternative, to censor patient at LVAD and HTx would underestimate risk in HFrEF, since censoring precludes subsequent competing events. Furthermore, the different pathophysiological pathways in HFpEF and HFrEF mean difficulty in comparing the two groups of patients with different age and sex distribution and comorbidity profile, although we tried to correct for clinical differences. The higher hazard ratio in HFpEF than in HFrEF is suggestive of a potentially stronger prognostic role, but by no means conclusive. The modest sample size precluded more direct comparisons of discrimination, calibration, and e.g. net reclassification. Finally, it remains unclear whether the association between sST2 concentrations and the prognosis of patients with HF reflects what is happening at a cardiac level, or whether the sST2 concentrations reflect other pathologic processes, such as pulmonary disease.
Conclusion
In patients with HFpEF compared to HFrEF, levels of sST2 were lower but potentially more strongly associated with outcomes. This may have implications for therapeutic targeting of the appropriate phase of the progressive HFpEF syndrome.
| Demographics | HFpEF n = 86 | HFrEF n = 86 | Control n = 21 | p-value | |||
|---|---|---|---|---|---|---|---|
| HFpEF: HFrEF | HFpEF: control | HFrEF: control | Overall | ||||
| Age years | 73 (67; 80) | 63 (52; 68) | 67 (59; 70) | <.001 | <.001 | .159 | <.001 |
| Gender male/female | 42/44 (49/51) | 70/16 (81/19) | 9/12 (43/57) | <.001 | .624 | <.001 | <.001 |
| Medical history | |||||||
| Atrial fibrillation/flutter | 49 (57) | 45 (52) | .541 | ||||
| Diabetes mellitus | 28 (33) | 25 (29) | .621 | ||||
| Smoking | 7 (8) | 8 (9) | .101 | ||||
| COPD | 17 (20) | 10 (12) | .143 | ||||
| CABG | 11 (13) | 21 (24) | .051 | ||||
| PCI | 9 (11) | 20 (23) | .026 | ||||
| NYHA I | 19(22) | 1 (1) | <.001 | ||||
| NYHA II | 47(55) | 4 (5) | |||||
| NYHA III | 20 (23) | 67 (80) | |||||
| NYHA IV | 0 | 14 (11) | |||||
| Clinical Measurements | |||||||
| BMI kg/m2 | 29 (25;34) | 27 (23;30) | 25 (22;26) | .001 | <.001 | .043 | <.001 |
| SBP mmHg | 140 (129;151) | 105 (96;121) | 125 (120;140) | <.001 | .001 | <.001 | <.001 |
| MAP mmHg | 100 (92;107) | 82 (74;93) | 94 (89;103) | <.001 | .111 | <.001 | <.001 |
| HR beats/min | 70 (60;80) | 70 (60;75) | .482 | ||||
| LVEF % | 64 (58;68) | 21 (15;28) | <.001 | ||||
| LVEDD mm | 47 (43;53) | 66 (61;75) | <.001 | ||||
| Treatment | |||||||
| ARB/ACE-I | 67 (78) | 77 (90) | .037 | ||||
| Beta blocker | 69 (80) | 85 (99) | <.001 | ||||
| MRA | 18 (21) | 58 (67) | <.001 | ||||
| Loop diuretic | 61 (71) | 75 (87) | .009 | ||||
| Digoxin | 10 (12) | 21 (24) | .030 | ||||
| Statin | 38 (44) | 38 (44) | 1.000 | ||||
| CRT | 0 | 46 (54) | <.001 | ||||
| ICD | 0 | 70 (81) | <.001 | ||||
| Laboratory | |||||||
| ST2 (µg/L) | 23 (17; 31) | 35 (23; 52) | 25 (21; 32) | <.001 | .145 | .029 | <.001 |
| NTproBNP (ng/L) | 1000 (465;2335) | 3290 (1405;6115) | 67 (31;110) | <.001 | <.001 | <.001 | <.001 |
| eGFR mL/min/1.73m2 | 68(50;81) | 58 (42;73) | 82 (71;91) | .036 | .006 | <.001 | <.001 |
| Sodium (mmol/L) | 141 (140;143) | 138 (136; 140) | 140 (138; 141) | <.001 | .010 | .022 | <.001 |
| Hemoglobin (g/L) | 13.1 (12.2;14.2) | 13.3 (12.2;14.4) | 14.0 (13.4;15.1) | .618 | .004 | .033 | .025 |
HFpEF: heart failure with preserved ejection fraction; HFrEF: heart failure with reduced ejection fraction; COPD: chronic obstructive pulmonary disease; CABG: coronary artery bypass grafting; PCI: percutaneus coronary intervention; NYHA: New York Heart Association; BMI: body mass index; SBP: systolic blood pressure; MAP: mean arterial pressure; HR: heart rate; LVEF: left ventricular ejection fraction; LVEDD: left ventricular end diastolic diameter; ARB: angiotensin receptor blocker; ACE-I, angiotensin converter enzyme inhibitor; MRA; mineralocorticoid receptor antagonist; CRT: cardiac resynchronization therapy; ICD: implantable cardioverter defibrillator; eGFR: estimated glomerular filtration rate.
| sST2 | |||||||
|---|---|---|---|---|---|---|---|
| HFpEF | HFrEF | Controls | |||||
| Variable | rs | p-value | rs | p-value | rs | p-value | |
| NT-proBNP | 0.392 | <.001 | 0.466 | <.001 | −0.020 | .931 | |
| Age | 0.116 | .295 | 0.043 | .698 | 0.215 | .350 | |
| NYHA class | 0.307 | .005 | 0.270 | .014 | |||
| eGFR | −0.073 | .513 | −0.224 | .044 | −0.141 | .542 | |
| MAP | 0.047 | .668 | −0.015 | .901 | −0.350 | .120 | |
| BMI | 0.116 | .301 | −0.035 | .755 | −0.115 | .501 | |
| E/e’ | 0.148 | .248 | −0.096 | .523 | |||
| LAVI | 0.276 | .019 | 0.247 | .224 | |||
| LVMI | −0.005 | .795 | 0.622 | .018 | |||
| LVEF | −0.057 | .634 | 0.005 | .962 | |||
HFpEF: heart failure with preserved ejection fraction; HFrEF: heart failure with reduced ejection fraction; eGFR: estimated glomerular filtration rate; MAP: mean arterial pressure; BMI: body mass index; E/E’: left ventricular transmitral early diastolic filling velocity/left ventricular early diastolic myocardial velocity; LAVI: Left atrial volume index; LVMI: Left ventricular mass index; LVEF: left ventricular ejection fraction.
Acknowledgments
The authors are grateful to Gunilla Förstedt at Karolinska University hospital for blood sampling and patient care, Kambiz Shahgaldi and Maria Westerlind for echocardiogram assessments.
Disclosure statement
LHL: No disclosures directly related to the present work. Unrelated disclosures are: Research grants from Astra Zeneca, Boston Scientific; consulting or speaker’s honoraria from Novartis, Astra Zeneca, Bayer, St Jude, Medtronic, Vifor Pharma. The other authors have no relationships that could be construed as a conflict of interest. CH: consulting fees from Novartis and speaker and honoraria from MSD. Authors take responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation.