Structured Abstract
Objectives
To investigate relationships between insulin-like growth factor-binding protein 7 (IGFBP7) and parameters of diastolic function or functional capacity in patients with heart failure and preserved ejection fraction (HFpEF), randomized to receive sildenafil or placebo.
Background
IGFBP7 was previously found to be associated with diastolic function in HF with reduced EF, but it is unclear whether these associations are present in HFpEF.
Methods
At baseline and 24 weeks, IGFBP7, imaging studies, and peak oxygen consumption (VO2max) were obtained and compared in 160 HFpEF patients randomized to receive sildenafil or placebo.
Results
Patients with supramedian baseline IGFBP7 concentrations were older, had signs of systemic congestion and worse renal function, and had higher concentrations of prognostic HF biomarkers including amino-terminal pro-B-type natriuretic peptide (P<0.05). Higher baseline IGFBP7 was modestly correlated with worse diastolic function: higher E velocity (ρ=0.40), E/E’ (ρ=0.40), left atrial volume index (ρ=0.39), and estimated right ventricular systolic pressure (RVSP; ρ=0.41; all P<0.001) and weakly correlated with transmitral E/A (ρ=0.26, P=0.006). Notably, change (Δ) in IGFBP7 was significantly correlated with change in E, E/A, E/E’, and RVSP. Elevated baseline IGFBP7 was associated with lower baseline VO2max (13.2 versus 11.1 mL/min/kg; P<0.001) and ΔIGFBP7 was weakly inversely correlated with ΔVO2max (ρ=−0.19, P=0.01). Subjects receiving sildenafil had a decrease in IGFBP7 over 24 weeks, in contrast to placebo-treated patients (median ΔIGFBP7 −1.5 versus +13.6 ng/mL; P<0.001).
Conclusions
In patients with HFpEF, IGFBP7 may be a novel biomarker of diastolic function and exercise capacity.
Keywords: Heart failure with preserved ejection fraction, Insulin-like growth factor-binding protein 7, diastolic function, biomarkers
Heart failure with preserved ejection fraction (HFpEF) affects approximately half of patients with the clinical syndrome of HF, and it is associated with considerable morbidity and mortality (1,2). A pivotal abnormality in HFpEF is the presence of impaired myocardial diastolic function; echocardiographic parameters to assess such impaired function can be complicated to accurately obtain and interpret (3). This is further compounded by the fact the clinical diagnosis of HFpEF may be challenging to recognize and manage. Taken together, these may be potential reasons why therapeutic trials have been disappointing for this common type of HF.
The use of cardiac biomarkers in HF is a rapidly growing area of interest (4), and may provide options to assist in the care of patients with HFpEF through better phenotyping of the syndrome. B-type natriuretic peptide (BNP) and its amino-terminal precursor (NT-proBNP) have been examined for both diagnosis and risk stratification in HFpEF (5,6) and their concentrations are associated with echocardiographic parameters used to define diastolic dysfunction (7,8). However, as the prime trigger for NP release is myocardial diastolic wall stress, they may be nonspecific for diastolic function; indeed, many other echocardiographic abnormalities such as left ventricular (LV) EF, chamber size, and valvular regurgitation contribute to circulating concentrations of BNP or NT-proBNP (9). Other biomarkers have been therefore explored for profiling of HFpEF (10–13) but remain non-specific for the diagnosis, similar to the NPs.
Recently, insulin-like growth factor-binding protein 7 (IGFBP7) has been identified as a biomarker associated with cardiac hypertrophy and HF through systematic proteomic candidate searches (14). IGFBP7 concentrations were recently found to be prognostic for prediction of worsening HF, hospitalization for HF, and cardiovascular death in a well-phenotyped cohort of patients with HF and reduced EF (HFrEF) (15); while IGFBP7 was not associated with echocardiographic parameters of systolic function or remodeling, it was significantly correlated with multiple parameters of diastolic function (16). Diastolic abnormalities are highly prevalent in patients with HFrEF (9), but it is unclear whether IGFBP7 would demonstrate a similar relationship in a population of patients with HFpEF, a group of patients in whom diastolic abnormalities are even more important. We therefore sought to examine baseline concentrations and change (Δ) over time in IGFBP7 in a well-characterized group of patients with HFpEF from the Phosphodiesterase-5 Inhibition to Improve Clinical Status and Exercise Capacity in HFpEF (RELAX) study (17). We also investigated associations between IGFBP7 and peak oxygen consumption (VO2max), as well as treatment with sildenafil. We hypothesized that baseline and Δ in IGFBP7 would be significantly associated and correlated with imaging parameters of diastolic dysfunction and VO2max in patients with HFpEF.
Methods
Patient population and study overview
The RELAX study evaluated the effect of sildenafil on exercise capacity and clinical status in 216 patients with EF ≥ 50%, stable HF symptoms with objective evidence of HF (defined as previous HF hospitalization, acute HF therapy with administration of intravenous diuretic therapy, chronic loop diuretic therapy with left atrial enlargement, or invasively documented elevated LV filling pressures), and VO2max of ≤60% of age and sex predicted value (with respiratory exchange ratio of ≥ 1.0) as well as either NT-proBNP ≥ 400 pg/mL or elevated LV filling pressures at the time of NT-proBNP measurement of < 400 pg/mL (17). The primary endpoint was change in VO2max at 24 weeks; secondary endpoints included change (Δ) in 6 minute walk distance, Δ in clinical status score based on time to death, time to cardiovascular or cardiorenal hospitalization, and change in quality of life (for those without hospitalization) at 24 weeks (17). The RELAX trial was funded by the National Heart, Lung, and Blood Institute and was conducted by the Heart Failure Network. All patients provided written informed consent and that trial was approved by the institutional review board at each site (17). The current analysis examines IGFBP7 in 160 patients from RELAX with available blood samples. Echocardiographic measurements, MRI data, and VO2max were compared at baseline and 24 weeks in patients receiving placebo and sildenafil in order to characterize the mechanistic links between IGFBP7 and diastology as well as evaluate the effects of sildenafil therapy on IGFBP7 in patients with HFpEF.
Laboratory evaluation and Imaging studies
Standard biomarker measurements were performed in a blinded fashion as previously reported by the Heart Failure Network biomarker core laboratory (University of Vermont; Burlington, VT) (18). IGFBP7 was measured using an Elecsys assay (Roche Diagnostics, Penzburg, Germany). The limit of detection for the IGFBP7 assay was 0.01 ng/mL. The inter-run and intra-run coefficients of variation were ≤5% and ≤2%.
Blinded core laboratories assessed data from echocardiography (Mayo Clinic Rochester), MRI (Duke University) and cardiopulmonary exercise testing (CPET; Massachusetts General Hospital). Diastolic function parameters assessed included mitral inflow peak early filling velocity (E) and late diastolic filling velocity (A) as well as their ratio (E/A), tissue Doppler early diastolic mitral annular velocity (E’), and estimation of filling pressures using the ratio of E/E’. Additionally, we examined left atrial volume index (LAVi), estimated right ventricular systolic pressure (RVSP), and left ventricular hypertrophy (LVH) as these are considered important echocardiographic measures of diastolic dysfunction (3).
Statistical Analysis
A total of 161 patients were included in the analysis, but one subject was missing baseline IGFBP7 concentration and another was missing the 24 week IGFBP7 concentration; 159 have data for change in IGFBP7 concentrations. Subjects were initially divided into groups using the median baseline IGFBP7 concentration of 219 ng/mL, and baseline characteristics were compared using the Wilcoxon rank sum test for continuous variables and Χ2 test for categorical variables. Continuous variables are displayed as medians and 25th, 75th percentiles (IQR). IGFBP7 was compared to other established and emerging biomarkers using bivariate correlations. Baseline imaging parameters were compared between subjects with inframedian IGFBP7 concentrations and compared to those with supramedian concentrations using the Wilcoxon rank sum test. To further investigate the independent predictive value of IGFBP7, logistic regression was performed with IGFBP7 as the independent variable and parameters of diastolic dysfunction as the dependent variables, adjusting for age, renal function, and NT-proBNP. Receiver operating curves were also constructed for both IGFBP7 and NT-proBNP to identify appropriate cutpoints with parameters of diastolic dysfunction. Relationships between IGFBP7 and imaging parameters are presented as Spearman correlations for both baseline and 24 week change measures. Δ was defined as final minus baseline value for each parameter. Rho values, adjusted for baseline variables, are presented for the relationship between the week 24 difference in log-transformed IGFBP7 and the change in imaging variable. Baseline and ΔVO2max were examined in a similar fashion; the relationship between ΔVO2max and ΔIGFBP7 is shown with a scatterplot. Analyses of echocardiographic and MRI parameters of diastolic function were repeated using median splits of baseline NT-proBNP and galectin-3, as well as ΔNT-proBNP and Δgalectin-3 at 24 weeks.
To evaluate the effects of sildenafil therapy on IGFBP7 concentrations, median baseline, 24 week, and ΔIGFBP7 concentrations were compared between those receiving sildenafil and those receiving placebo using the Wilcoxon-rank sum test. An adjusted P value for treatment effect comes from a linear regression model that adjusts for log-transformed baseline IGFBP7 concentration.
Data were analyzed with SAS version 9.4 (Cary, NC). All P values are two-sided with results ≤0.05 considered significant.
Results
Baseline IGFBP7 concentrations and baseline characteristics
The median IGFBP7 concentration was 219 ng/mL. Supplemental Table 1 portrays baseline characteristics of the subjects included in the current analysis with those who did not have IGFBP7 concentrations available. Compared to the subjects included in the current analysis, those without baseline IGFBP7 concentrations appeared to have characteristics of more advanced HF. Table 1 portrays baseline characteristics of the subjects in the current analysis divided by the median IGFBP7 concentration. Patients with supramedian IGFBP7 concentrations were older, more likely to have atrial fibrillation or flutter as well as signs of congestion, and they were more likely to be taking a diuretic. Additionally, these patients had significantly elevated cystatin C, NT-proBNP, pro-collagen type III amino-terminal peptide (PIIINP), highly sensitive troponin I (hsTnI), and galectin-3 concentrations as well as reduced estimated glomerular filtration rate (eGFR) when compared to those with inframedian IGFBP7.
Table 1.
Baseline characteristics divided by median baseline IGFBP7 concentration.
| Characteristic | N | IGFBP7 < 219 ng/mL | N | IGFBP7 > 219 ng/mL | P value |
|---|---|---|---|---|---|
| Clinical data | |||||
| Age (years), median (IQR) | 80 | 67 (60.5, 73) | 80 | 69 (65, 78) | 0.02 |
| Male sex (%) | 80 | 34 (42.5) | 80 | 45 (56.3) | 0.08 |
| Self-reported race white (%) | 80 | 76 (95.0) | 80 | 72 (90.0) | 0.23 |
| Ischemic heart disease (%) | 80 | 26 (32.5) | 80 | 34 (42.5) | 0.19 |
| Hypertension (%) | 80 | 65 (81.3) | 80 | 69 (86.3) | 0.39 |
| Presence of atrial fibrillation / flutter (%) | 80 | 30 (37.5) | 80 | 49 (61.3) | 0.003 |
| Chronic obstructive pulmonary disease (%) | 80 | 13 (16.3) | 80 | 16 (20.0) | 0.54 |
| Diabetes Mellitus (%) | 80 | 27 (33.8) | 80 | 38 (47.5) | 0.08 |
| Hospitalization for heart failure in last year (%) | 80 | 22 (27.5) | 80 | 29 (36.3) | 0.24 |
| Calculated definition of anemia (%) | 80 | 15 (18.8) | 80 | 34 (42.5) | 0.001 |
| Physical exam | |||||
| Body Mass Index (kg/m2), median(IQR) | 80 | 32.8 (29.0, 37.7) | 80 | 32.9 (28.1, 39.5) | 0.85 |
| Systolic Blood Pressure (mmHg), median(IQR) | 80 | 126 (113, 138) | 80 | 124 (113, 135) | 0.82 |
| Diastolic Blood Pressure (mmHg), median(IQR) | 80 | 70 (64, 80) | 80 | 69.5 (61, 78) | 0.26 |
| Heart rate (bpm), median(IQR) | 80 | 68.5 (62, 78.5) | 80 | 68 (60, 79) | 0.62 |
| Jugular venous pressure ≥8 cm H20 (%) | 77 | 21 (27.3) | 77 | 42 (54.5) | <0.001 |
| Rales (%) | 80 | 4 (5.0) | 80 | 6 (7.5) | 0.51 |
| Peripheral edema (%) | 80 | 10 (12.5) | 80 | 22 (27.5) | 0.02 |
| NYHA classification >II (%) | 80 | 37 (46.3) | 80 | 46 (57.5) | 0.15 |
| Orthopnea (%) | 73 | 42 (57.5) | 76 | 46 (60.5) | 0.71 |
| ACE inhibitor or angiotensin receptor blocker (%) | 80 | 58 (72.5) | 80 | 52 (65.0) | 0.31 |
| Beta blocker (%) | 80 | 58 (72.5) | 80 | 61 (76.3) | 0.59 |
| Aldosterone antagonist (%) | 80 | 7 (8.8) | 80 | 11 (13.8) | 0.32 |
| Calcium channel blocker (%) | 80 | 18 (22.5) | 80 | 30 (37.5) | 0.04 |
| Statin (%) | 80 | 53 (66.3) | 80 | 50 (62.5) | 0.62 |
| Daily diuretic (%) | 80 | 56 (70.0) | 80 | 77 (96.3) | <0.0001 |
| Functional data | |||||
| MLWHF Total score, median(IQR) | 78 | 44 (31, 65) | 75 | 42 (26, 59) | 0.28 |
| MLWHF Physical Dimension score, median(IQR) | 80 | 24 (18.5, 30.5) | 78 | 21 (13, 29) | 0.08 |
| Baseline walk distance (meters), median(IQR) | 80 | 357 (288, 401) | 80 | 302 (239, 359) | 0.002 |
| Core labs | |||||
| Creatinine (mg/dL)†, median(IQR) | 80 | 0.93 (0.75, 1.13) | 78 | 1.22 (1.01, 1.54) | <0.0001 |
| Cystatin C (mg/L)†, median(IQR) | 80 | 1.09 (0.91, 1.25) | 80 | 1.59 (1.19, 1.93) | <0.0001 |
| eGFR (mL/min/1.73m2), median(IQR) | 80 | 75.8 (61.1, 88.9) | 78 | 55.5 (42.7, 68.6) | <0.0001 |
| NT- proBNP (pg/mL)†, median(IQR) | 80 | 291 (90, 620) | 79 | 1174 (620, 1919) | <0.0001 |
| Pro-collagen III NTP (ug/L)†, median(IQR) | 80 | 6.78 (5.17, 7.99) | 80 | 8.7 (6.5, 11.6) | <0.0001 |
| High sensitivity troponin I (pg/mL)†, median(IQR) | 80 | 5.9 (3.5, 11.4) | 78 | 11.2 (7.8, 21.9) | <0.0001 |
| Galectin 3 (ng/mL), median(IQR) | 80 | 12.9 (10.3, 15.9) | 75 | 15.2 (11.7, 20.4) | 0.002 |
IQR= interquartile range, bpm=beats per minute, NYHA=New York heart association, ACE=angiotensin converting enzyme, MLWHF=Minnesota Living with Heart Failure, eGFR= estimated glomerular filtration rate, NT-proBNP=amino-terminal pro-B-type natriuretic peptide, NTP= amino-terminal peptide
Table 2 shows baseline imaging and CPET results. In the current analysis, 159 patients underwent echocardiography, 88 obtained an MRI, and 160 underwent CPET at baseline. Baseline echocardiography revealed modest, but significant associations and correlations between baseline IGFBP7 and transmitral E velocity (ρ=0.40; P<0.0001), E/E’ (ρ=0.40; P<0.0001), LAVi (ρ=0.39; P<0.0001), and estimated RVSP (ρ=0.41; P<0.0001). It was also weakly associated and correlated with the transmitral E/A ratio (ρ=0.26; P=0.006). Scatterplots of these correlations are portrayed in Figure 1. These results demonstrated stronger correlations than our previous findings in patients with HFrEF (16). Notably, there were no significant associations or correlations with measures of systolic function including LV end diastolic volume index (ρ=0.19, P=0.10), LV end systolic volume index (ρ=0.16, P=0.16), and LVEF (ρ=−0.13, P=0.09). LV hypertrophy (ρ=0.23, P=0.01) as well as LV mass index by both echo and MRI were significantly associated with IGFBP7 concentration (ρ=0.23, P=0.01 and ρ=0.22, P=0.04, respectively). MRI estimates at baseline revealed significantly reduced aortic distensability in those with supramedian IGFBP7 (ρ= −0.30, P=0.02); however, there were no significant relationships between IGFBP7 and MRI measures of systolic function. Baseline CPET demonstrated that those with supramedian IGFBP7 concentrations had a significantly reduced VO2max. Similarly, those with inframedian VO2max (≤11.7 mL/kg/min) had significantly higher baseline IGFBP7 concentrations (243 versus 197 pg/mL; P<0.0001).
Table 2.
Baseline imaging and cardiopulmonary exercise testing results, divided by median baseline IGFBP7 concentration.
| Spearman correlation | ||||||
|---|---|---|---|---|---|---|
| Characteristic | N | IGFBP7 < 219ng/mL | IGFBP7 > 219 ng/mL | Wilcoxon with rank sum baseline P value | IGFBP7 (ρ) | Spearman correlation P value |
| Baseline echocardiography | ||||||
| MV inflow: E velocity at leaf tip (m/sec), median (IQR) | 151 | 0.9 (0.7, 1.1) | 1.1 (0.9, 1.4) | <0.0001 | 0.395 | <0.0001 |
| MV inflow: A velocity at leaf tip (m/sec), median (IQR) | 112 | 0.7 (0.5, 0.9) | 0.6 (0.4, 0.9) | 0.30 | −0.066 | 0.49 |
| E/A ratio, median (IQR) | 112 | 1.18 (1.00, 1.80) | 1.80 (1.13, 3.25) | 0.01 | 0.258 | 0.006 |
| LV relaxation septal - E' (m/sec), median (IQR) | 146 | 0.06 (0.05, 0.08) | 0.06 (0.04, 0.07) | 0.047 | −0.203 | 0.01 |
| Filling pressure septal (medial) - E/E' (m/sec), median (IQR) | 143 | 13.3 (10.0, 18.2) | 18.3 (13.3, 26.7) | <0.001 | 0.398 | <0.0001 |
| LA volume index (mL/m2), median (IQR) | 113 | 40.3 (31.7, 47.3) | 53.9 (39.8, 63.1) | <0.001 | 0.386 | <0.0001 |
| RVSP (mmHg), median (IQR) | 103 | 35.7 (32.0, 43.6) | 48.4 (36.4, 56.0) | <0.001 | 0.413 | <0.0001 |
| LV mass index (g/m2), median (IQR) | 117 | 65.6 (56.5, 83.3) | 80 (67, 103) | 0.006 | 0.227 | 0.01 |
| LV hypertrophy, N (%) | 117 | 22 (37.9) | 35 (59.3) | 0.021 | 0.234 | 0.01 |
| LV end diastolic volume index (mL/m2), median (IQR) | 79 | 48.8 (42.8, 55.0) | 55.8 (45.0, 65.2) | 0.09 | 0.189 | 0.10 |
| LV end systolic volume index (mL/m2), median (IQR) | 79 | 18.4 (15.9, 22.5) | 23.0 (16.4, 30.7) | 0.13 | 0.158 | 0.16 |
| LVEF (%),median (IQR) | 159 | 60.5 (57.5, 68.0) | 60 (55, 65) | 0.24 | −0.134 | 0.09 |
| Baseline cardiac MRI | ||||||
| LV stroke volume (mL), median (IQR) | 88 | 67.2 (60.3, 82.3) | 81.2 (69.2, 94.1) | 0.05 | 0.121 | 0.27 |
| LV mass index (g/m2), median (IQR) | 88 | 59.7 (49.3, 73.0) | 63.8 (54.1, 85.5) | 0.10 | 0.218 | 0.04 |
| LV end diastolic volume index (mL/m2), median (IQR) | 88 | 50.4 (43.1, 58.4) | 60.6 (53.9, 69.3) | 0.004 | 0.225 | 0.04 |
| LV end systolic volume index (mL/m2), median (IQR) | 88 | 17.3 (14.0, 22.1) | 22.3 (18.8, 27.3) | 0.02 | 0.186 | 0.08 |
| LVEF (%), median (IQR) | 88 | 67.2 (59.6, 72.1) | 63.7 (57.3, 68.3) | 0.20 | −0.101 | 0.35 |
| Cardiac index (L/min/m2), median (IQR) | 87 | 2.23 (1.82, 2.52) | 2.4 (2.01, 2.77) | 0.17 | 0.106 | 0.33 |
| Aortic distensibility (10−3 mmHg−1), median (IQR) | 61 | 1.22 (0.74, 2.05) | 0.79 (0.55, 1.70) | 0.08 | −0.301 | 0.02 |
|
| ||||||
| CPET | ||||||
| Baseline peak VO2 (ml/min/kg), median (IQR) | 160 | 13.2 (11.1, 16.1) | 11.1 (9.9, 13.4) | <0.0001 | −0.339 | <0.0001 |
MV= mitral valve, IQR= interquartile range, LV= left ventricle, LA= left atrium, RVSP= estimated right ventricular systolic pressure, LVEF= left ventricular ejection fraction, MRI= magnetic resonance imaging, CPET= cardiopulmonary exercise testing, VO2max= peak oxygen consumption
Figure 1. Correlation between baseline IGFBP7 and baseline parameters of diastolic dysfunction.
There were significant correlations between IGFBP7 and E/A, E/E’, left ventricular (LV) mass index, E’, left atrial (LA) volume index, and estimated right ventricular systolic pressure (RVSP).
Comparison between baseline IGFBP7 and other biomarkers and clinical factors
Correlations between IGFBP7 and other established and emerging biomarkers were performed, as shown in Table 3. Many of these correlations were significant, with IGFBP7 most strongly correlated with cystatin C (ρ=0.64; P<0.0001) and the carboxy-terminal telopeptide of collagen type 1 (ρ=0.65, P<0.0001). It was also strongly correlated with NT-proBNP (ρ=0.59; P<0.0001). Given the significant correlations with parameters of renal function, we further investigated the relationship between IGFBP7 and eGFR. Baseline eGFR was not significantly associated with ΔIGFBP7. However, eGFR at 24 weeks and ΔeGFR were significantly associated with ΔIGFBP7 [β= −0.001; 95% CI (−0.002, −0.0001); P=0.02 and β= −0.003; 95% CI (−0.004, −0.0009); P=0.003, respectively].
Table 3.
Bivariate correlations of other biomarkers with IGFBP7.
| Biomarker | Spearman correlation with IGFBP7 (ρ) | Spearman correlation P value |
|---|---|---|
| Creatinine (mg/dL) | 0.494 | <0.0001 |
| eGFR (mL/min/1.73m2) | −0.471 | <0.0001 |
| Uric acid (mg/dL) | 0.356 | <0.0001 |
| Aldosterone (pg/mL) | 0.114 | 0.15 |
| Cystatin C (mg/L) | 0.637 | <0.0001 |
| NT-pro BNP (pg/mL) | 0.589 | <0.0001 |
| Pro-collagen III amino-terminal peptide (ug/L) | 0.442 | <0.0001 |
| High sensitivity troponin I (pg/mL) | 0.438 | <0.0001 |
| Endothelin-1 (pg/mL) | 0.571 | <0.0001 |
| High sensitivity C-reactive protein (mg/L) | 0.046 | 0.56 |
| Carboxy-Terminal Telopeptide of Collagen Type I (ug/L) | 0.651 | <0.0001 |
| Galectin 3 (ng/mL) | 0.350 | <0.0001 |
eGFR= estimated glomerular filtration rate, NT-proBNP= amino-terminal pro-B-type natriuretic peptide
Baseline concentrations of NT-proBNP and galectin-3 were also examined with relationship to echocardiographic and MRI parameters, divided by the respective median concentrations, as shown in Supplemental Table 2. As anticipated, baseline NT-proBNP showed similar correlations to echocardiographic parameters of diastolic function when compared to IGFBP7; however, unlike IGFBP7, baseline NT-proBNP also demonstrated significant correlations with measures of systolic function including LV stroke volume and cardiac index, whereas IGFBP7 did not. Furthermore, NT-proBNP was not significantly correlated with LV hypertrophy whereas IGFBP7 was weakly, but significantly correlated. NT-proBNP also did not demonstrate significant inverse correlation with aortic distensability (ρ= −0.23, P=0.07). Baseline galectin-3 did not show significantly consistent correlation with any of these measures.
IGFBP7 was not independently predictive of measures of diastolic dysfunction such as E/A >1.5, E/E’ >15, E’ ≤ 0.08 m/sec, and LV mass index after adjustment for NT-proBNP, age, and eGFR; however, NT-proBNP retained independent value in each model except LV mass index. To further investigate the relationship between IGFBP7 and NT-proBNP, we identified the optimal cutpoint using receiver operating curves for each marker with respect to the diastolic function parameters of E/A >1.5 and E/E’ >15. The area under the curve (AUC) for IGFBP7 was 0.68 for E/A >1.5 and 0.71 for E/E’>15; for NT-proBNP the values were 0.74 and 0.67, respectively. The majority of patients with E/A >1.5 and E/E’>15 had elevated concentrations of both markers (74% and 78% for each parameter, respectively). Notably, fewer patients with abnormal diastolic function had either elevated IGFBP7 concentrations without elevated NT-proBNP concentrations (44% for E/A >1.5 and 25% for E/E’>15) or elevated NT-proBNP without elevated IGFBP7 concentrations (64% for E/A >1.5 and 54% for E/E’>15).
Change in IGFBP7, imaging, and functional capacity over 24 weeks
The relationship between ΔIGFBP7 and imaging parameters are shown in Table 4. As seen with baseline measurements, significant, but weak correlations were found between ΔIGFBP7 and Δechocardiographic measures of diastolic function including Δ in transmitral E velocity (ρ=0.11; P=0.006), E/A ratio (ρ=0.22; P=0.02), E/E’ (ρ=0.17; P=0.002), and estimated RVSP (ρ=0.28; P=0.04). When examining the relationship between IGFBP7 and VO2max at 24 weeks, patients with supramedian VO2max (> 11.9 mL/kg/min) had a significantly lower median IGFBP7 concentration of 206 ng/mL versus 244 ng/mL in those with inframedian VO2max (P=0.004). Similarly, patients with supramedian 24 week IGFBP7 concentrations (>228 ng/mL) had a numerical trend towards lower peak VO2max compared to those with inframedian concentrations (11.4 versus 12.8 mL/min/kg; P=0.10). Spearman correlation (adjusted for baseline VO2max and IGFBP7) revealed a weak, but significant inverse relationship between ΔIGFBP7 and ΔVO2max over the 24 week study period (ρ= -0.19, P=0.013), as shown in Figure 2.
Table 4.
Change in IGFBP7 and imaging parameters over 24 weeks.
| Diastolic function measure | N | Spearman correlation with change in IGFBP7 (ρ) | Spearman correlation P value | Adjusted P value |
|---|---|---|---|---|
| Echocardiography change from baseline to week 24 | ||||
| MV inflow: E velocity at leaf tip (m/sec) | 144 | 0.114 | 0.17 | 0.006 |
| MV inflow: A velocity at leaf tip (m/sec) | 102 | −0.141 | 0.16 | 0.14 |
| E/A ratio | 102 | 0.225 | 0.02 | 0.02 |
| LV relaxation septal (medial) - E' (m/sec) | 135 | −0.096 | 0.27 | 0.06 |
| Filling pressure septal (medial) - E/E' (m/sec) | 133 | 0.174 | 0.05 | 0.001 |
| LA volume index | 99 | 0.024 | 0.81 | 0.52 |
| RVSP (mmHg) | 88 | 0.282 | 0.008 | 0.03 |
| Cardiac MRI change from baseline to week 24 | ||||
| LV stroke volume | 78 | −0.674 | 0.56 | 0.75 |
| Cardiac index | 77 | −0.139 | 0.23 | 0.25 |
| Aortic distensibility | 46 | 0.067 | 0.66 | 0.90 |
MV= mitral valve, LV= left ventricle, LA= left atrium, RVSP= estimated right ventricular systolic pressure, MRI= magnetic resonance imaging, CPET= cardiopulmonary exercise testing, VO2max= peak oxygen consumption
Figure 2. Correlation between ΔIGFBP7 and ΔVO2max over 24 weeks.
There was a significant inverse correlation over the study period. The P value comes from a linear regression model that adjusts for baseline IGFBP7 and VO2max.
Correlations between ΔNT-proBNP and Δgalectin-3 with Δ in measures of diastolic function from echocardiography and MRI were also examined, as shown in Supplemental Table 3. ΔNT-proBNP again demonstrated correlation with Δ in echocardiographic diastolic function parameters. Similar to baseline measurements, Δgalectin-3 did not demonstrate any significant correlations with Δ in imaging parameters.
Relationship between Sildenafil and IGFBP7
Baseline and 6 month IGFBP7 concentrations were similar in those randomized to sildenafil compared to those receiving placebo (240 versus 215 ng/mL, P=0.17), as shown in Supplemental Table 4. However, when examining ΔIGFBP7 over 24 weeks, those receiving sildenafil had significantly lower concentrations of IGFBP7, whereas those receiving placebo showed a significant increase in IGFBP7 concentration (−1.5 ng/mL versus +13.6 ng/mL, P=0.001).
Discussion
In our analysis of the relationship between IGFBP7 and indices of diastolic function and functional capacity in patients with HFpEF, we have identified several interesting findings. First, patients with supramedian IGFBP7 concentrations were more likely to show a higher risk profile, with older age, more atrial arrhythmia, greater signs of congestion on history or physical exam, and more elevated concentrations of other established biomarkers. Second, we confirm our hypothesis that both baseline and ΔIGFBP7 were modestly but significantly associated with echocardiographic parameters of diastolic function such as transmitral E velocity, E/E’, and estimated RVSP, and weakly correlated with E/A ratio. Third, as hypothesized, supramedian IGFBP7 concentrations were also significantly associated with lower VO2max at baseline; furthermore, ΔIGFBP7 was weakly inversely correlated with ΔVO2max over 24 weeks. Fourth, subjects receiving sildenafil had a significant reduction in IGFBP7 concentrations over the study period when compared to those receiving placebo. To our knowledge, this is the first report investigating IGFBP7 relative to imaging parameters, functional capacity and sildenafil therapy in patients with HFpEF.
IGFBP7 is a biomarker that is associated with G1 phase cell cycle arrest (19) part of the cellular senescence associated secretome. Higher levels of IGFBP7 may therefore accelerate cellular senescence, potentially contributing to aging of the myocardium (20). IGFBP7 has been localized to the Weibel-Palade bodies (storage organelles) of endothelial cells (21). Furthermore, IGFBP7 has also been associated with collagen deposition (22) and may therefore also be linked with myocardial fibrosis. The syndrome of HFpEF has been recently proposed to be a pro-inflammatory state with production of reactive oxygen species by endothelial cells that subsequently results in concentric LV remodeling, stiffening of cardiomyocytes, and increased collagen deposition which all ultimately lead to diastolic dysfunction (23). Given association between IGFBP7, senescence, and collagen deposition as well as its localization to endothelial cells, IGFBP7 may be linked to later stages of this pro-inflammatory cascade, which could elucidate its relationship with echocardiographic parameters of diastolic dysfunction and functional capacity. IGFBP7 is less likely to be a direct marker of inflammation, however, as it was not correlated with C-reactive protein. Finally, another potential mechanism of elevated IGFBP7 concentrations in patients with HFpEF could be via the cardiorenal axis, as IGFBP7 has recently been studied as a marker of acute kidney injury and the strongest of the correlations with other biomarkers is with the renal marker cystatin C (19).
The syndrome of HFpEF remains a clinical conundrum, and biomarkers may provide further insight into the complexities of HFpEF potentially leading to better understanding of this condition, and supplement what is ascertained by echocardiography and physical exam. The NPs have previously been associated with diastolic dysfunction (7,8) and may imply propensity for treatment response in patients with HFpEF (24), however other biomarkers may add to information provided by BNP or NT-proBNP. In the current study, both NT-proBNP and IGFBP7 had similar correlations with parameters of diastolic dysfunction, similar AUC with regards to the diastolic function parameters of E/A and E/E’, and were strongly correlated with each other; however, they may reflect different aspects of diastolic dysfunction given that NT-proBNP also demonstrated correlation with cardiac structure and function beyond diastolic abnormalities, while IGFBP7 did not. This specificity of IGFBP7 may result in further improved phenotyping of patients with HFpEF above what has been achieved with the natriuretic peptides. Furthermore, the majority of patients with significant diastolic dysfunction had elevations in both IGFBP7 and NT-proBNP, whereas fewer had elevations in only one of the biomarkers, raising the possibility that the combination of these markers may be superior to either in isolation for identifying diastolic dysfunction. Future investigation of IGFBP7 as a tool to better phenotype patients with HFpEF is needed in order to explore the hypothesis that elevation in the biomarker identifies a therapeutic imperative. To this point, it is noteworthy that patients receiving sildenafil therapy had a reduction in IGFBP7 concentrations compared to those receiving placebo, whereas in the original study those receiving sildenafil had increase in NT-proBNP (17); whether this observation may identify a subgroup where sildenafil therapy could be beneficial is unclear.
It is also notable IGFBP7 showed a significant inverse correlation with VO2max at baseline, and ΔIGFBP7 was correlative with ΔVO2max. As worse diastolic function is inversely linked to exercise capacity, it is tempting to speculate the correlation between IGFBP7 values and functional capacity may reflect changes in diastolic function; invasive analyses exploring this hypothesis are planned. Serial data with biomarkers, echocardiography, and functional capacity are not often readily available in HF trials, thus our study is important, lending comprehensive analyses with both baseline and change in each of these parameters.
While our study introduces novel findings regarding IGFBP7 in patients with HFpEF, limitations are present. First, our study is retrospective, and thus our findings can only be regarded as hypothesis-generating. Our study is also limited by a small sample size, missing data for some parameters, and a short follow up of 24 weeks. However, even with these shortcomings, we confirmed our hypothesis of modest, but significant correlation between IGFBP7 and parameters of diastolic function and functional capacity. Second, given limited sample size we could not investigate outcomes as a function of IGFBP7, as our primary focus was mechanistic, but we have previously demonstrated prognostic value of IGFBP7 in another cohort of patients with HFpEF (25). Third, supramedian IGFBP7 concentrations were associated with other established biomarkers, raising the concern IGFBP7 provides redundant biological information, particularly with regards to NT-proBNP. However, IGFBP7 was much less correlated with systolic parameters, suggesting IGFBP7 and NT-proBNP are likely reflecting different aspects of diastolic function. Our goal was to pursue characterization of a marker more specific to the molecular processes leading to echocardiographic parameters of diastolic dysfunction to assist with further understanding of the pathophysiology and mechanism of HFpEF, with the ultimate possibility of objectively phenotyping this heterogeneous population of patients. While speculative, given the paucity of treatments for HFpEF (whose efficacy was tested largely on a clinical picture of HF with an LVEF above 50%), it may be possible that a combination of clinical data, echocardiographic imaging, and molecular phenotyping may more effectively inform responses to therapy. Finally, given the relationship between IGFBP7 and sildenafil therapy, there may be an association between IGFBP7 and cyclic guanosine monophosphate, but these data were not available in the current study.
In conclusion, IGFBP7 may be a novel biomarker of diastolic dysfunction in patients with HFpEF. Furthermore, IGFBP7 also appears to be modestly related to functional capacity, itself an important component to the syndrome of HFpEF. A better understanding of the link(s) between IGFBP7 and diastolic function would be of value, as the search continues for further clarity with regards to diagnosis and treatment of patients with HFpEF. More studies are needed in order to elucidate the optimal diagnostic and treatment strategies for patients who suffer from this complex condition.
Supplementary Material
Perspectives.
Competency in Medical Knowledge
The syndrome of HFpEF remains a clinical challenge, given its complicated diagnosis and lack of viable therapeutic options. The current study examines the novel biomarker IGFBP7 and demonstrates its weak to modest correlation with diastolic function and functional capacity in patients with HFpEF.
Translational Outlook
Biomarkers such as IGFBP7 can provide insight into the complex pathophysiology of HFpEF. Given the weak to modest relationship of IGFBP7 with diastolic function and functional capacity as well as potential interaction between IGFBP7 and therapy, it is conceivable that in the future, IGFBP7 may play a role in personalizing the care of patients with HF.
Acknowledgments
Funding Sources: Dr. Gandhi is supported by the Dennis and Marilyn Barry Fellowship in Cardiology. Dr. Gaggin is supported in part by the Clark Fund for Cardiac Research Innovation. Dr. Januzzi is supported in part by the Roman W. DeSanctis Clinical Scholar Endowment and the Hutter Family Professorship. The Heart Failure Clinical Research Network is supported by grants HL084861, HL084875, HL084877, HL084889, HL084890, HL084891, HL084899, HL084904, HL084907, and HL084931.
Abbreviations
- HFpEF1
Heart Failure with Preserved Ejection Fraction
- NP
Natriuretic Peptides
- IGFBP7
Insulin-like growth factor-binding protein 7
- HFrEF
Heart Failure with reduced Ejection Fraction
- LAVi
Left Atrial Volume Index
- RVSP
Right Ventricular Systolic Pressure
- VO2max
Peak oxygen consumption
- CPET
Cardiopulmonary Exercise Testing
- PIIINTP
Pro-collagen type III amino-terminal peptide
- hsTnI
Highly sensitive troponin I
Footnotes
Disclosures: Dr. Gaggin has received grant support from Roche and Portola; consulting income from Roche Diagnostics, American Regent, Boston Heart Diagnostics and Critical Diagnostics; research payments for clinical endpoint committees for EchoSense. Dr. Januzzi has received grant support from Siemens, Prevencio, and Singulex; consulting income from Roche Diagnostics, Critical Diagnostics, Sphingotec, and Novartis; and research payments for clinical endpoint committees/data safety monitoring boards from Novartis, Amgen, Janssen, and Boehringer Ingelheim. Dr. Liu has received research grants from Genome Canada, with which Roche Diagnostics is a partner. The remaining authors have nothing to disclose.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Bhatia RS, Tu JV, Lee DS, et al. Outcome of heart failure with preserved ejection fraction in a population-based study. N Engl J Med. 2006;355:260–9. doi: 10.1056/NEJMoa051530. [DOI] [PubMed] [Google Scholar]
- 2.Owan TE, Hodge DO, Herges RM, Jacobsen SJ, Roger VL, Redfield MM. Trends in prevalence and outcome of heart failure with preserved ejection fraction. N Engl J Med. 2006;355:251–9. doi: 10.1056/NEJMoa052256. [DOI] [PubMed] [Google Scholar]
- 3.Nagueh SF, Appleton CP, Gillebert TC, et al. Recommendations for the evaluation of left ventricular diastolic function by echocardiography. J Am Soc Echocardiogr. 2009;22:107–33. doi: 10.1016/j.echo.2008.11.023. [DOI] [PubMed] [Google Scholar]
- 4.Januzzi JL, Jr, Felker GM. Surfing the biomarker tsunami at JACC: heart failure. JACC Heart Fail. 2013;1:213–5. doi: 10.1016/j.jchf.2013.03.007. [DOI] [PubMed] [Google Scholar]
- 5.Komajda M, Carson PE, Hetzel S, et al. Factors associated with outcome in heart failure with preserved ejection fraction: findings from the Irbesartan in Heart Failure with Preserved Ejection Fraction Study (I-PRESERVE) Circ Heart Fail. 2011;4:27–35. doi: 10.1161/CIRCHEARTFAILURE.109.932996. [DOI] [PubMed] [Google Scholar]
- 6.van Veldhuisen DJ, Linssen GC, Jaarsma T, et al. B-type natriuretic peptide and prognosis in heart failure patients with preserved and reduced ejection fraction. J Am Coll Cardiol. 2013;61:1498–506. doi: 10.1016/j.jacc.2012.12.044. [DOI] [PubMed] [Google Scholar]
- 7.Tschope C, Kasner M, Westermann D, Gaub R, Poller WC, Schultheiss HP. The role of NT-proBNP in the diagnostics of isolated diastolic dysfunction: correlation with echocardiographic and invasive measurements. Eur Heart J. 2005;26:2277–84. doi: 10.1093/eurheartj/ehi406. [DOI] [PubMed] [Google Scholar]
- 8.Yu CM, Sanderson JE, Shum IO, et al. Diastolic dysfunction and natriuretic peptides in systolic heart failure. Higher ANP and BNP levels are associated with the restrictive filling pattern. Eur Heart J. 1996;17:1694–702. doi: 10.1093/oxfordjournals.eurheartj.a014753. [DOI] [PubMed] [Google Scholar]
- 9.Yancy CW, Jessup M, Bozkurt B, et al. 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association Task Force on practice guidelines. Circulation. 2013;128:e240–327. doi: 10.1161/CIR.0b013e31829e8776. [DOI] [PubMed] [Google Scholar]
- 10.Barasch E, Gottdiener JS, Aurigemma G, et al. Association between elevated fibrosis markers and heart failure in the elderly: the cardiovascular health study. Circ Heart Fail. 2009;2:303–10. doi: 10.1161/CIRCHEARTFAILURE.108.828343. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Krum H, Elsik M, Schneider HG, et al. Relation of peripheral collagen markers to death and hospitalization in patients with heart failure and preserved ejection fraction: results of the I-PRESERVE collagen substudy. Circ Heart Fail. 2011;4:561–8. doi: 10.1161/CIRCHEARTFAILURE.110.960716. [DOI] [PubMed] [Google Scholar]
- 12.de Boer RA, Lok DJ, Jaarsma T, et al. Predictive value of plasma galectin-3 levels in heart failure with reduced and preserved ejection fraction. Ann Med. 2011;43:60–8. doi: 10.3109/07853890.2010.538080. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Manzano-Fernandez S, Mueller T, Pascual-Figal D, Truong QA, Januzzi JL. Usefulness of soluble concentrations of interleukin family member ST2 as predictor of mortality in patients with acutely decompensated heart failure relative to left ventricular ejection fraction. Am J Cardiol. 2011;107:259–67. doi: 10.1016/j.amjcard.2010.09.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Chugh S, Ouzounian M, Lu Z, et al. Pilot study identifying myosin heavy chain 7, desmin, insulin-like growth factor 7, and annexin A2 as circulating biomarkers of human heart failure. Proteomics. 2013;13:2324–34. doi: 10.1002/pmic.201200455. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Motiwala SR, Szymonifka J, Belcher A, et al. Measurement of novel biomarkers to predict chronic heart failure outcomes and left ventricular remodeling. J Cardiovasc Transl Res. 2014;7:250–61. doi: 10.1007/s12265-013-9522-8. [DOI] [PubMed] [Google Scholar]
- 16.Gandhi PU, Gaggin HK, Sheftel AD, et al. Prognostic usefulness of insulin-like growth factor-binding protein 7 in heart failure with reduced ejection fraction: a novel biomarker of myocardial diastolic function? Am J Cardiol. 2014;114:1543–9. doi: 10.1016/j.amjcard.2014.08.018. [DOI] [PubMed] [Google Scholar]
- 17.Redfield MM, Chen HH, Borlaug BA, et al. Effect of phosphodiesterase-5 inhibition on exercise capacity and clinical status in heart failure with preserved ejection fraction: a randomized clinical trial. JAMA. 2013;309:1268–77. doi: 10.1001/jama.2013.2024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Redfield MM, Borlaug BA, Lewis GD, et al. PhosphdiesteRasE-5 Inhibition to Improve CLinical Status and EXercise Capacity in Diastolic Heart Failure (RELAX) trial: rationale and design. Circ Heart Fail. 2012;5:653–9. doi: 10.1161/CIRCHEARTFAILURE.112.969071. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kashani K, Al-Khafaji A, Ardiles T, et al. Discovery and validation of cell cycle arrest biomarkers in human acute kidney injury. Crit Care. 2013;17:R25. doi: 10.1186/cc12503. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Severino V, Alessio N, Farina A, et al. Insulin-like growth factor binding proteins 4 and 7 released by senescent cells promote premature senescence in mesenchymal stem cells. Cell Death Dis. 2013;4:e911. doi: 10.1038/cddis.2013.445. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.van Breevoort D, van Agtmaal EL, Dragt BS, et al. Proteomic screen identifies IGFBP7 as a novel component of endothelial cell-specific Weibel-Palade bodies. J Proteome Res. 2012;11:2925–36. doi: 10.1021/pr300010r. [DOI] [PubMed] [Google Scholar]
- 22.Guo XH, Liu LX, Zhang HY, et al. Insulin-like growth factor binding protein-related protein 1 contributes to hepatic fibrogenesis. J Dig Dis. 2014;15:202–10. doi: 10.1111/1751-2980.12126. [DOI] [PubMed] [Google Scholar]
- 23.Paulus WJ, Tschope C. A novel paradigm for heart failure with preserved ejection fraction: comorbidities drive myocardial dysfunction and remodeling through coronary microvascular endothelial inflammation. J Am Coll Cardiol. 2013;62:263–71. doi: 10.1016/j.jacc.2013.02.092. [DOI] [PubMed] [Google Scholar]
- 24.Pitt B, Pfeffer MA, Assmann SF, et al. Spironolactone for heart failure with preserved ejection fraction. N Engl J Med. 2014;370:1383–92. doi: 10.1056/NEJMoa1313731. [DOI] [PubMed] [Google Scholar]
- 25.Gandhi PU, Chow SL, Rector TS, et al. Prognostic Value of Insulin-like Growth Factor-binding Protein 7 in patients with Heart Failure and Preserved Ejection Fraction. J Cardiac Fail. 2016 doi: 10.1016/j.cardfail.2016.06.006. (submitted) [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.