Methods

The study was approved by the Partners Healthcare Institutional Review Board and complied with the Declaration of Helsinki. Informed consent was obtained from all patients. The PROTECT study was a randomized, single-center, prospective proof-of-concept trial that compared standard HF care with N-terminal pro–B-type natriuretic peptide (NT-proBNP)–guided care with a goal of reducing natriuretic peptide levels to ≤1,000 pg/ml in patients with HFrEF. The methods and results of the PROTECT study have been published previously. Of the 151 patients enrolled in the PROTECT study, 124 had available samples for serial IGFBP7 measurement and baseline echocardiographic studies; they were seen during a mean follow-up period of 10 months, with a total of 882 office visits. Details regarding biomarker measurements are included in the Supplementary Index .

Detailed 2-dimensional transthoracic echocardiography was performed at baseline and when possible at completion of the study, a mean of 10 months later. The follow-up echocardiographic examination was performed ±2 weeks from the last study visit and on the same day as the final blood draw. Further details regarding the echocardiography protocol and abnormal values for this analysis are provided in the Supplementary Index . Interpretation of the ultrasound studies was done by 2 readers (R.B.W. and A.L.B.) blinded to study assignment and biomarker results.

All continuous variables are expressed as mean ± SD for normally distributed values; non-normal variables were identified using the Kolmogorov-Smirnov test and are expressed as medians and interquartile ranges. Comparisons of variables as a function of IGFBP7 concentrations were performed using Student’s t test, the Mann-Whitney U test, or the chi-square test, as appropriate. For continuous variables expressed in multiple categories, the Kruskal-Wallis test was used. To examine linear associations between IGFBP7 concentrations and other continuous variables, we first ln-transformed non-normal results, to achieve normality. Then, Spearman correlations were performed, with the ρ value expressed accompanied by the corresponding p value. Selected scatterplots were generated to provide a graphical representation of correlations. After univariable correlations, we examined the available data for independent predictors of the dependent variable of IGFBP7 concentrations; all candidate variables required a retention p value ≤0.10 to be included in the multivariable linear regression model. The variables included were age, ischemic cause of cardiomyopathy, hepatomegaly on examination, NT-proBNP, left atrial volume index (LAVi), right ventricular systolic pressure (RVSP), right atrial size (superior-inferior), A peak velocity, E/A ratio, S-wave deceleration time, E/E′ ratio, and E′/A′ ratio. After the construction of the best-fit linear model, β coefficients for each predictor of IGFBP7 were calculated. Furthermore, we performed Spearman correlations with several other novel biomarkers of HF and echocardiographic diastolic parameters of E/A, E/E′, pulmonary venous flow S/D ratio, RVSP, and LAVi and compared them with IGFBP7.

Consistent with previous work, clinical categorizations were made based on an IGFBP7 level of 117.8 ng/ml, a threshold prognostic for events. Additionally, exploration of secular trends in IGFBP7 to predict change in diastolic function was based on time spent at <117.8 ng/ml integrated over time in study; this was done using baseline and final values as well as time spent at <117.8 ng/ml. Furthermore, correlation between baseline and final values was performed using Spearman’s ρ. Last, the percentage of time spent with IGFBP7 <117.8 ng/ml was also used to analyze risk for adverse outcomes, using univariate followed by multivariable Cox proportional-hazards, modeling the association with the incidence of any cardiovascular events. The multivariable model was adjusted for epidemiologically and clinically relevant risk factors and included the following: treatment arm allocation in PROTECT, age, gender, New York Heart Association class, LV ejection fraction, E/A ratio, E/E′ ratio, LAVi, and RVSP, as well as baseline estimated glomerular filtration rate and NT-proBNP. Statistical analysis was performed with PASW version 17.0 (SPSS, Inc., Chicago, Illinois), with 2-sided p values considered significant if <0.05.

Results

Table 1 lists the baseline characteristics of the subjects in the present analysis. Of note, the previously determined prognostic threshold of ≥117.8 ng/ml ³ represented the highest tertile for the group. Supplementary Table 1 lists the baseline characteristics subdivided by the IGFBP7 concentration of 117.8 ng/ml. Supplementary Table 2 compares baseline echocardiographic measurements of study participants, also divided using this prognostic threshold. No clear associations were present between IGFBP7 concentrations and LV mass, LV volumes, or LV ejection fraction. In contrast, significant associations were found between IGFBP7 concentrations and measures that describe abnormalities of diastolic function, namely, LAVi, higher transmitral Doppler E-wave velocity, lower transmitral Doppler A-wave velocity (corresponding direct association with higher E/A ratio), and higher E/E′ ratio. Furthermore, more patients with high IGFBP7 had E/E′ ratio >15 and were significantly more likely to have higher RVSP, along with a more dilated right ventricle.

We then focused on correlations between ln-IGFBP7 and various echocardiographic parameters, listed in Table 2 . In univariate analyses, we detected numerous significant correlations between ln-IGFBP7 concentrations and cardiac structure and function, such as atrial size, right ventricular pressures, and significant correlations with diastolic parameters, such as transmitral Doppler E and A waves (as well as their ratio), pulmonary vein Doppler velocity and slope of the D wave (as well as marginal correlation with pulmonary vein S/D ratio), and E/E′ ratio. In contrast, we found no correlation with LV volume, LV mass, or the LV ejection fraction. Supplementary Figure 1 shows scatterplots of transmitral Doppler E/A, E/E′, pulmonary vein S/D ratio, LAVi, and RVSP in relation to IGFBP7 concentrations.

Table 2

Univariable correlation and multivariable linear regression analyses of insulin-like growth factor-binding protein 7 concentrations and baseline echocardiographic parameters

Characteristic	Univariable		Multivariable
	ρ	p	Standardized β Coefficient	T	p
Left ventricular ejection fraction	0.018	0.08
Left ventricular end systolic volume index	0.087	0.59
Left ventricular end diastolic volume index	0.090	0.58
Left ventricular mass	0.120	0.18
Ventricular septal thickness	0.116	0.20
Posterior wall thickness	0.034	0.71
Left ventricular outflow tract velocity	0.128	0.16
Left atrial size, anterior-posterior	0.265	0.003
Left atrial size, superior-inferior	0.231	0.01
Left atrial volume index	0.237	0.008	0.386	2.857	0.01
Right ventricular end diastolic dimension	0.202	0.03
Right ventricular end systolic dimension	0.009	0.90
Right ventricular free wall thickness	0.222	0.02
Right ventricular fractional area change	0.240	0.009
Right ventricular systolic pressure	0.316	0.001	0.317	4.137	0.001
Right atrial size, inflow	0.029	0.76
Right atrial size, superior-inferior	0.287	<0.001	0.584	5.408	<0.001
E peak velocity	0.230	0.01
A peak velocity	−0.208	0.03	−0.592	−5.484	<0.001
E/A	0.304	0.001	0.601	5.826	0.005
Pulmonary vein peak systolic velocity	−0.147	0.14
Pulmonary vein peak systolic velocity deceleration slope	−0.120	0.24
Pulmonary vein peak systolic velocity deceleration time	0.120	0.24	0.437	5.612	<0.001
Pulmonary vein peak anterograde diastolic velocity	0.411	<0.001
Pulmonary vein peak anterograde diastolic velocity deceleration slope	0.316	0.001
Pulmonary vein peak anterograde diastolic velocity deceleration time	−0.048	0.64
Pulmonary vein peak systolic velocity/peak anterograde diastolic velocity	−0.173	0.09
Tissue Doppler peak E velocity	−0.01	0.92
Tissue Doppler peak A velocity	0.287	0.003
Tissue Doppler E/E′	0.257	0.005	0.222	2.673	0.02
Tissue Doppler E′/A′	−0.364	0.005	−0.759	7.293	<0.001

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Coronary Atheroma Regression and Plaque Characteristics Assessed by Grayscale and Radiofrequency Intravascular Ultrasound After Aerobic Exercise
2. Comparing Changes in Severe Versus Mild Perfusion Defect Size in Patients Who Underwent Serial Rubidium-82 Positron Emission Tomography Myocardial Perfusion Imaging
3. Aldosterone-Cortisol Imbalance Immediately After Fontan Operation With Implications for Abnormal Fluid Homeostasis
4. Gluten Sensitivity: New Epidemic or New Myth? Every Major Change in Our Diet Carries With It the Possibility of Unforeseen Risks

Stay updated, free articles. Join our Telegram channel

Join