Cardiac troponin I in congenital heart defects with pressure or volume overload

Abstract

Objective. To evaluate the prevalence of cardiac troponin I (cTnI) and autoantibodies to cTn in children with congenital heart defects with volume or pressure overload fulfilling the criteria for treatment, and in healthy children. Design. The study groups comprised 78 children with volume overload caused by an atrial septal defect or a patent ductus arteriosus, and 60 children with pressure overload caused by coarctation of the aorta or stenosis of the aortic or the pulmonary valve, and 74 healthy controls. Serum levels of natriuretic peptides, cTnI, and autoantibodies to cTn were analyzed at baseline, prior to treatment and in 64 patients 6 months after treatment. Results. At baseline, one child with volume overload, 12 children with pressure overload, and one healthy control had positive cTnI. Further analysis of the pressure overload subgroup revealed that the children with positive cTnI were younger than those with negative cTnI, and had higher levels of natriuretic peptides. The pressure gradient at the coarctation site or stenotic valve was higher in those with positive TnI. Six months after treatment, 63 of 64 children examined were cTnI negative. Conclusions. The cTnI release is more frequently associated with pressure than volume overload which resolves after treatment in most children.

Key words::

• autoantibodies
• congenital heart defect
• troponin

Introduction

The troponin complex comprises troponin C, troponin I, and troponin T. Of these, cardiac troponins T (cTnT) and I (cTnI) are myofibrillar regulatory proteins which are highly specific to myocardial cells. They are expressed throughout ventricular and atrial tissues. Only small portion is located in the cytosolic compartment, with the vast majority bound to the contractile apparatus (1). cTnI is a very specific and sensitive marker of myocardial injury (2). Use of a high-sensitivity assay for cardiac troponin makes it possible to find detectable levels of circulating cardiac troponins in heart failure (3). In addition, in patients with chronic heart failure, nonacutely elevated cTnT concentrations have been associated with adverse prognosis (4).

The natriuretic peptide system plays an important role in regulation of fluid homeostasis, blood pressure/vascular tone, and cardiomyocyte growth. This system functions to retard the progression of heart failure, opposing the effect of neurohumoral systems (5). Levels of natriuretic peptides are higher immediately after birth and decrease thereafter, during the first 2–4 months (6). There is some evidence that serum levels of natriuretic peptides and cardiac troponin may have a weak correlation (7).

Some studies exist concerning perioperative values of cardiac troponins and their association with outcome and complications in patients with congenital heart defects. Little is known, however, about serum levels of cTnI in patients with congenital heart defects causing volume or pressure overload prior to treatment (8,9). Advanced methods are needed for differential diagnostics in patients with congenital heart defects and symptoms like chest pain, decreased exercise tolerance, and shortness of breath. Additionally, further evaluation is needed in order to titrate medications and to time interventions properly. Children at risk for subtle heart failure are especially a challenge. The timing of treatment for congenital heart defect is based on the hemodynamic and anatomic situations, with consideration of myocardial cell adaptation and chamber remodeling. It is, therefore, important to have multiple methods available for decision-making regarding the timing of intervention and follow-up evaluation.

Methods

This prospective follow-up study was carried out at the Children's Hospital, Helsinki University Hospital, University of Helsinki, Helsinki, Finland, between February 2003 and February 2006. All parents agreed to have their children participate in this clinical trial approved by the hospital ethics committee and gave their written informed consent. We wanted to study the effect of volume and pressure overload on cTnI levels separately. As examples of volume overload-causing chamber dilatation, we chose to examine patients with an atrial septal defect (ASD) causing right-sided volume overload as well as those with patent ductus arteriosus (PDA) causing left-sided volume overload. As examples of pressure overload causing ventricular wall hypertrophy, we chose to examine patients with an aortic valve stenosis (AS) and coarctation of the aorta (CoA) causing left ventricular pressure overload as well as those with pulmonary valve stenosis (PS) causing right ventricular pressure overload.

The study groups comprised 78 children with volume overload caused by an ASD or a PDA, and 60 children with pressure overload caused by CoA, AS or PS. In addition, controls were 74 healthy children (Table I). All patients fulfilled the criteria for treatment of the cardiac defect: specifically, they were:

Table I. Data on healthy control children and patients according to the type of congenital heart defect at baseline. Values are median (range).

	Controls	ASD	PDA	CoA	PS	AS
N	74	41	37	40	10	10
Age (years)	5.4 (0.0–15.6)	6.4 (2.3–18.5)	2.3 (0.6–10.6)	6.2 (0.0–19.0)	0.1 (0.0–8.4)	0.2 (0.0–6.5)
Female/Male	31/43	9/32	13/24	26/14	8/2	7/3
Weight (kg)	20.0 (3.1–66.5)	22.0 (12.4–64.7)	13.0 (5.0–32.8)	23.8 (2.7–73)	4.5 (3.0–25.4)	6.6 (2.7–15.4)
Height (cm)	114 (48–175)	118 (83–164)	88 (61–141)	123 (48–187)	56 (49–126)	64 (44–109)

ASD = atrial septal defect, PDA = patent ductus ateriosus, CoA = coarctation of aorta, PS = stenosis of pulmonary valve, AS = stenosis of aortic valve.

• 1. Forty-one children undergoing cardiac evaluation before percutaneous or surgical closure of an ASD. The indication for intervention was a significant volume overload of the right side of the heart. Twenty patients were examined 6 months after ASD closure (percutaneous n = 13, surgical n = 7).

• 2. Thirty-seven children with PDA undergoing cardiac evaluation before percutaneous (n = 33) or surgical (n = 4) closure of a PDA. The indication for intervention was an audible murmur. All patients with PDA had normal pulmonary artery pressures as measured by echocardiographic Doppler gradient between the aorta and the pulmonary artery and in those treated percutaneously also invasively in the catheterization laboratory. In this patient group, one child had clinical signs of congestive heart failure and was treated with diuretics prior to PDA closure. All other patients were asymptomatic. One child was diagnosed with Mulibrey nanism. Twenty-five patients were examined 6 months after percutaneous closure of the PDA.

• 3. Forty children were diagnosed with unoperated (native) CoA or recurrent CoA (reCoA) requiring intervention. The indication for intervention was duct dependency or, beyond infancy, an arm to leg gradient of 20 mm Hg or more and a significant anatomic narrowing in echocardiography ± CT angiography. Fifteen children were operated on, and 25 children underwent percutaneous angioplasty of the coarctation site. Fifteen children were examined 6 months after the procedure (percutaneous n = 4, surgical n = 11).

• 4. Ten children with AS were examined before percutaneous valvuloplasty. In this patient group, the indication for intervention was duct dependency, or a peak-to-peak gradient of over 50 mm Hg. One child with CATCH-22 syndrome had previously been operated on for interrupted aortic arch and ventricular septal defect. Two children were examined 6 months after the procedure.

• 5. Ten children with PS were examined before percutaneous valvuloplasty. The indication for intervention was duct dependency, or a Doppler gradient of over 50 mm Hg. One child was diagnosed with Noonan syndrome. Two children were examined six months after the procedure.

All patients and controls underwent clinical cardiovascular examination and blood test sampling for measurement of natriuretic peptides, cTnI, and cardiac troponin-specific autoantibodies (cTnAAb) just prior to intervention. Echocardiography was performed according to published guidelines (10). We had 74 control children, and they were examined once. They were asymptomatic and showed no abnormalities in clinical examination, ECG, or echocardiography.

Demographics of children with ASD, PDA, CoA, AS, and PS, and the healthy controls are in Table I. Characteristics of the patients divided into subgroups according to loading conditions (volume or pressure overload) are in Table II.

Table II. Characteristics of healthy controls and patients in subgroups according to loading conditions at baseline. The pressure-overload group comprised children with coarctation of the aorta, aortic valve stenosis, or pulmonary valve stenosis. The volume-overload group comprised children with atrial septal defect or patent ductus arteriosus. Values are median (range).

	Controls	Pressure overload	Volume overload
N	74	60	78
age (years)	5.4 (0.0–15.6)	0. 7 (0.0–19.0)*	3.9 (0.6–18.5)§
weight (kg)	20.0 (3.1–66.5)	10.7 (2.7–73.0)*	16.8 (5.0–64.7)§
height (cm)	114 (48–175)	74 (44–187)*	103 (61–164)§
ANPN (nmol/l)	0.30 (0.10–1.60)	0.63 (0.10–36.00)†	0.41 (0.11–4.90)†‡
NT-proBNP (ng/l)	68 (14–3224)	167 (4–35001)†	107 (5–6027)†§

ANPN = N-terminal proatriopeptide, NT-proBNP = N-terminal pro-brain natriuretic peptide.

*P < 0.001 as compared with controls.

†P < 0.05 as compared with controls.

‡P < 0.001 as compared with patients with pressure overload.

§P < 0.05 as compared with patients with pressure overload.

Measurement of natriuretic peptides

Serum samples were frozen at − 20°C. Serum concentrations of N-terminal proatriopeptide (ANPN) were measured by immunofluorometric assay. The reagents were from Medix Biochemica (Espoo, Finland) and instruments by Delfia Research Fluorometer (Wallac, Turku, Finland). Serum concentrations of N-terminal pro-brain natriuretic peptide (NT-proBNP) were measured by the electrochemiluminometric method. The reagent kit was from Roche (Mannheim, Germany), and the samples were analyzed at Limbach Laboratory (Heidelberg, Germany). The upper limit for the assays for NT-proBNP was 35,000 ng/l. In statistical analysis, an NT-proBNP concentration of 35,001 ng/l served for values higher than 35,000 ng/l.

Immunoassay for detection of cTnI

Serum samples for analysis of cTnI levels were frozen at − 70°C and concentrations of cTnI were determined according to manufacturer's instructions using the second generation Innotrac Aio!^™ Troponin I kit and the fully automated Innotrac Aio!^™ immunoanalyzer (Radiometer/Innotrac Diagnostics Oy, Turku, Finland). This assay is designed to suffer minimally from the presence of cTnI- specific cTnAAbs, (11) which cause negative interference in many commercial cTnI assays (12). According to the manufacturer, its analytical sensitivity is typically less than 0.01 μg/L. All samples were analyzed in duplicate and the cTnI results calculated from mean signals with MultiCalc Software (Perkin-Elmer/Wallac, Turku, Finland).

Immunoassay for detection of cardiac troponin specific autoantibodies (cTnAAb)

Human cTnAAbs were determined essentially as described earlier (13). Samples with an ITC-specific signal of 100 counts or higher were regarded as cTnAAb-positive if Student's T-test gave a P value of less than 0.05 in comparison of signals obtained from the same serum with and without added ITC were compared.

Statistical analysis

Analyses were performed with the PASW Statistics version 18 for Windows (SPSS Inc., Chicago, IL). For variables derived from blood samples, median and range were calculated. Because distribution of parameters tested by Kolmogorov-Smirnov´s goodness-of-fit test was not normal, the Mann- Whitney test was used for statistical analysis between groups, and the Wilcoxon Signed-Rank Test for analysis within groups. Correlations between serum levels of cTnI and natriuretic peptides were measured with Spearman´s correlation coefficient. The level of significance was P < 0.05.

Results

cTnI and natriuretic peptides

In the healthy controls, only one newborn had a positive cTnI value.

In the subgroup with volume overload, only one 5-year-old child with an ASD had a positive cTnI value. In the subgroup with pressure overload, serum levels of cTnI were positive in 12 children: 6/40 (15%) with CoA, 4/10 (40%) with AS, and 2/10 (20%) with PS (Table III). Levels of natriuretic peptides in patients were divided into subgroups according to loading conditions (volume or pressure overload) and in controls are in Table II.

Table III. Comparison of children with pressure overload and baseline positive or negative levels of cardiac troponin I. Values are median (range).

	TnI +	TnI +	P
N	12	48
age (years)	0.0 (0.0–13.0)	3.5 (0.0–19.0)	< 0.001
weight (kg)	3.7 (2.7–43.9)	15.5 (2.7–73)	< 0.001
ANPN (nmol/l)	11.61 (0.53–36.00)	0.53 (0.10–11.00)	< 0.001
NT -proBNP (ng/l)	35001 (53–35001)	103 (4–19527)	< 0.001
stenosis gradient (mm Hg)	54 (23–146)	33 (0–110)	0.019

ANPN = N-terminal proatriopeptide, NT-proBNP = N-terminal pro-brain natriuretic peptide.

Further analysis of the pressure overload subgroup revealed that the children with positive cTnI were significantly younger than those with negative cTnI 0.16 (0.00–156.00) months vs. 42.02 (0.00–228.0) months, respectively, (P < 0.001), altogether 14/60 patients were less than one-month old, ten of them cTnI positive and four cTnI negative. Serum levels of natriuretic peptides were higher among patients with positive cTnI (Table III) but there was no correlation between cTnI and natriuretic peptide levels. The pressure gradients across the stenotic valves in patients with aortic or pulmonary valve stenosis, and the arm to leg blood pressure gradients in patients with coarctation of the aorta were higher in patients with positive cTnI than in those with negative cTnI (Table III).

At the follow-up visit 6 months after catheter or surgical intervention, of all 64 children examined, only one child with CoA was cTnI positive, all others were cTnI-negative. Of the 13 children with positive serum levels of cTnI at baseline, seven underwent follow-up measurements, all of them negative.

Autoantibodies

None of our 74 healthy controls had autoantibodies to cTn. At baseline, only one patient with PDA had autoantibodies against cTn but with a negative serum level of cTnI. During the 6-month follow-up, none of our patients had developed cTnAAbs.

Discussion

In this study, we prospectively evaluated the incidence of cTnI and autoantibodies against it before and 6 months after interventions in patients with congenital cardiac defects and in healthy controls.

In our study, only one healthy baby had a positive cTnI. In some studies, cTn levels have been variable and in healthy infants sometimes even higher than adult levels (14,15). Levels of cTnI may increase after birth and decrease thereafter (16). No data are available regarding cTnI levels in untreated patients with pure pressure overload. In our study, children with positive cTnI levels were younger than those with negative levels. Most of them were less than one-month old and had congenital heart defects with severe pressure overload. The elevated cTnI levels may be partly due to younger age of these children. However, the pressure gradients were higher in cTnI positive than cTnI negative children. Unfortunately, the number of children aged less than one month in the control group is not sufficient to allow comparisons with children with pressure overload and positive cTnI.

After percutaneous closure of ASD, levels of cTnI have increased more than after a diagnostic catheterization procedure without defect closure. The highest increase has been noticeable soon after the closure (17). Additionally, after surgical ASD closure, levels of cTnI have been higher than after percutaneous closure, and increased levels have been measured for up to one week after surgery (18). In another study, cTnI levels after interventional catheterization have been higher than after diagnostic cardiac catheterization. Unfortunately, in that study, patients with baseline levels above the normal range were excluded (19).

We wished to study the levels of cTnI in patients with various types of loading conditions, specifically before treatment and six months thereafter. At baseline, most of the children with positive serum levels of cTnI were neonates with CoA, AS, or PS, whereas in only one child with ASD, the cTnI level was positive. Pressure gradients across the coarctation site or stenotic aortic or pulmonary valve were higher in patients with pressure overload and positive cTnI than in those with negative cTnI. Similarly, in adult population, positive cTnI has been shown to be associated with more severe aortic stenosis (20). Positive levels of TnI in our study may have resulted from mild subendocardial ischemia and reversible membrane damage in pressure-loaded hypertrophied myocardium, not necessarily from cell death. Sugimoto et al have shown a correlation between cTnI and pulmonary pressure. Their cTnI levels were higher in patients with volume and pressure overload caused by left-to-right shunt than in control children (21). Unfortunately, the 6-month follow-up was completed in only 64/138 children including 7/13 with positive cTnI at baseline. In all but one of those patients, cTnI had normalized by the time of the 6-month follow-up after treatment, indicating cessation of volume and pressure overload.

In adults, cTnI levels have been shown to be higher in patients with heart failure (3,22). Proposed mechanisms leading to cTn release in heart failure are cardiomyocyte damage from inflammatory cytokines or oxidative stress, subendocardial ischemia, and apoptosis. Neurohumoral activation may play a role in cTn release (3). In neonates, apoptosis is a possible mechanism of TnI release (23). Our patients with positive cTnI levels were younger, and their pressure gradients and natriuretic peptide levels were higher than in patients with negative cTnI levels in the subgroup with pressure overload. This is in accordance with a reported finding that in an interventional cardiac catheterization group, cTnI levels correlate inversely with age and weight (19).

Because several assays are available for cTnI detection, comparison between studies is difficult. Numerous manufacturers currently market immunoassays for cTnI measurement, but results of these assays are not always interchangeable because of differences in antibody reactivity, calibration materials, and assay formats, and because of the pronounced heterogeneity of the cTnI molecule in the bloodstream. Use of a common calibrator has reduced the bias in measured cTnI values but has been unable to provide acceptable cTnI assay standardization (24). A similar problem has not been evident with TnT assays commercially available only from a single manufacturer. The assay we used has been designed to suffer minimally from the presence of cTn-specific cTnAAbs (11) which cause negative interference in many commercial cTnI assays (12). On the other hand, Eriksson et al have shown that autoantibodies to cTn can interfere with the measurement of serum levels of cTnI (11). It has recently been demonstrated that autoantibodies to cTns (cTnAAb), which are able to interfere with cTn detection, exist in a high proportion (5%–20%) of individuals (adults) with or without cardiac diseases (11,25,26). Whether the development of cTnAAbs exerts positive effects on cardiac function and remodeling, and whether they are beneficial for patient outcome, is still controversial.

Limitations to the study: The age range between the volume- and pressure-overloaded children is different. This is due to the differences at ages at which these lesions are typically treated. Also, the number of healthy controls aged less than one month is small, and therefore, the subgroup of control neonates is not large enough to allow comparisons with neonates with pressure overload.

Conclusion

cTnI release is more frequently associated with pressure than with volume overload in children with congenital heart defect. It seems to resolve in most children after treatment. In patients with pressure overload and positive cTnI levels, pressure gradients and natriuretic peptide levels are higher than in patients with negative cTnI levels. High-sensitivity troponin assays may provide an additional tool for detecting mild heart failure or subendocardial ischemia and for the follow-up of this group of children.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

This study was supported by grants from the special governmental subsidy for health sciences research, Tampere, Finland, and the Paavo Nurmi Foundation, Helsinki, Finland.