International Journal of Cardiology 195 (2015) 7–14

Contents lists available at ScienceDirect

International Journal of Cardiology

j ourna l homepage: www.e lsev ie r .com/ locate / i j ca rd

High sensitivity troponin T in adult congenital heart disease☆

Justyna Rybicka a,⁎, Piotr Dobrowolski a, Magdalena Lipczyńska a, Ewa Kowalik a, Anna Klisiewicz a,Piotr Hoffman a, Piotr Szymański ba Adult Congenital Heart Disease Department, Institute of Cardiology, Warsaw, Polandb Acquired Valve Disease Department, Institute of Cardiology, Warsaw, Poland

☆ This author takes responsibility for all aspects of the rof the data presented and their discussed interpretation.⁎ Corresponding author at: Adult Congenital Heart D

Cardiology, Alpejska 42, 04-628 Warsaw, Poland.E-mail address: jrybicka@ikard.pl (J. Rybicka).

http://dx.doi.org/10.1016/j.ijcard.2015.05.0770167-5273/© 2015 Published by Elsevier Ireland Ltd.

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 15 January 2015Received in revised form 21 April 2015Accepted 14 May 2015Available online 15 May 2015

Keywords:High sensitivity troponin TAdult congenital heart diseaseACHDVentricular dysfunction

Background:High sensitivity troponin T (hsTnT) assays enable us to detect chronic heart failure (CHF). Adult con-genital heart disease (ACHD) patients are classified as being in at least stage B of CHF. The purpose of the studywas to assess hsTnT levels in ACHD patients and determine its clinical significance.Methods: This is a prospective cross-sectional study. We assessed hsTnT in 131 ACHD patients and in 30 healthycontrols. All ACHD patients underwent routine clinical and echocardiographic evaluation and had hsTnT and N-terminal brain natriuretic peptide (NT-pro-BNP) level measurements.Results: The cut-off value defining an abnormal hsTnT level was established as N0.005 ng/mL. 35.1% (n= 46) ofACHD patients had abnormal hsTnT compared to 6.7% (n=2) of healthy controls (p= 0.002). The prevalence ofelevated hsTnT did not differ between simple and complex and between non-cyanotic and cyanotic congenitalheart disease (CHD). The sensitivity and specificity of hsTnT for the detection of moderate or severe (significant)

systemic ventricular dysfunction was 78.6% and 69.8%, respectively (OR 8.49; CI 95% 2.23–32.30; p b 0,0001)whereas for significant pulmonary ventricular dysfunction it was 66.7% and 68.2%, respectively (OR 4.29; CI95% 1.56–11.79; p = 0.003). In multivariate logistic regression models elevated hsTnT, but not NT-pro-BNP,was independently associatedwith both significant systemic ventricular dysfunction (p=0.004) and significantpulmonary ventricular dysfunction (p = 0.011).Conclusions: A troponin leak is observed in a substantial number of ACHD patients and is associated with signif-icant systemic and pulmonary ventricular impairment. Compared to NT-pro-BNP, hsTnT is a more specific inde-pendent predictor of ventricular dysfunction in ACHD.

© 2015 Published by Elsevier Ireland Ltd.

1. Introduction

The development of a high sensitivity test for the detection of cardiactroponins has enabled us to detect them not only in the setting of acutemyocardial necrosis but also in chronic myocardial injury when minutequantities of troponin are released by the overloaded and/or necroticmyocardial cells [1]. Elevated troponin levels (≥0.01 ng/mL) are detect-ed in 10% of patients with chronic heart failure and their presence is as-sociatedwith amore severe disease and aworse outcome [1,2]. Patientswith congenital heart disease (CHD) are by definition classified as hav-ing at least stage B of chronic heart failure (CHF) and, even if apparentlyasymptomatic, as a rule have amuch lower exercise tolerance than their

eliability and freedom from bias

isease Department, Institute of

counterparts [3,4]. The risk of developing symptomatic CHF throughoutpatients' lifetime is very high especially in those with complex CHD.These patients are often treated with only palliative surgical proceduresand subsequently their myocardial muscle is continuously submitted tochronic pressure and/or volume overload. In patients with cyanoticheart disease, myocytes are also subjected to continuous ischemia be-cause of a low blood oxygen saturation. It can be assumed that thesemechanisms may lead to natriuretic peptide and troponin releasefrom stressed cardiomyocytes. However, while we have data on thechronic troponin release in classic heart failure [1,2], the data on theprevalence of a chronic troponin leak, in particular the minute quanti-ties detected with high sensitivity assays and their clinical significancein adults with congenital heart disease, are very limited [5–7]. There-fore,we assessed troponin releasewith high sensitivity assays in a seriesof patients with congenital heart disease and hypothesized that abnor-mal cardiac anatomy and exposure of themyocardium to an altered vol-ume and pressure load as well as to decreased oxygen delivery incyanotic lesions lead to myocardial damage and to a microscopic tropo-nin leak. NT-pro-BNP levels served as a benchmark [8,9].

8 J. Rybicka et al. / International Journal of Cardiology 195 (2015) 7–14

2. Methods

2.1. Patients

This is a prospective cross-sectional study. 131 adult congenitalheart disease (ACHD) patients, both with simple and complex lesions(with the exception of patients with an isolated bicuspid aortic valve)were consecutively recruited from an ambulatory care setting andfrom the inpatient ward of the tertiary care adult congenital heart dis-ease center. None of the patients were hospitalized within the previoussix months and their management was stable within a minimum of3 months before inclusion into the study. The exclusion criteria wereas follows: diagnosed coronary artery disease, pulmonary embolism,myocarditis and/or pericarditis, infections, neoplasms, primary muscleand neurovascular system disorders, systemic connective tissue dis-eases, serum creatinine N 200 μmol/L and elevated liver transaminases(N3 fold upper limits of normal). The control group was comprised of30 healthy staff members at the same institution.

All patients underwent routine clinical evaluation. Blood sampleswere collected to determine full blood count, hsCRP, creatinine, sodi-um and liver transaminases as well as hsTnT and NT-pro-BNP levels.All patients with CHD were given chest X-rays and transthoracicechocardiography.

2.2. Measurement of hsTnT and NT-pro-BNP

Blood was collected in tubes containing EDTA. hsTnT was measuredby the chemiluminescence and photon emission method (Cobas e 601analyzer, Roche Diagnostics, Germany). The lower detection limit ofthe highly sensitive TnT assay was 0.003 ng/mL. In the population of545 healthy subjects, the 99th percentile was 0.014 ng/mL and the95% interassay coefficient was 0.012–0.025 ng/mL. The interassay coef-ficient of variation was ≤10% at 0.013 ng/mL. NT-pro-BNP was deter-mined using immunoradiometric assay (Cobas e 601 analyzer, RocheDiagnostics, Germany).

2.3. Assessment of other clinical variables

A postero-anterior chest X-ray was obtained at 120 kVp using digi-tally processed technology (Philips) and the cardiothoracic ratio (CTR)was calculated. Central cyanosis associated with congenital heart dis-ease was defined as bluish discoloration and an oxygen saturation ofless than 90% in the absence of severe pulmonary disease. Oxygen satu-ration was measured at rest using transcutaneous pulse oximetry(Wellch Allyn Propaq LT, 80ZLTN, USA).

2.4. Echocardiography

All patients underwent a complete transthoracic echocardiographicstudy (GE Medical Systems Vivid 6, 7 or 9). Studies were performedby a group of cardiologists experienced in congenital heart disease. Forthe assessment of basic cardiac anatomy, the segmental approach wasapplied. Systemic and pulmonary ventricular function is assessed qualita-tively as normal, mildly, moderately or severely impaired. Classificationwasmade by an experienced observer; it was based on the integrated as-sessment of either the ventricular ejection fraction depending on anato-my with Simpson's single or biplane method (in the case of systemicleft ventricles and selected univentricular hearts), and/or fractional areachange (in the case of systemic and subpulmonary right ventricles), aswell as the long axis function measured by tissue Doppler velocitiesand/or mitral and/or tricuspid annular plane systolic excursion or solelyqualitatively, in the case of ventricles with severely distorted anatomy(where standard methods are not considered applicable).

2.5. Statistical analysis

Continuous variables are presented as the median and IQR and fre-quency as a percentage. Categorical variables are compared using thePearson χ2 test, whereas the differences between continuous variables,which had a non-normal distribution, are determined by means of theKruskal–Wallis test and then by theGames–Howell post-hoc test, as ap-propriate. The value of p b 0.05 is considered statistically significant.Correlations between variables were investigated using Pearson'scorrelation.

Troponin levels are compared in subsets of patients with normalNew York Heart Association (NYHA) class (NYHA class I) and impaired(NYHA class II or more) exercise tolerance, simple and complex lesions,cyanotic vs non-cyanotic lesions as well as impaired vs preserved func-tion of the systemic and pulmonary ventricle.

Receiver operating characteristic (ROC) analyses are used to assessthe predictive accuracy of hsTnT for detecting complex CHD, cyanoticlesions, impaired physical capacity (NYHA class II or III), pulmonaryarterial hypertension as well as systemic and pulmonary ventriculardysfunction. Comparison between areas under the curve (AUCs) is per-formed using the DeLong method. The cut-off value defining abnormalhsTnT levels in our population was determined on the basis of theAUC analysis for the identification of moderately to severely impairedsystemic ventricular dysfunction and established as N0.005 ng/mL. Itwas characterized by 78.6% sensitivity and 69.8% specificity (Fig. 1).

In order to estimate the predictive value of hsTnT and NT-pro-BNP,univariate andmultivariate logistic regressionmodels are used. Statisti-cal analyses are performed with commercially available computer soft-ware PASWStatistics 20 (SPSS Inc., Chicago, IL, USA) andMedCalc 12.7.0(MedCalc Software, Ostend, Belgium).

The study protocol conforms to the ethical guidelines of the 1975Declaration of Helsinki. It was approved by the Institutional ReviewBoard and the Ethical Committee of the Institute of Cardiology inWarsaw. Written informed consent was obtained from all patientsand members of the control group.

3. Results

131 ACHD patients and 30 healthy age and sex matched controlswere included in the study. Table 1 shows clinical characteristics ofthe study group as well as hsTnT and NT-pro-BNP levels. The hsTnTlevels in ACHD patients ranged from 0.003 to 0.085 ng/mL (IQR0.003–0.009) and from 0.003 to 0.01 ng/mL in healthy controls. NT-pro-BNP levels in ACHD patients ranged from 22.0 to 15,640.0 pg/mLand from 8.6 to 123.6 pg/mL in healthy controls; 35.1% (n = 46) ofACHD patients had an abnormal hsTnT level (N0.005 ng/mL) comparedto 6.7% (n = 2) of healthy controls (p = 0.002); 24.4% (n = 32) ofpatients and 3% (n = 1) of controls had hsTnT values of ≥0.01 ng/mL(p= 0.01). 77.9% (n= 102) of ACHD patients and none of the controlshad an elevated NT-pro-BNP level (N125 pg/mL) (p b 0.0001). For acomparison of the four groups identified on the basis of hsTnT andNT-pro-BNP elevation see Table 2. 3.8% (n= 5) of patients had atrial fi-brillation. NT-pro-BNP levels did not differ between patients with sinusrhythm and atrial fibrillation [802.9 (22.0–15,640.0) vs 1496.0 (538.0–3231.0) pg/mL, p = 0.47]. Also, hsTnT levels did not significantly differbetween patients with sinus rhythm and atrial fibrillation [0.007(0.003–0.085) vs 0.009 (0.004–0.015) ng/mL, p = 0.59]. Furthermore,no significant differences in the prevalence of elevated NT-pro-BNPand hsTnT are noted between the groups with sinus rhythm and atrialfibrillation (p = 0.23 and p = 0.17, respectively). Both hsTnT and NT-pro-BNP levels have a significant positive correlation with age (r =0.336, p b 0.0001 and r = 0.280, p = 0.001, respectively), creatininelevels (r = 0.322, p b 0.0001 and r = 0.377, p b 0.0001, respectively)and cardiothoracic ratio (r = 0.192, p = 0.049 and r = 0.507,p b 0.0001). Compared with patients with normal hsTnT, patientswith elevated hsTnT had higher NT-pro-BNP [214.6 (117.9–630.4) vs

Fig. 1. ROC curve for identification of moderately to severely impaired systemic ventricular dysfunction. Comparison between hsTnT (dotted line) and NT-pro-BNP (solid line). For hsTnTAUC = 0.733; for NT-pro-BNP AUC = 0.838, p = 0.136. ROC = receiver-operating characteristic; AUC = area under the receiver-operating characteristic curve.

9J. Rybicka et al. / International Journal of Cardiology 195 (2015) 7–14

878.9 (393.4–1632) pg/mL, p b 0.001] and hsCRP [0.13 (0.07–0.29) vs0.18 (0.09–0.48) mg/dL, p = 0.07].

3.1. hsTnT and NT-pro-BNP and NYHA class

There was a stepwise, albeit insignificant, increase in the proportionof patients with elevated hsTnT in groups with NYHA class I, II and III(p = 0.30) (Fig. 2). NT-pro-BNP levels are significantly higher inNYHA class III than in NYHA class I patients [481.7 (195.1–1288.3) vs164.2 (51.8–645.6) pg/mL, p = 0.031]. In all NYHA classes NT-pro-BNP is significantly more frequently elevated than hsTnT (Fig. 2).

3.2. hsTnT and NT-pro-BNP and anatomy

No significant differences in the prevalence of elevated hsTnT arenoted between the groups with simple and complex CHD (p = 0.66).This also pertains to elevated NT-pro-BNP levels (p = 0.46) (Fig. 3).However, both in simple and complex CHD NT-pro-BNP is significantlymore frequently elevated than hsTnT (Fig. 3).

3.3. hsTnT and NT-pro-BNP and cyanosis

Hemoglobin concentration was significantly higher in cyanotic pa-tients compared to non-cyanotic [18.4 (10.2–24.5) vs 14.9 (10.5–20.9) g/L, p b 0.0001]. There was also strong significant negative corre-lation between hemoglobin concentration and saturation (r=−0.606,p b 0.0001). No significant differences in the prevalence of elevatedhsTnT and NT-pro-BNP are noted between the groups with non-cyanotic and cyanotic CHD (p = 0.80 and p = 0.18, respectively). Asin previous subgroups, NT-pro-BNP is significantly more frequently ele-vated than hsTnT in non-cyanotic and cyanotic patients (Fig. 4).

3.4. hsTnT and NT-pro-BNP and pulmonary arterial hypertension

Pulmonary arterial hypertension (PAH) was diagnosed when meanpulmonary arterial pressure was 25 mm Hg or more at rest as assessed

by heart catheterization, according to the ESC/ERS Guidelines for the di-agnosis and treatment of pulmonary hypertension, 2009. 27 out of 35(77%) patients with severe pulmonary hypertension received therapywith endothelin-1 receptor antagonist (bosentan). 37.1% (n = 13) ofpatients with PAH had elevated hsTnT level. 80% (n = 28) of patientswith PAH had elevated NT-pro-BNP level. No significant differences inhsTnT levels were found between the groups with PAH and withoutPAH [0.010 (0.003–0.085) vs 0.008 (0.003–0.072) ng/mL, p = 0.30].Also, no significant differences in NT-pro-BNP levels were shown be-tween the groups with PAH and without PAH [933.5 (34–5301) vs968.3 (22–15,640) pg/mL, p = 0.93]. Furthermore, no statistically sig-nificant differences are noted in the prevalence of elevated hsTnT be-tween the groups with PAH and without PAH (p = 0.46). This alsopertains to NT-pro-BNP (p = 0.46). As in other analyzed subgroups,NT-pro-BNP is more frequently elevated than hsTnT in patients withPAH and without PAH.

There are no significant differences in the number of patients withmoderately or severely impaired systemic ventricular function betweenthe groups with PAH and without PAH (p= 0.07). Moreover, no signif-icant differences in the prevalence of moderately or severely impairedpulmonary ventricular function are noted between the groups withPAH and without PAH (p = 0.30).

3.5. hsTnT and NT-pro-BNP and ventricular dysfunction

hsTnT levels are significantly more likely to be abnormal in patientswith moderately or severely impaired systemic and pulmonary ventric-ular function compared with patients with normal or mildly impairedsystemic and pulmonary ventricular function (p b 0.0001 and p =0.003, respectively) (Figs. 5 and 6). The same pattern was observedwith respect to the frequency of elevated NT-pro-BNP levels in patientswith moderately or severely impaired systemic ventricular functioncompared with patients with normal or mildly impaired systemic ven-tricular function (p = 0.034). However, there are no significant differ-ences in the number of patients with an elevated NT-pro-BNP levelbetween the groups with normal, mildly, moderately or severely

Table 1Clinical characteristics of study group.

Variables Congenital heart disease patients(n = 131)

Age, years 32 (23–48.5)Gender, male/female 64 (49%)/67 (51%)BSA (m2) 1.74 (1.6–1.9)NYHA functional class I/II/III/IV 26 (20%)/59 (45%)/46 (35%)/0Atrial fibrillation 5 (3.8%)hs Troponin T (ng/mL) 0.004 (0.003–0.009)NT-pro-BNP level (pg/mL) 369.4 (144.1–964.6)Creatinine (μmol/L) 74 (64–87)Cardiothoracic ratio 54 (48–60)Simple CHD 34

ASD 19Intervention: none/reparative 19/0

VSD 8Intervention: none/reparative 7/1

Other lesionsa 7Intervention: none/reparative 7/0

Complex CHD 97Tetralogy of Fallot 16Surgery: none/palliative/reparative 4/0/12

Systemic right ventricle 23Surgery: none/palliative/reparative 15/6/2

Single ventricle physiology 25Surgery: none/palliative 1/24

Other lesionsb 33Surgery: none/palliative/reparative 19/2/12

Cyanotic CHD 55 (42%)Pulmonary arterial hypertensionc 35 (27%)Moderate or severe SV dysfunctiond 14 (11%)Moderate or severe PV dysfunctione 21 (16%)

Values are expressed as median (IQR).BSA= body surface area; NYHA= New York Heart Association; CHD= congenital heartdisease; ASD = atrial septal defect; VSD = ventricular septal defect; SV = systemic ven-tricle; PV = pulmonary ventricle.

a Patent ductus arteriosus (PDA), Ebstein anomaly, aortopulmonary window.b Atrio-ventricular septal defect (AVSD), pulmonary atresia with ventricular septal de-

fect (PA+ VSD), transposition of the great arteries with ventricular septal defect and pul-monary stenosis (TGA + VSD + PS), atrial septal defect with partial anomalouspulmonary venous connection (ASD + PAPVC), ventricular septal defect with patentductus atreriosus (VSD + PDA), pulmonary stenosis with atrial septal defect (PS +ASD), double outlet right ventricle with ventricular septal defect and pulmonary stenosis(DORV + VSD + PS), truncus arteriosus (TA), pulmonary atresia with ventricular septaldefect (PA + VSD), coarctation of the aorta with patent ductus arteriosus and ventricularseptal defect (CoA + PDA + VSD).

c 14 (41.2%) patients with simple CHD had pulmonary arterial hypertension comparedto 21 (21.6%) patients with complex CHD (p = 0.027).

d 2 (5.9%) patients with simple CHD had moderately to severely impaired systemic ven-tricular function vs 12 (12.5%) with complex CHD (p = 0.29).

e 8 (23.5%) patients with simple CHD had moderately to severely impaired pulmonaryventricular function compared to 13 (17.3%) with complex CHD (p = 0.45).

10 J. Rybicka et al. / International Journal of Cardiology 195 (2015) 7–14

impaired pulmonary ventricular function (p = 0.18). Elevated NT-pro-BNP levels are significantly more frequently observed than elevatedhsTnT levels in patients with normal or mildly impaired systemic andpulmonary ventricular function (Figs. 5 and 6). The differences in thefrequency of abnormal NT-pro-BNP and hsTnT levels in patients withmoderately or severely impaired systemic and pulmonary ventricularfunction are non-significant (Figs. 5 and 6). 21.4% and 33.3% of patientshaving significant systemic and pulmonary ventricular dysfunction

Table 2Univariate and multivariate logistic regression models for moderately or severelyimpaired systemic ventricular function.

Variables Univariate model Multivariate model

Exp(B)

95% CI p Exp(B)

95% CI p

Age 1.02 0.98–1.05 0.37 0.97 0.93–1.01 0.219Gender 2.87 0.85–9.68 0.09 3.12 0.83–11.6 0.09NT-pro-BNP N 125 pg/mL – 0.00 1.00 – – 1.00hsTnT N 0.0055 ng/mL 8.49 2.23–32.3 0.002 9.64 2.07–44.9 0.004

Significant differences are shown in bold.

respectively do not have elevated hsTnT. There is no association be-tween NYHA class and systemic and pulmonary ventricular function(p = 0.375 and p = 0.635, respectively).

The sensitivity and specificity of hsTnT for the detection of signifi-cant systemic ventricular dysfunction (moderately to severely impairedfunction of the ventricle) is 78.6% and 69.8%, respectively (OR 8.49; CI95% 2.23–32.30; p b 0,0001) whereas pulmonary ventricular dysfunc-tion sensitivity and specificity was 66.7% and 68.2%, respectively (OR4.29; CI 95% 1.56–11.79; p = 0.003). The sensitivity and specificity ofNT-pro-BNP for the detection of systemic ventricular dysfunction is100% and 25%, respectively and for pulmonary ventricular dysfunctionit is 90.5 and 22.7%, respectively (OR 2.79; CI 95% 0.60–13.00; p =0.18). The combination of both markers increased sensitivity for the di-agnosis of significant ventricular dysfunction at the cost of lower speci-ficity—100% and 44.4% for the systemic ventricle; 100% and 41% for thepulmonary ventricle.

In a multivariate logistic regression model the factor independentlyassociated with significant systemic ventricular dysfunction is an ele-vated hsTnT level, but not an elevatedNT-pro-BNP level (Table 2). In an-other multivariate logistic regression model, the factor independentlyassociated with significant pulmonary ventricular dysfunction, is alsoan elevated hsTnT but not an elevated NT-pro-BNP level (Table 3).

4. Discussion

Previous studies have shown that patients with stable chronic heartfailure of various etiologies have a detectable hsTnT,which is associatedwith an adverse outcome [1,10–12]. Latini et al. [1] found that in thegroup of 4053 patients with stable chronic heart failure, 10.4% had anelevated troponin T level (≥0.01 ng/mL). In our series, a quarter ofpatients had hsTnT ≥ 0.01 ng/mL and one third had hsTnT levels ofN0.005 ng/mL. Hence, the frequency of elevated hsTnT levels in our pa-tients is more than double compared to the stable chronic heart failurepopulation. Comparisons with other ACHD populations are difficult toarrive at as few reports describe troponin T elevation in ACHD patientsand these studies include either very small groups of patients or thosewhose illnesswas detected early after a surgical or interventional proce-dure [3,13,14]. There are some limited data for children [4,5,12]. In astudy of adults with CHD and pulmonary arterial hypertension (PAH)(86% with Eisenmenger syndrome) elevated hsTnT levels (defined asN0.014 μg/L) were found in 26% of patients [7]. It has also been shownthat tachyarrhythmias can cause elevation of both troponin T and NT-pro-BNP [15,16]. Thus, we checked for these potential confounders. Inour study population atrial fibrillation was present in only 3.8% of pa-tients. Furthermore, we did not find differences in the prevalence of el-evated hsTnT and NT-pro-BNP between the groups with sinus rhythmand atrial fibrillation.

The present study is the first that evaluates hsTnT levels in a largegroup of ACHD patients.

Previous studies proved that in adult patients with CHD routinelyused heart failure indices such as CTR [17,18] also apply to the CHDpop-ulation as they relate to NYHA class and systemic ventricular functionand, notably, not to anatomy. Regarding biomarkers, in our studypatients with impaired physical capacity (NYHA class II and III) had sig-nificantly higher NT-pro-BNP than patients in NYHA class I. This corre-sponds well with the results of projects that addressed natriureticpeptide levels in chronic heart failure. In contrast, we did not show sta-tistically significant differences in the hsTnT level among NYHA classes.However, we did not demonstrate the relationship between NYHA classand ventricular function, neither (systemic nor pulmonary). According-ly, it has been demonstrated that in ACHD patients a cardiopulmonarytest is superior to theNYHA classification for the assessment of objectivephysical capacity [19].

Robust information derived from our study is that the hsTnT leveldiscriminates patients with ventricular dysfunction (suggesting therole of a chronic troponin leak in its development).

Fig. 2. Frequency of elevated hsTnT and NT-pro-BNP levels in adults with congenital heart disease according to the NYHA functional class and a comparison of the frequency of elevatedhsTnT and NT-pro-BNP levels among patients in NYHA class I, II and III.

11J. Rybicka et al. / International Journal of Cardiology 195 (2015) 7–14

Anatomy did not distinguish between patients with elevated andnon-elevated troponin T and NT-pro-BNP. However, 97% of patientswith simple CHDdid not undergo any palliative or reparative procedure.This may explain the lack of differences in the prevalence of elevatedhsTnT between patients with simple and complex CHD. On the otherhand, Bolger et al. [20] and Ross et al. [21] did not find a difference inthe prevalence of increased levels of biomarkers in groups of CHD pa-tients with various anatomic lesions.

Interestingly,we did not showdifferences in theprevalence of hsTnTelevation between cyanotic and non-cyanotic groups. Given this, wespeculate that chronic myocardial hypoxemia associated with cyanoticheart disease does not causemyocyte injury that would lead to a chron-ic troponin release.

Also, we did not show differences in hsTnT and NT-pro-BNP levelsbetween the groupswith PAH andwithout PAH. Furthermore, no statis-tically significant differences are noted in the prevalence of elevatedhsTnT and NT-pro-BNP between the groups with PAH and withoutPAH. However, Schuuring MJ et al. [7] showed that ACHD patientswith PAH and elevated hsTnT levels had significantly higher mortalityrate compared to patients with non-elevated hsTnT levels (62 vs 13%,p= 0.005). Troponin release in this groupmay be secondary to ventric-ular pressure overload. Thus, patients with PAH and elevated hsTnTlevel may require more aggressive therapy.

Fig. 3. Frequency of elevated hsTnT and NT-pro-BNP levels in adults with congenital heart diseaBNP level among patients with simple and complex congenital heart disease.

Bolger et al. [20] proved an association between biomarkers such asatrial natriuretic peptide (ANP), BNP, endothelin-1 and norepinephrineand systemic ventricular impairment. In addition, Latini et al. [1] foundthat patients with CHF who had elevated TnT (≥0.01 ng/mL) had moreadvanced ventricular impairment. We demonstrated statistically signif-icant differences in the prevalence of elevated hsTnT among ACHD pa-tients with normal or mildly impaired and moderately or severelyimpaired systemic and pulmonary ventricular function. Conversely, anelevated NT-pro-BNP level did not distinguish patients with significantpulmonary ventricular dysfunction.Notably, 21.4% and 33.3% of patientshaving significant systemic and pulmonary ventricular dysfunction re-spectively do not have elevated hsTnT. Hence, myocardial injury isclearly multifactorial.

The presence of troponin T in plasma reflects cardiac myocyte dam-age or death [22]. However, mechanisms of continuous troponin releasefrom themyocardium remain unclear [23,24]. This may be secondary toongoingmyocyte death,which seems to be independent of ischemic or-igin [25,26]. Another possible process is a leakage of troponins from thecytosol due to a stretch of myocyte and loss of membrane integrity [27].In contrast, BNP is synthesized and released by cardiac myocytes inresponse to cardiac volume and pressure overload [11,28]. Thesedisparities may explain the differences in the prevalence of elevatedhsTnT levels and NT-pro-BNP levels in various subgroups in our study,

se according to the anatomy and comparison of frequency of elevated hsTnT and NT-pro-

Fig. 4. Frequency of elevatedhsTnT andNT-pro-BNP levels in adultswith congenital heart disease according to thepresence or absence of cyanosis and comparison of frequency of elevatedhsTnT and NT-pro-BNP level among patients with non-cyanotic and cyanotic congenital heart disease.

12 J. Rybicka et al. / International Journal of Cardiology 195 (2015) 7–14

especially in patients with normal or mildly impaired ventricularfunction.

Presumably, in the majority of these patients, ventricular dysfunc-tion is secondary to volume or pressure overload and myocytes remainundamaged. It needs to be determinedwhether dysfunction of the ven-tricle is, in this case, reversible.

Similarly, it is unknown whether patients with ventricular dysfunc-tion and non-elevated hsTnT have better outcomes than thosewith tro-ponin elevation. The only preliminary prognostic data available concerna specific subgroup of patients with Eisenmenger's syndrome [7]. Ap-proximately one third of these patients had elevated hsTnT and pre-served ventricular function. Troponin release in this group may becaused by transient loss of membrane integrity and leakage of TnT.Whether patients with elevated hsTnT and preserved ventricular func-tion have worse outcomes than those without troponin elevation re-mains an open question.

Moreover, 75% and 77.3% of patients with preserved systemic andpulmonary ventricular function respectively have elevated NT-pro-BNP. Given this, there are manymore factors than just ventricular func-tion that are responsible for NT-pro-BNP elevation, whereas release ofTnT is more specifically associated with, at least moderate, ventriculardysfunction. We are the first report that hsTnT is more specific thanNT-pro-BNP for the detection of significant systemic and pulmonaryventricular dysfunction. NT-pro-BNP has a very low specificity. Notably,the combination of both markers increased sensitivity for the diagnosisof significant ventricular dysfunction but at the cost of lower specificity.

Fig. 5. Frequency of elevated hsTnT andNT-pro-BNP levels in adultswith congenital heart diseashsTnT and NT-pro-BNP level among patients with normal to mildly and moderately to severel

The prognostic value of combined (both hsTnT and NT-pro-BNP) eleva-tion vs isolated hsTnT or NT-pro-BNP elevation remains to be established.

The correlation between hsTnT and NT-pro-BNP and clinical vari-ables shows that older patients and those with higher creatinine arelikely to have myocardial injury and increased cardiac filling volumeand/or pressure. We also demonstrated an association between cardiacsize assessed by chest X-ray and biomarker release. Thisfinding is in linewith other studies evaluating the usefulness of markers for detectingmyocardial dysfunction. In a multivariate model hsTnT but not NT-pro-BNP predicted significant systemic and pulmonary ventriculardysfunction.

5. Study limitations

Some of the subgroups consisted of a small number of patients.Therefore, the statistical power of our study would be greater if itincluded a larger number of subjects. However, to the best of our knowl-edge this is the first study to investigate hsTnT levels and their signifi-cance in a heterogeneous ACHD population and it will form the basisfor further, more specific research. Due to the heterogeneity of cardiacanatomyand the lack of guidelines for the evaluation of systolic functionin patientswith CHD,we used a classification based on integrated quan-titative and/or qualitative assessment. It divided patients into fourgroups: those with normal, mildly, moderately and severely impairedventricular function. A further limitation of the study is that we did

e according to systemic ventricular function and a comparison of the frequency of elevatedy impaired systemic ventricular function.

Fig. 6. Frequency of elevated hsTnT andNT-pro-BNP levels in adultswith congenital heart disease according to pulmonary ventricular function and comparison of the frequency of elevatedhsTnT and NT-pro-BNP levels among patients with normal to mildly and moderately to severely impaired pulmonary ventricular function.

13J. Rybicka et al. / International Journal of Cardiology 195 (2015) 7–14

not perform serial measurements of hsTnT for risk stratification. How-ever, this will be the subject of our future research.

6. Conclusions

In light of our results, we support the notion that CHF is a complexsyndrome [29], and is also so in CHD patients. Accordingly, NT-pro-BNP release reflects the physical capacity of ACHD patients as deter-mined by means of an NYHA score, which is conditioned by volumeand/or pressure ventricular overload whereas hsTnT release reflectsmyocardial damage and therefore ventricular dysfunction. Thus, wepostulate that combined assessment of NT-pro-BNP and hsTnT may becomplementary and highly valuable in assessing the clinical status ofACHD patients. The present study, for the first time, provides evidencethat hsTnT can be used for the identification of significant systemicand pulmonary ventricular impairment. Given this, we plan to evaluatea predictive value of hsTnT for adverse outcomes in the studiedgroup—those hospitalized due to decompensated heart failure and theincreased risk of mortality. At present, the presence of hsTnT seems tobe useful in managing care for ACHD patients with complex lesions inan inpatient setting.

Conflict of interest

The authors report no relationships that could be construed as a con-flict of interest.

Table 3Univariate and multivariate logistic regression models for moderately or severelyimpaired pulmonary ventricular function.

Variables Univariate model Multivariate model

Exp(B)

95% CI p Exp(B)

95% CI p

Age 1.02 0.99–1.05 0.32 0.99 0.96–1.03 0.65Gender 1.53 0.59–3.99 0.39 1.32 0.47–3.72 0.60NT-pro-BNP N 125 pg/mL 2.79 0.60–13.03 0.19 2.42 0.45–12.97 0.30hsTnT N 0.0055 ng/mL 4.29 1.56–11.79 0.005 4.35 1.41–13.43 0.011

Significant differences are shown in bold.

Acknowledgments

This work was supported by the Polish Ministry of Science andHigher Education (grant number 2.23/VI/11).

References

[1] R. Latini, S. Masson, I.S. Anand, E. Missov, M. Carlson, T. Vago, L. Angelici, S. Barlera,G. Parrinello, A.P. Maggioni, G. Tognoni, J.N. Cohn, Prognostic value of very lowplasma concentrations of troponin T in patients with stable chronic heart failure,Circulation 116 (2007) 1242–1249.

[2] I. Tentzeris, R. Jarai, S. Farhan, T. Perkmann, M.A. Schwarz, G. Jakl, J. Wojta, K. Huber,Complementary role of copeptin and high-sensitivity troponin in predicting out-come in patients with stable chronic heart failure, Eur. J. Heart Fail. 13 (2011)726–733.

[3] A.P. Bolger, A.J. Coats, M.A. Gatzoulis, Congenital heart disease: The original heartfailure syndrome, Eur. Heart J. 24 (2003) 970–976.

[4] P.M. Fredriksen, G. Veldtman, S. Hechter, J. Therrien, A. Chen, M.A. Warsi, M.Freeman, P. Liu, S. Siu, E. Thaulow, G. Webb, Aerobic capacity in adults with variouscongenital heart diseases, Am. J. Cardiol. 87 (2001) 310–314.

[5] M.P.H. Bojan, M. Lilot, S. Vicca, P. Pouard, P. Vouhé, Early elevation of cardiac tropo-nin I is predictive of short-term outcome in neonates and infants with coronaryanomalies or reduced ventricular mass undergoing cardiac surgery, J. Thorac.Cardiovasc. Surg. 144 (2012) 1436–1444.

[6] M.O.K. Sugimoto, A. Kajihama, K. Nakau, H. Manabe, H. Kajino, Volume overload andpressure overload due to left-to-right shunt-induced myocardial injury. Evaluationusing a highly sensitive cardiac troponin-I assay in children with congenital heartdisease, Circ. J. 75 (2011) 2213–2219.

[7] M.J. Schuuring, A.C. van Riel, J.C. Vis, M.G. Duffels, J.P. van Straalen, S.M. Boekholdt,J.G. Tijssen, B.J. Mulder, B.J. Bouma, High-sensitivity troponin T is associated withpoor outcome in adults with pulmonary arterial hypertension due to congenitalheart disease, Congenit. Heart Dis. 8 (2013) 520–526.

[8] S.A. Hunt, W.T. Abraham, M.H. Chin, A.M. Feldman, G.S. Francis, T.G. Ganiats, M.Jessup, M.A. Konstam, D.M. Mancini, K. Michl, J.A. Oates, P.S. Rahko, M.A. Silver,L.W. Stevenson, C.W. Yancy, E.M. Antman, S.C. Smith Jr., C.D. Adams, J.L. Anderson,D.P. Faxon, V. Fuster, J.L. Halperin, L.F. Hiratzka, A.K. Jacobs, R. Nishimura, J.P.Ornato, R.L. Page, B. Riegel, ACC/AHA 2005 Guideline Update for the Diagnosis andManagement of Chronic Heart Failure in the Adult: a report of the American Collegeof Cardiology/American Heart Association Task Force on Practice Guidelines (Writ-ing Committee to update the 2001 Guidelines for the Evaluation and Managementof Heart Failure): developed in collaboration with the American College of ChestPhysicians and the International Society for Heart and Lung Transplantation: En-dorsed by the Heart Rhythm Society, Circulation 112 (2005) e154–e235.

[9] P.A. McCullough, R.M. Nowak, J. McCord, J.E. Hollander, H.C. Herrmann, P.G. Steg, P.Duc, A. Westheim, T. Omland, C.W. Knudsen, A.B. Storrow, W.T. Abraham, S. Lamba,A.H. Wu, A. Perez, P. Clopton, P. Krishnaswamy, R. Kazanegra, A.S. Maisel, B-type na-triuretic peptide and clinical judgment in emergency diagnosis of heart failure: anal-ysis from breathing not properly (BNP) multinational study, Circulation 106 (2002)416–422.

[10] C.H. Del Carlo, A.C. Pereira-Barretto, C. Cassaro-Strunz, Mdo R. Latorre, J.A. Ramires,Serial measure of cardiac troponin T levels for prediction of clinical events in decom-pensated heart failure, J. Card. Fail. 10 (2004) 43–48.

14 J. Rybicka et al. / International Journal of Cardiology 195 (2015) 7–14

[11] Y. Sato, R. Taniguchi, T. Makiyama, K. Nagai, H. Okada, T. Yamada, A. Matsumori, Y.Takatsu, Serum cardiac troponin T and plasma brain natriuretic peptide in patientswith cardiac decompensation, Heart 88 (2002) 647–648.

[12] J.S. Healey, R.F. Davies, S.J. Smith, R.A. Davies, D.S. Ooi, Prognostic use of cardiac tro-ponin T and troponin I in patients with heart failure, Can. J. Cardiol. 19 (2003)383–386.

[13] R. Taniguchi, Y. Sato, Y. Nishio, T. Kimura, T. Kita, Measurements of baseline andfollow-up concentrations of cardiac troponin-T and brain natriuretic peptide in pa-tients with heart failure from various etiologies, Heart Vessels 21 (2006) 344–349.

[14] K. Liesemer, T.C. Casper, K. Korgenski, S.C. Menon, Use andmisuse of serum troponinassays in pediatric practice, Am. J. Cardiol. 110 (2012) 284–289.

[15] Y. Otaki, T. Arimoto, H. Takahashi, S. Kadowaki, D. Ishigaki, T. Narumi, Y. Honda, T.Iwayama, S. Nishiyama, T. Shishido, T. Miyashita, T. Miyamoto, T. Watanabe, I.Kubota, Prognostic value of myocardial damage markers in patients with chronicheart failure with atrial fibrillation, Intern. Med. 53 (2014) 661–668.

[16] N. Ben Yedder, J.F. Roux, F.A. Paredes, Troponin elevation in supraventriculartachycardia: primary dependence on heart rate, Can. J. Cardiol. 27 (2011) 105–109.

[17] J.N. Cohn, G.R. Johnson, R. Shabetai, H. Loeb, F. Tristani, T. Rector, R. Smith, R.Fletcher, Ejection fraction, peak exercise oxygen consumption, cardiothoracicratio, ventricular arrhythmias, and plasma norepinephrine as determinants of prog-nosis in heart failure. The V-HEFT VA cooperative studies group, Circulation 87(1993) VI5–VI16.

[18] K. Dimopoulos, G. Giannakoulas, I. Bendayan, E. Liodakis, R. Petraco, G.P. Diller, M.F.Piepoli, L. Swan, M. Mullen, N. Best, P.A. Poole-Wilson, D.P. Francis, M.B. Rubens,M.A. Gatzoulis, Cardiothoracic ratio from postero-anterior chest radiographs: a sim-ple, reproducible and independent marker of disease severity and outcome in adultswith congenital heart disease, Int. J. Cardiol. 166 (2013) 453–457.

[19] Recommendations for exercise testing in chronic heart failure patients, Eur. Heart J.22 (2001) 37–45.

[20] A.P. Bolger, R. Sharma, W. Li, M. Leenarts, P.R. Kalra, M. Kemp, A.J. Coats, S.D. Anker,M.A. Gatzoulis, Neurohormonal activation and the chronic heart failure syndrome inadults with congenital heart disease, Circulation 106 (2002) 92–99.

[21] R.D. Ross, S.R. Daniels, D.C. Schwartz, D.W. Hannon, R. Shukla, S. Kaplan, Plasma nor-epinephrine levels in infants and children with congestive heart failure, Am. J.Cardiol. 59 (1987) 911–914.

[22] K. Setsuta, Y. Seino, T. Ogawa, M. Arao, Y. Miyatake, T. Takano, Use of cytosolic andmyofibril markers in the detection of ongoing myocardial damage in patients withchronic heart failure, Am. J. Med. 113 (2002) 717–722.

[23] J. Feng, B.J. Schaus, J.A. Fallavollita, T.C. Lee, J.M. Canty Jr., Preload induces troponin Idegradation independently of myocardial ischemia, Circulation 103 (2001)2035–2037.

[24] H.D. White, Pathobiology of troponin elevations: do elevations occur with myocar-dial ischemia as well as necrosis? J. Am. Coll. Cardiol. 57 (2011) 2406–2408.

[25] G. Olivetti, R. Abbi, F. Quaini, J. Kajstura, W. Cheng, J.A. Nitahara, E. Quaini, C. DiLoreto, C.A. Beltrami, S. Krajewski, J.C. Reed, P. Anversa, Apoptosis in the failinghuman heart, N. Engl. J. Med. 336 (1997) 1131–1141.

[26] J. Narula, P. Pandey, E. Arbustini, N. Haider, N. Narula, F.D. Kolodgie, B. Dal Bello, M.J.Semigran, A. Bielsa-Masdeu, G.W. Dec, S. Israels, M. Ballester, R. Virmani, S. Saxena,S. Kharbanda, Apoptosis in heart failure: release of cytochrome c frommitochondriaand activation of caspase-3 in human cardiomyopathy, Proc. Natl. Acad. Sci. U. S. A.96 (1999) 8144–8149.

[27] M.H. Hessel, D.E. Atsma, E.J. van der Valk, W.H. Bax, M.J. Schalij, A. van der Laarse,Release of cardiac troponin I from viable cardiomyocytes is mediated by integrinstimulation, Pflugers Arch. 455 (2008) 979–986.

[28] Y. Iwanaga, S. Miyazaki, Heart failure, chronic kidney disease, and biomarkers—anintegrated viewpoint, Circ. J. 74 (2010) 1274–1282.

[29] Y. Sato, T. Kita, Y. Takatsu, T. Kimura, Biochemical markers of myocyte injury in heartfailure, Heart 90 (2004) 1110–1113.