Biomarkers and early detection of late onset anthracycline-induced cardiotoxicity in children

Abstract

Background

The main strategy for minimizing anthracycline cardiotoxicity is early detection of high-risk patients.

Aim of the study

To investigate the role of cardiac biomarkers; cardiac troponin T (cTnT) and N-terminal probrain natriuretic peptide (NT-pro-BNP), and tissue Doppler imaging (TDI), as early predictors of chronic cardiotoxicity in survivors of acute leukemia.

Patients and methods

We carried a retrospective study on 50 asymptomatic survivors of acute leukemia who received anthracycline in their treatment protocols. All patients underwent blood sampling to determine the levels of NT-pro-BNP and cTnT along with conventional echocardiography and TDI.

Results

None had abnormal cTnT levels. About 20% had abnormal NT-pro-BNP levels. Diastolic dysfunction of the left ventricle was the most significant in conventional echocardiography. TDI was superior as it detected myocardial affection in 10% more than echo. TDI demonstrated global myocardial damage with significant aberrations in peak myocardial velocities and ratios.

Conclusions

NT-pro-BNP can be used as a sensitive cardiac biomarker in monitoring of anthracycline-induced cardiotoxicity. Follow up is essential to validate the role of NT-pro-BNP as an early marker for late onset anthracycline-induced cardiotoxicity. Tissue Doppler is marvelous as it could detect early cardiac dysfunction even in those with normal study by conventional echocardiography.

Keywords:

• Biomarkers
• Anthracycline
• Cardiotoxicity
• Leukemia
• Children

Introduction

The increased 5-year survival rates approaching 80% for the majority of pediatric malignancies have resulted in an increasing focus on the late effects of therapy and quality of life in the growing population of childhood cancer survivors.^1,2

Unfortunately, cardiotoxicity is one of the most serious chronic complications of cancer therapy. Mortality related to cardiac causes is 10-fold higher among childhood cancer survivors as compared with age-matched control subjects.³

Anthracycline chemotherapy and mediastinal and neck radiation are the most common causes of therapy-related cardiovascular complications in childhood cancer survivors.⁴ The early identification of patients at risk for cardiotoxicity is a primary goal for both cardiologists and oncologists, allowing for the planning of personalized anti-neoplastic therapeutic strategies, the support of cardiac function, and the monitoring of the progression of cardiac damage.⁵

To detect subclinical myocardial damage, time and expensive monitoring of cardiac functions is still recommended, during and after chemotherapy.^4,6,7 Nevertheless, most of the approaches commonly used in clinical practice echocardiographic left ventricular ejection fraction showed low diagnostic sensitivity and low predictive power in detecting subclinical myocardial injury.^4,6,8

Tissue Doppler imaging (TDI) has shown to be more sensitive to identify anthracycline-induced cardiomyopathy,^9,10 and aberration in TDI may be identified before abnormalities can be detected by conventional echocardiography.¹¹

There is a growing expectation for newer, non-invasive, cost-effective diagnostic tools for early identification of patients prone to developing drug-induced cardiotoxicity.¹² Use of easily detectable cardiac biomarkers in blood, such as cardiac troponins and cardiac natriuretic peptides, has been evaluated in an animal model and in some clinical studies.^13–15

In the current work we evaluated the role of the biochemical markers N-terminal proBrain natriuretic peptide (NT-pro-BNP), cardiac troponin T (cTnT), and the new imaging modality, TDI, in early detection of chronic cardiotoxicity in patients treated for acute leukemia.

Patients and methods

We carried out a retrospective study on 50 asymptomatic survivors of childhood leukemia, who visited the late effects, Clinic of Pediatric Oncology, Zagazig University Hospitals and received anthracycline as a part of their therapy for acute leukemia for more than 1 year age. Informed consents were obtained from the patients or their parents.

Our inclusion criteria were: (1) Patients at the time of evaluation should have no signs or symptoms of cardiac impairment and have stable general condition. (2) Normal hepatic and renal function tests.

Patients with history of cardiac disease and hypertension were excluded from the study.

All survivors underwent all the following:

1.	Complete history taking and thorough physical examination including weight, body surface area (BSA), and measurement of blood pressure.
2.	The cumulative dose of anthracyclines was calculated for each patient according to his BSA.
3.	Methods for evaluation of subclinical cardiotoxicity: (a) Conventional echocardiography: Detailed conventional echocardiography was performed at Pediatric Cardiology Unit, Zagazig University Hospitals. For all patients, standard measurements were: left ventricular posterior wall (LVPW) thickness at diastole (IVS), left ventricular dimensions at end-systole (LVES), left ventricular dimensions at end-diastole (LVED), ejection fraction (EF) estimated by cube method, and fraction shortening (SF). All measurements were compared with the normal values of LVPW, IVS, LVES, and LVED which were taken from the standard tables according to ages and BSA.16 An EF of less than 55% was considered abnormal and SF of less than 29% was considered abnormal.17 (b) TDI was obtained according to the methods described by Kapusta et al.18 In the parasternal, long-axis view measurements of peak myocardial velocities were made guided by color-coded TDI. Peak myocardial velocity during systole (S), early diastole (De), and late diastole (Da) were measured at the right and left ventricular sides of interventricular septum (IVSR and IVSL), respectively, and the endocardial and epicardial side of LVPW. The transmyocardial velocity gradients were defined as the velocity difference between the left and right sides of IVS and between the endocardial and epicardial sides of the LVPW. The transmyocardial gradients were calculated for each cardiac cycle phase (S, De, and Da) separately, in the apical four chamber view. The Doppler velocity–time sonography was recorded using single-gated TDI. Peak longitudinal myocardial velocities were assessed during S, De, and Da within basal, middle, and apical parts of the right ventricular wall (RVW) and left ventricular wall (LVW). Ratios of peak velocities in De/Da and De/S were calculated for all TDI measurements. (c) Biochemical analysis: Venous blood samples for assessment of biochemical makers were obtained from an indwelling catheter after 30 minutes of rest in supine position. The blood samples were withdrawn into chilled tubes containing EDTA. The whole blood was centrifuged; plasma was decanted, immediately frozen, and stored at −27C until assayed (within 6 months after sampling). Plasma concentration of biochemical makers was measured by electro-chemi-luminescence immunoassay on Elecsys 1010 analyzer according to the manufacturer's guidelines (Roche Diagnostics, Minneapolis, MN, USA). Concentration of biochemical makers diagnostic for cardiotoxicity of oncology treatment has not been established yet. In our study, values above the reference range based on a number of studies and recommended by the manufacturer were considered elevated and suggesting cardiac injury associated with the treatment. The abnormal level was defined as a cTnT level more than 0.010 ng/ml.19 For NT-pro-BNP normal values of children, we used the age-dependent reference values by Albert and his colleagues.20

Statistical analysis

Data were analyzed using Microsoft Office 2007 (Excel) and Statistical Package for Social Science (SPSS) version 19.0.0, SPSS Inc., Chicago, IL, USA. Data were summarized using the arithmetic mean, standard deviation (SD), screening test, Student t-test, Chi-square test (χ²), and Mann–Whitney U test.

Probability (P) value was considered for statistical significance if it was less than 0.05.

Results

Subject characteristics

Fifty asymptomatic survivors of childhood acute leukemia were included in our work. None of the survivors had history of acute cardiotoxicity following anthracycline dosage. Table 1 shows the characteristics of the study population.

Table 1. Characteristics of the study population

Number of the study population	50
Male gender	30 (60%)
Female gender	20 (40%)
Mean age at diagnosis ± SD in years (range)	8.4 ± 3.46 (3–15)
Mean age at follow up ± SD in years (range)	3.75 ± 2.25 (1.5–6)
Mean age at evaluation ± SD in years (range)	11.63 ± 2.17 (8–16)
CAD <150–300 mg/m^2	18 (36%)
CAD >300 mg/m^2	32 (64%)
Diagnosis
ALL	39 (78%)
AML	11 (22%)

CAD: cumulative anthracyclin dose; ALL: acute lymphoblastic leukemia; AML: acute myeloid leukemia.

Biomarkers and cumulative anthracycline dosage

None of the survivors showed elevated cTnT levels. Ten of the 50 survivors (20%) had abnormal NT-pro-BNP levels compared to sex and age appropriate norms. There was a significant relation between abnormal NT-pro-BNP levels and age at diagnosis, follow-up duration, and cumulative anthracycline dosage (Table 2).

Table 2. Relation of serum level of NT-pro-BNP and risk factors of cardiotoxicity in study population

	Abnormal NT-Pro-BNP, n = 10 (20%)	Normal NT-Pro-BNP, n = 40 (80%)	P value
Age at diagnosis (year)	4.7 ± 2.3	6.3 ± 51.7	0.016*
Age at follow up (year)	11.9 ± 2.3	11.5 ±2.2	0.665
Duration of follow up (year)	4.60 ± 1.17	2.60 ± 1.41	<0.001**
CAD (mg/m^2)	484.6 ± 133.3	344.5 ± 76.6	<0.001**
Sex (rank) Mann–Whitney	31.0 (sum of rank 310)	24.12 (sum of rank 965)	0.189
Sex (Chi-square)	Male = 17(42.5%)	Male = 7(70.0%)	0.123
	Female = 23(57.5%)	Female = 3(30.0%)

*Significant; **highly significant.

Echocardiographic parameters and biomarkers

Significant correlation between abnormal NT-pro-BNP and increased LVED, LVES indexed for BSA and FS (Table 3). Also, there was significant increase in serum level of NT-pro-BNP in relation to normal values for age in 13 (26%) patients who show abnormalities in TDI parameters (Table 4).

Table 3. NT-pro-BNP in relation to ECHO parameters in our survivors

	Abnormal NT-Pro-BNP, n = 10 (20%)	Normal NT-Pro-BNP, n = 40 (80%)	P value
EF	63.0 ± 68	66.33 ± 5.08	0.091
FS	30.6 ± 4.69	34.45 ± 5.05	0.012*
IVSD	5.6 ± 0.84	5.23 ± 1.14	0.345
LVPW	5.7 ± 0.483	5.70 ± 1.017	1.000
LVEDS	30.20 ± 4.8	27.4 ± 2.6	0.015*
LVEDD	48.40 ± 3.6	37.9 ± 2.0	<0.001**

EF: ejection fraction; FS: fraction shortening; IVSD: interventricular septum thickness at end diastole; LVPW: left ventricular posterior wall; LVEDS: left ventricular dimension during systole; LVEDD: left ventricular dimension during systole; *significant; **highly significant.

Table 4. Relation between serum levels of NT-pro-BNP and TDI parameters in group (II)

	Abnormal TDI, n = 13 (26%)	Normal TDI, n = 37 (74%)	P value
NT-PBNP	1127.70 ± 689.91	159.24 ± 59.33	0.000 (<0.001)**

**Highly significant.

Echocardiographic parameters and cumulative anthracycline dosage

Conventional M-mode echocardiographic

Table 5 shows the echocardiographic results of the 50 survivors when divided into two anthracycline dosage groups. LVED and LVES were significantly increased with increased cumulative anthracycline dosage. Reduced left ventricular systolic function, expressed by a decrease in EF and FS was not significantly related to increased cumulative anthracycline dosage.

Table 5. Relation of cumulative anthracycline doses (CAD) in (mg/m²) and ECHO parameters in study population

	CAD < 300, N = 19 (38%)	CAD > 300, N = 31 (62%)	P value
EF	67.37 ± 2.79	62.61 ± 6.54	0.089
FS	34.41 ± 2.63	33.19 ± 4.84	0.326
IVSD	5.46 ± 1.22	5.20 ± 1.01	0.430
LVPW	5.74 ± 0.45	5.68 ± 1.14	0.829
LVEDS	25.89 ± 2.31	29.23 ± 3.23	<0.001**
LVEDD	37.37 ± 0.96	41.71 ± 6.29	0.005**

EF: ejection fraction; FS: fraction shortening; IVSD: interventricular septum thickness at end diastole; LVPW: left ventricular posterior wall; LVEDS: left ventricular dimension during systole; LVEDD: left ventricular dimension during systole; **highly significant.

Tissue Doppler imaging

1.

The parasternal long-axis view: The results of the peak myocardial velocity measurements are shown in Table 6 .The myocardial velocities of the interventricular septum was significantly decreased during S and De in both the right and left ventricular sides but not in the transmyocardial side. The myocardial velocities of LVPW were significantly increased on the epicardial side during S, De, and Da and also on endocardial side during De. The transmyocardial velocity gradient was significantly decreased during De.

2.

The apical four-chamber view: The myocardial velocities of the RVW were significantly decreased in the apical part during S and both De and Da, and also decreased significantly in the middle part during S and De. The myocardial velocities of the IVS were significantly increased in the middle and basal parts during S and Da, but decreased in the apical and middle parts during De. The myocardial velocities of the LVPW were significantly decreased in the apical parts in S and De but increased in the middle and basal parts in De and Da (Table 7).

Table 6. Mean peak myocardial velocities (mm/s) in childhood cancer survivors in the parasternal long-axis view

	S	P value	De	P value	Da	P value	De/S	P value	De/Da	P value
IVSR	−0.90	<0.001**	+1.50	<0.001**	+0.94	NS	0.76	<0.001**	0.90	<0.001**
IVSL	−2.04	<0.001**	+2.00	<0.001**	+1.09	NS	1.04	<0.001**	0.91	<0.001**
TransIVS	−1.14	NS	−0.50	NS	−0.14	<0.001**	0.00	–	0.00	–
LVPWepi	+3.33	<0.001**	−7.00	<0.001**	−2.66	<0.001**	1.17	<0.001**	0.86	<0.001**
LVPWen	+4.71	NS	−9.43	<0.001**	−3.17	NS	2.06	<0.001**	1.54	<0.001**
TransLVPW	+1.39	NS	−3.00	NS	−1.03	NS	0.89	<0.001**	0.00	–

(−): Direction away from the transducer; (+): direction toward the transducer; NS: non-significant; IVSR: right ventricular side of interventricular septum; IVSL: left ventricular side of interventricular septum; LVPWepi: epicardial side of left ventricular posterior wall; LVPWen: endocardial side of left ventricular posterior wall; LVPW: left ventricular posterior wall; **highly significant.

Table 7. Mean peak myocardial velocities (mm/s) in the apical four chamber view, in our survivors

	S	P value	De	P value	Da	P value	De/S	P value	De/Da	P value
RVW api	+5.26	<0.001**	−6.26	<0.001**	−2.94	<0.001**	1.94	NS	1.21	NS
RVW mid	+8.20	<0.001**	−9.31	<0.001**	−5.06	NS	2.03	NS	0.90	<0.001**
RVW Basal	+12.86	NS	−11.71	<0.001**	−6.57	NS	1.06	<0.001**	1.09	<0.001**
IVS api	+5.14	NS	−5.29	<0.001**	−2.00	NS	1.56	<0.001**	1.27	<0.001**
IVS mid	+8.03	<0.001**	−8.43	<0.001**	−3.00	<0.001**	1.50	<0.001**	1.37	<0.001**
IVS basal	+9.63	<0.001**	−12.86	NS	−5.00	<0.001**	1.53	<0.001**	1.47	<0.001**
LVW api	+5.03	<0.001**	−6.57	<0.001**	−2.06	NS	1.36	<0.001**	0.90	<0.001**
LVW mid	+8.14	NS	−20.71	<0.001**	−7.57	<0.001**	1.94	<0.001**	1.97	NS
LVW basal	+8.86	NS	−23.14	<0.001**	−8.43	<0.001**	5.06	<0.001**	1.21	<0.001**

S: systole; De: early diastole; Da: late diastole; +: direction toward the transducer; −: direction away from the transducer; RVW: right ventricular free wall; api: apical part of; mid: middle part of; basal: basal part of; IVS: interventricular wall; LVW: left ventricular wall; **highly significant.

Discussion

Cardiotoxicity remains a major limitation of chemotherapy, strongly affecting the quality of life and the overall survival of cancer patients, regardless of their oncologic prognosis. The time elapsed from the end of cancer therapy to the beginning of heart failure therapy for chemotherapy-induced cardiac dysfunction is an important determinant of recovery. This highlights the need for a real-time diagnosis of cardiac injury. The current standard for monitoring cardiac function detects cardiotoxicity only when a functional impairment has already occurred, precluding any chance of preventing its development.²¹ The role of biomarkers in the early detection on anthracycline-induced cardiotoxicity in children was recently reviewed.²² In the current study the role of biomarkers in relation to anthracycline-induced cardiotoxicity was investigated in 50 asymptomatic survivors of childhood acute leukemia. None of the survivors in our study had elevated cTnT levels. Cardiac troponins reflect myocardial injury and are related to acute anthracycline-induced cardiotoxicity changes.²³ The clinical studies with cardiac troponins are limited and the results are ambiguous. In some studies the administration of anthracyclines did not cause any elevation in cardiac troponins.^23–25 In another study, cardiac troponins became positive after anthracycline treatment, correlated with the disease severity, and might predict subsequent major cardiac events during follow-up.^26–29 The absence of abnormal cTnT in our survivors suggests that cTnT does not play a role in the detection of late onset anthracycline-induced cardiotoxicity.

Natriuretic peptides are released by the myocardium in response to volume and pressure overload, but are also stimulated by cytokines and growth factors.³⁰

The applicability of natriuretic peptides (ANP, BNP, and NT-pro-BNP) as a marker of anthracycline-induced cardiotoxicity has been investigated in a few recent studies. The results suggested that natriuretic peptides may be of value in detecting clinical and subclinical cardiotoxicity after anthracycline,^6,31 as well as markers of cardiac toxicity after a high dose of chemotherapy and hematopoietic stem cell transplantation.^32–34 However, definitive clinical evaluation has been missing and natriuretic peptides have not been routinely used for monitoring of cardiotoxicity in clinical practice.

Our current study showed that abnormal NT-pro-BNP were detected in 20% of the asymptomatic survivors and abnormal NT-pro-BNP were significantly related to increasing cumulative dose of anthracycline, duration of follow up and younger age at diagnosis. That is in agreement with Mavinkurve-Groothuis et al.,²³ who reported abnormal level of NT-pro-BNP in 13% of 122 long-term survivors of childhood cancer.

An increase in left ventricular dimensions is a common abnormal echocardiographic finding in anthracycline-induced cardiotoxicity.²³ In agreement with our results, Mavinkurve-Groothuis et al.²³ found a significant relation between abnormal NT-pro-BNP level and increased LVEDD indexed by BSA but contrary to us they failed to demonstrate the correlation with systolic function.

In the current study we found significant correlation between abnormal NT-pro-BNP level and both increased LVESD and LVEDD indexed by BSA and FS which suggest its clinical usefulness in detection of anthracycline-induced cardiotoxicity and highlight its role as an early marker of cardiotoxicity induced by anthracycline.

Echocardiography is the most frequently used non-invasive method for evaluation of cardiac function including toxic effect of oncology treatment.³⁵ In the present study, subclinical cardiotoxicity was detected in 16% of our survivors in the form of increase of left ventricular dimension and EF. A review of anthracycline-related subclinical cardiotoxicity noted a frequency varying from 0 to 57% in 25 studies with doses of anthracyclines varying from 45 to 1275 mg/m².³⁶ One of the main risk factors for anthracycline cardiotoxicity is high cumulative dose which is associated with a higher incidence of subclinical dysfunction.³⁷ The analysis of echocardiographic parameters in relation to cumulative anthracycline dosage showed that LVES and LVED significantly increase with the increase of cumulative anthracycline dose. Similar results were reported by Horacek et al.³⁸ who found that the incidence of left ventricular systolic and diastolic dysfunction advances with increasing cumulative doses of anthracyclines. Unfortunately, Lipshultz et al.³⁹ revealed that even patients who received cumulative doses as low as 45 mg/m² eventually experienced cardiac abnormalities, including significantly reduced left ventricular mass and dimensions. Even low doses are problematic if the goal is to avoid late cardiotoxicity. Although most studies have used echocardiography to make diagnosis of ventricular dysfunction, the assessment methods and threshold of determining abnormality have been variable.^40,41

TDI has shown to be sensitive to identify anthracycline-induced cardiomyopathy and aberration in TDI parameters may be identified before abnormalities can be detected by conventional echocardiography.⁴² In our study, we found 26% of our survivors had global myocardial damage in the form of significant aberrations in peak myocardial velocities and in the ratios between velocities in De/Da Abnormal myocardial velocities were found in all cardiac cycles (S, De, and Da). Thus TDI aberration was detected in instructions that appeared normal by M-mode echocardiography, and might precede structural changes.^42,43 Similar results were reported by Rathe et al.⁴⁴ Aberration in TDI parameters has shown to be highly sensitive to identify anthracycline-induced cardiomyopathy.⁴²

In conclusion, measurement of NT-pro-BNP can be used as a sensitive cardiac biomarker in monitoring of anthracycline induced-cardiotoxicity in survivors of acute leukemia. However, follow up of these survivors is essential to answer the question whether NT-pro-BNP is an early marker for late onset anthracycline-induced cardiotoxicity.

TDI appears to be an interesting method of surveillance for the growing number of survivors as it could detect cardiac dysfunction earlier and in structures that appear normal by conventional echocardiography.