Early diagnostic value of high‐sensitivity cardiac troponin T for cancer treatment‐related cardiac dysfunction: a meta‐analysis

Abstract

Early diagnosis of cancer treatment‐related cardiac dysfunction (CTRCD) is important as cancer therapy increases the risk of cardiac dysfunction. High‐sensitivity cardiac troponin T (hs‐cTnT) is a highly specific marker of myocardial injury. However, its diagnostic value for CTRCD has not been systematically evaluated. This meta‐analysis aimed to evaluate whether hs‐cTnT could be used as an early diagnostic biomarker for CTRCD. We systematically surveyed PubMed, Embase, Cochrane Library, and Web of Science databases for studies of hs‐cTnT for the diagnosis of CTRCD before 1 April 2022. Patients of all ages and all cancer types who underwent echocardiographic left ventricular ejection fraction assessment and blood hs‐cTnT and received anticancer therapy (including chemotherapy, radiotherapy, targeted therapy, immune checkpoint inhibitors, and other treatments) were included in this study, resulting in a total of eight studies with 1294 patients. The occurrence of CTRCD was associated with elevated hs‐cTnT [sensitivity: 0.78, 95% confidence interval (CI): 0.64–0.88; specificity: 0.75, 95% CI: 0.59–0.86; area under the curve (AUC): 0.83, 95% CI: 0.80–0.86]. We further performed subgroup analysis and found that the AUC of hs‐cTnT elevation for the diagnosis of CTRCD increased from 0.83 to 0.90 (95% CI: 0.87–0.92) at 3–6 months, suggesting a higher early diagnostic value of hs‐cTnT compared with echocardiography for CTRCD. In terms of clinical applicability, the Fagan plot showed pre‐test and post‐test probabilities of 51% and 9%, respectively, indicating that hs‐cTnT testing can improve the accuracy of clinical diagnosis of CTRCD. However, it was not possible to determine the optimal cut‐off value for early diagnosis of CTRCD with hs‐cTnT. The Deeks funnel plot was largely symmetrical (P = 0.74); hence, publication bias was not observed. Hs‐cTnT allowed early CTRCD diagnosis at 3–6 months. However, further high‐quality research is needed to determine the optimal cut‐off value for early CTRCD diagnosis with this biomarker.

Introduction

The number of cancer survivors has continued to increase owing to advances in cancer diagnosis and treatment.¹ However, cancer treatment-related cardiovascular complications are a significant cause of death among cancer survivors, and treatments such as chemoradiotherapy for tumours can cause structural and functional damage to the myocardium, leading to cancer treatment-related cardiac dysfunction (CTRCD). Anthracyclines, trastuzumab, pertuzumab, immune checkpoint inhibitors (ICIs), and radiation therapy can all lead to CTRCD in cancer patients.^2,3 The 2002 Clinical Practice Guidelines in Cardio-Oncology updated the definition of CTRCD³ as follows: (i) left ventricular ejection fraction (LVEF) ≥ 50% with a relative decrease in overall longitudinal echocardiographic strain of >15% and/or a new increase in cardiac biomarkers; (ii) an LVEF decrease to 40–49% accompanied by a relative decrease in overall longitudinal echocardiographic strain of >15% and/or a decrease in LVEF < 10% and a new increase in cardiac biomarkers; or (iii) an LVEF decrease to ≤40%. CTRCD monitoring methods include electrocardiography, cardiac magnetic resonance imaging (MRI), nuclear imaging techniques, and serum markers in addition to echocardiography.^3–8 However, limitations exist among current detection methods. Therefore, early CTRCD detection is critical for the prevention of long-term cardiovascular morbidity in cancer survivors.

Serum biomarkers are important baseline risk assessment and diagnosis tools of CTRCD to guide cardioprotective therapy during anticancer treatment in cancer patients and to provide prognostic prediction.⁹ The 2016 European Society of Cardiology (ESC) position paper on cancer therapy and cardiovascular toxicity also recommends the use of high-sensitivity troponins to predict left ventricular (LV) dysfunction in patients treated with anthracyclines and/or trastuzumab.⁸ The 2020 European Society for Medical Oncology (ESMO) consensus states that measurement of high-sensitivity troponin at the baseline examination for oncology therapy may be useful in detecting or predicting cardiovascular toxicity, particularly cardiomyopathy and/or heart failure (HF).¹⁰ Recent Clinical Practice Guidelines in Cardio-Oncology suggested that myocardial markers, such as troponin and brain natriuretic peptide, are included as diagnostic indicators of CTRCD,³ suggesting that the diagnostic value of troponin for CTRCD has been gradually emphasized. Michel et al. analysed 61 trials with 5691 patients and showed that troponin had a negative predictive value of 93% for CTRCD, with a high exclusionary diagnostic value.¹¹ Additionally, several studies suggested that serum high-sensitivity troponin could be used as a marker for early monitoring of CTRCD in patients undergoing cancer chemoradiotherapy.^12–24 However, no systematic and comprehensive evaluation of the early diagnostic value of serum high-sensitivity cardiac troponin T (hs-cTnT) has been performed.

In this systematic review and meta-analysis, all studies that used hs-cTnT as a diagnostic index of CTRCD and LVEF as the diagnostic gold standard were included to assess the early diagnostic value of hs-cTnT in CTRCD.

Methods

This study was registered with PROSPERO (CRD42022297497) and followed the Cochrane Handbook for Systematic Reviews of Interventions and Preferred Reporting Items for a Systematic Review and Meta-Analysis of Diagnostic Test Accuracy Studies: the PRISMA-DTA Statement.^25,26

Study selection

The authors searched for studies on the role of hs-cTnT in CTRCD diagnosis in PubMed, Embase, Cochrane Library, and Web of Science databases from the database-establishment dates until 1 April 2022. The search terms were as follows: [cardiotoxicity (MeSH) OR cardiotoxicity OR cardiac toxicity OR cardiac toxicities OR toxicity, cardiac OR cardiovascular toxicity OR cardiovascular dysfunction OR cardiac dysfunction OR cardiac insufficiency] AND (High-Sensitivity Troponin OR high-sensitivity cardiac troponin OR hs-cTn). References of the included studies were also manually searched for potentially relevant studies. The two authors (L. X. F. and W. X.) performed all database searches, abstract screenings, and full-text reviews independently. References selected through database searches were imported into Endnote 20, and their titles and abstracts were independently reviewed after duplicates had been removed. Echocardiographic LVEF assessment and blood hs-cTnT examination were performed in patients of all age groups and all cancer types, who received chemotherapy, radiotherapy, targeted therapy, ICIs, and other treatments. During baseline examinations and follow-ups of oncology treatment, diagnostic sensitivity and specificity were clearly reported. Literature published in English was included, whereas reviews, case reports, letters, comments, and animal studies were excluded.

Data extractions and quality assessments

A standardized data extraction table was developed to extract data, including the names of first authors, publication years, countries, study designs, tumour types, treatment methods, patient sex, age, baseline LVEF and hs-cTnT values, hs-cTnT thresholds, testing time, true positive values, false positive values, false negative values, and true negative values. The two authors (L. X. F. and W. X.) independently extracted all data, and all discrepancies were resolved by Z. X. D.

The authors (L. X. F. and W. X.) independently assessed the quality of included studies using the Quality Assessment of Studies in Diagnostic Accuracy (QUADAS-2) tool.²⁷ For the Patient Selection domain, if all patients receiving tumour treatment (with any tumour types, treatment methods, ages, etc.) were consecutively or randomly included in a study, they were considered to be at a low risk of bias. If pre-specified thresholds were used and results were interpreted without the knowledge of LVEF results, the Index Test domain was considered to be at a low risk of bias. If LVEF values for all included patients were calculated by transthoracic two-dimensional echocardiography using the Simpson modified biplane method and were obtained without the knowledge of hs-cTnT results, the Reference Standard domain was considered to be at a low risk of bias. Regarding the risk of bias for the Flow and Timing domain, if hs-cTnT and LVEF were examined simultaneously in a single treatment cycle, they were treated as acceptable values. Any discrepancies in the quality evaluation process were resolved by a third reviewer (Z. X. D.).

Statistical analysis

The sensitivity and specificity of hs-cTnT were estimated using a random-effects model with 95% confidence interval (CI). Additionally, positive likelihood ratio (PLR), negative likelihood ratio (NLR), and diagnostic odds ratio (DOR) were calculated. The summary receiver operating characteristic (SROC) curve was constructed, and the area under the curve (AUC) was calculated. The following recommendations were provided for the interpretation of AUC values: low (0.5 ≤ AUC < 0.7), moderate (0.7 ≤ AUC < 0.9), or high (0.9 ≤ AUC ≤ 1) accuracy. Heterogeneity in combined studies was assessed using Cochrane Q and I² statistics. I² < 25% indicated no heterogeneity, 25% < I² < 50% indicated moderate heterogeneity, and I² > 50% indicated high heterogeneity among the results. If there was significant heterogeneity, a meta-regression analysis was performed to explore potential sources. Sensitivity analyses were conducted by analysing one article at a time. Publication bias was assessed using the Deeks funnel plot, and P < 0.05 was considered significant. Clinical applicability of the studies was evaluated by a Fagan diagram. Statistical analyses were performed by one of the authors (L. X. F.) in STATA 16.0 using the ‘MIDAS’ module and meta disc 1.4 software.

Results

Research features

A total of 3578 papers were retrieved, and 3570 ineligible papers were excluded (Figure 1). Eight studies included in this review were published between 2014 and 2021, with two from China, two from Japan, and four from the United States, Italy, Romania, and Germany. A total of 1294 patients were included in these articles, with 633 cases of breast cancer, 186 cases of lymphoma, 90 cases of upper gastrointestinal tract tumours, 41 cases of multiple myeloma, 36 cases of melanoma, 31 cases of ovarian cancer, 28 cases of neuroendocrine tumour, 27 cases of lung cancer, 18 cases of acute myeloid leukaemia, 15 cases of sarcoma, 3 cases of pleural mesothelioma, and 186 cases with no specific tumour type reported. The average patient age was 48–70 years, and 71.3% of them were women (922 cases). The study design included six prospective cohort studies and two retrospective studies. Most enrolled patients received anthracycline chemotherapy, some received trastuzumab monoclonal antibodies, and a minority of patients received ICIs or thoracic radiation therapy. Roche Diagnostics instruments were used for hs-cTnT diagnosis (all instruments from Roche Diagnostics had a range of 3–14 ng/L). Of the included studies, five studies used the Cobas instrument, one study used the Elecsys 2010 analyser, one study used ECLusys high-sensitivity Troponin T, and one study did not report the specific instrument and model. Cardiotoxicity was defined as an LVEF decrease in the range of ≥10% and <50%, measured by two-dimensional echocardiography after cancer therapy initiation, and the incidence rate of CTRCD ranged from 6.1% to 47%. Details of each study are presented in Table 1. Further details are provided in the supporting information.

[IMAGE OMITTED. SEE PDF]

Table 1 Basic characteristics of included studies

								Baseline
Source	Study design	Country of publication	Cancer type (%)	Treatment (%)	Age (years)	Female (%)	Cut-off (ng/L)	LVEF (%)	hs-cTnT (ng/L)	Study size (no.)	Detection time of hs-cTnT	CTRCD (%)	TP	FP	FN	TN
Demissei et al.,16 2020	Prospective	America	Breast cancer, 100	Doxorubicin, 61.6; trastuzumab, 22; doxorubicin + trastuzumab, 16.4	Mean: 48	100	14	Mean: 53	Mean: 3	170	2 months	18.2	19	52	12	87
Zhang et al.,17 2017	Retrospective	China	Large B-cell lymphoma, 100	CHOP/R-CHOP, 100	Mean (SD): 50 (12)	50	7.5	Mean (SD): 69 (12.4)	Mean (SD): 3 (3)	82	2–4 chemotherapy cycles	6.1	3	12	2	65
Bisoc et al.,18 2020	Prospective	Romania	Not provided	Doxorubicin, 100	Mean (SD): 56.6 (10.2)	60.3	8	Mean: 60.1	Not provided	68	3 months	22.1	10	17	5	36
Kitayama et al.,20 2017	Prospective	Japan	Breast cancer, 100	Anthracycline (epirubicin), 85; trastuzumab, 45; both agents, 30	No cardiotoxicity, mean (SD): 55 (2.0); cardiotoxicity, mean (SD): 57 (4.3)	100	7	No cardiotoxicity, mean (SD): 70 (1.0); cardiotoxicity, mean (SD): 73 (1.0)	No cardiotoxicity, mean (SD): 7 (1.7); cardiotoxicity, mean (SD): 5 (1.9)	40	3 months	10	4	0	0	36
Katsurada et al.,19 2014	Prospective	Japan	Breast cancer, 100	Radiation, 63; doxorubicin, 26; epirubicin, 73.6	Group N, mean (SD): 49 (7); Group R, mean (SD): 57 (9)	100	5.5	Group N, mean (SD): 68 (5.0); Group R, mean (SD): 71 (3.0)	Group N, mean: 3; Group R, mean: 3	19	6 months	47	7	2	2	8
Petricciuolo et al.,22 2021	Prospective	Italy	Lung cancer, 90; pleural mesothelioma, 10	Pembrolizumab, nivolumab, durvalumab, and atezolizumab, 100	Mean: 68	23	14	Mean: 57	Mean: 12	30	3 months	43	10	4	3	13
Kang et al.,23 2014	Prospective	China	Large B-cell non-Hodgkin's lymphoma, 100	R-CHOP, 100	Mean (SD): 53.9 (13.8)	45.3	>4 from baseline	Mean (SD): 65 (3.8)	Mean (SD): 1 (2)	75	The third chemotherapy cycle (21 days as a cycle)	18.7	11	22	3	39
Finke et al.,21 2021	Retrospective	Germany	Breast cancer, 49.9; upper GI tumour, 11.1; multiple myeloma, 5.1; melanoma, 4.4; ovarian cancer, 3.8; NET, 3.5; lymphoma, 3.6; AML, 2.2; sarcoma, 1.9; others, 14.6	Palliative, 21.5; adjuvant, 15.4; neoadjuvant, 29.3; terminated, 16.8; others, 17.0	Mean: 60	70.4	7	Mean: 60	Mean: 7	810	8.8 months	32.7	244	300	21	245

Quality assessment of included studies

As shown in Figure 2, two studies with no patient inclusion criteria were included, and the risk of bias in the Patient Selection domain was unclear. Risk of bias in the Reference Standard domain was also unclear because five studies did not provide blinded information on the interpretation of LVEF results. A high risk of bias was mainly observed in the domains of Index Test and Flow and Timing. In the Index Test domain, five studies with neither blinded information nor pre-determined thresholds were at a high risk of bias; two other studies without pre-determined thresholds were at a high risk of bias; and one study without blinded information had an unclear risk of bias. In the Flow and Timing domain, one study, which did not measure hs-cTnT and LVEF in the same treatment cycle and did not include all patients in the statistical analysis, was at a high risk of bias; one study, which included patients without the same gold standard assessment and did not include all patients in the statistical analysis, was at a high risk of bias; and two other studies, which did not include all patients in the statistical analysis, were also at a high risk of bias. One study had problems with its clinical applicability because it defined cardiotoxicity as an LVEF decrease of >5%, measured by two-dimensional echocardiography after anticancer therapy initiation.

[IMAGE OMITTED. SEE PDF]

Diagnostic value

The meta-analysis showed that the combined sensitivity was 0.78 (95% CI: 0.64–0.88) and the specificity was 0.75 (95% CI: 0.59–0.86) (Figure 3); the PLR was 3.1 (95% CI: 1.8–5.4), NLR was 0.29 (95% CI: 0.17–0.50), DOR was 11 (95% CI: 4–27), and AUC was 0.83 (95% CI: 0.80–0.86) (Figure 4). The analysis suggested that hs-cTnT had a moderate diagnostic value for CTRCD (0.7 ≤ AUC < 0.9).

[IMAGE OMITTED. SEE PDF]

[IMAGE OMITTED. SEE PDF]

Heterogeneity tests

SROC contours were drawn for threshold effect analyses, and the results showed that the corresponding points of each study were scattered, without a ‘shoulder-arm-like’ shape. The Spearman correlation coefficient between the logarithm of the calculated sensitivity and the logarithm of 1-specificity was 0.167 (P = 0.693), indicating that heterogeneity was not caused by threshold effects. A forest plot was drawn and the Cochran-Q test was applied to calculate the I² value for heterogeneity assessment, sensitivity had a Cochran-Q of 96.17 (P = 0.00) and I² of 92.72 (95% CI: 89.08–96.36), and specificity had a Cochran-Q of 201.88 (P = 0.00) and I² of 96.53 (95% CI: 95.15–97.92), indicating that the sensitivity and specificity of the eight included studies were highly heterogeneous.

The sources of heterogeneity were identified by meta-regression analysis, and covariates were selected as follows: (i) number of patients (>100 and <100); (ii) study design (prospective and retrospective studies); (iii) ethnicity (Asians vs. Europeans and Americans); (iv) cut-off threshold (≥7 and <7); and (v) detection time point of hs-cTnT (3–6 and <3 or >6 months). As shown in Table 2, ethnicity and detection time point of hs-cTnT were the main sources of heterogeneity. Subgroup analyses also showed that when the detection time point was 3–6 months, the AUC increased from the aforementioned value of 0.83 (95% CI: 0.80–0.86) to 0.90 (95% CI: 0.87–0.92), indicating a better diagnostic performance. In further sensitivity analyses after excluding the data of Finke et al., the pooled sensitivity was 0.75 (95% CI: 0.59–0.86) with a Cochran-Q of 8.63 and I² of 30.46 (95% CI: 0.00–89.58) (P = 0.20), and the pooled specificity was 0.79 (95% CI: 0.64–0.89) with a Cochran-Q of 44.50 and I² of 86.52 (95% CI: 77.86–95.17) (P = 0.00). Sensitivity heterogeneity was significantly reduced, suggesting that the study by Finke et al. might be a major source of heterogeneity. The Deeks' funnel plot was largely symmetrical (P = 0.74), indicating no publication bias (Figure 5).

Table 2 Meta-regression analysis

Covariate/subgroup	Studies, n	Sensitivity (95% CI)	P-value	Specificity (95% CI)	P-value
Number of patients included			0.18		0.26
>100	2	0.83 (0.70–0.96)		0.54 (0.32–0.75)
<100	6	0.76 (0.60–0.91)		0.80 (0.70–0.91)
Study design			0.58		0.39
Prospective	6	0.72 (0.61–0.83)		0.79 (0.63–0.94)
Retrospective	2	0.92 (0.87–0.98)		0.67 (0.37–0.97)
Ethnicity			0.63		0.01
Asian	4	0.77 (0.58–0.97)		0.84 (0.73–0.95)
European and American	4	0.79 (0.66–0.91)		0.62 (0.46–0.78)
Cut-off threshold (ng/L)			0.80		0.68
≥7	6	0.78 (0.64–0.91)		0.76 (0.61–0.91)
<7	2	0. 79 (0.57–1.00)		0.72 (0.42–1.00)
Time (months)			0.66		0.05
3–6	4	0.77 (0.59–0.95)		0.84 (0.71–0.97)
<3 or >6	4	0.79 (0.65–0.93)		0.65 (0.48–0.83)

[IMAGE OMITTED. SEE PDF]

The PLR of hs-cTnT for CTRCD diagnosis was 3.1 (95% CI: 1.8–5.4), NLR was 0.29 (95% CI: 0.17–0.50), and pre-test probability was set at 25% (the median of the included studies) to draw the Fagan nomogram. The results are shown in Figure 6. The pre-test and post-test probabilities were 51% and 9%, respectively, indicating that the possibility of cardiotoxicity increased from 25% to 51% after hs-cTnT diagnosis and that the possibility of no cardiotoxicity decreased from 25% to 9%. As such, the hs-cTnT test could improve the accuracy of CTRCD diagnosis.

[IMAGE OMITTED. SEE PDF]

Discussion

This meta-analysis evaluated the early diagnostic value of hs-cTnT in CTRCD. The results showed that elevated hs-cTnT levels had a high early diagnostic value for CTRCD (AUC = 0.90) at 3–6 months after the initiation of anthracyclines, trastuzumab, anthracyclines in combination with trastuzumab, ICIs, and radiotherapy in combination with anthracyclines.

Our meta-analysis of hs-cTnT for CTRCD diagnosis showed a combined sensitivity of 0.78 (95% CI: 0.64–0.88), specificity of 0.75 (95% CI: 0.59–0.86), and AUC of 0.83 (95% CI: 0.80–0.86), suggesting a moderate diagnostic performance. However, when the detection time point was 3–6 months, the AUC increased to 0.90 (95% CI: 0.87–0.92), indicating a high early diagnostic value. This finding is consistent with results of previous studies.^11,13,15,28 Meta-regression analysis showed that ethnicity and detection time point of hs-cTnT were the main sources of heterogeneity, and sensitivity analysis showed that I² for sensitivity and specificity decreased to 30.46 (95% CI: 0.00–89.58) and 86.52 (95% CI: 77.86–95.17), respectively, after the exclusion of the study by Finke et al.²⁰ This observation might result from different cut-off thresholds, study designs, patient populations, and treatments. Hs-cTnT had a PLR of 3.1 (95% CI: 1.8–5.4) and an NLR of 0.29 (95% CI: 0.17–0.50) for CTRCD diagnosis, suggesting that although the accuracy of diagnosing CTRCD patients was 3.1 times higher than that of judging healthy subjects, 29% of CTRCD patients might be missed. The Fagan line plot of clinical applicability evaluation showed that when the pre-test probability (CTRCD prevalence) was assumed to be 25%, the positive post-test probability (probability of CTRCD occurring when the test was positive) of hs-cTnT increased to 51%, and the negative post-test probability (probability of CTRCD occurring when the test was negative) decreased to 9%, suggesting that clinical assessment of hs-cTnT could improve the diagnostic rate of CTRCD.

In a cardiotoxicity assay of 35 patients with non-Hodgkin's lymphoma (NHL) receiving chemotherapy with cyclophosphamide, adriamycin, vincristine, and prednisone (CHOP), the overall hs-cTnT levels had increased significantly during the fourth cycle of chemotherapy (every 21 days) when LVEF did not show any abnormalities.¹⁵ In a study of 225 patients with non-small cell lung cancer (NSCLC), elevated hs-cTnT levels during concurrent dual chemotherapy with platinum and paclitaxel and chest radiotherapy were dependent on radiation heart dose. Hs-cTnT levels were associated with CTRCD and mortality, and routine monitoring of hs-cTnT could identify patients at high risk of CTRCD due to radiotherapy at an early stage.²⁹ The study by Finke et al. showed that echocardiography in combination with hs-cTnT for CTRCD diagnosis, with a low hs-cTnT cut-off value of 7 ng/L, could identify patients with high mortality risk, suggesting that monitoring hs-cTnT values before chemotherapy initiation could be an important tool for mortality risk stratification in cancer patients.²¹ Because acute myocardial infarction due to myocyte ischaemia and hypoxia could lead to myocardial cell necrosis and destruction of cell membranes and organelles, the release of troponin into the bloodstream can be detected. Possible pathophysiological causes of troponin elevation in patients with CTRCD include³⁰ excitation/contraction coupling and/or alterations in intracellular calcium homeostasis and/or mitochondrial function, leading to cardiomyocyte dysfunction; altered cardiac pre-loading and post-loading conditions; and alterations in extracellular matrix composition.^31,32 The injury in CTRCD begins at the cellular level before abnormal cardiac function. The ability of hs-cTnT to accurately identify small changes in troponin concentration (increase or decrease) within a short time period provides greater analytical accuracy and clinical sensitivity for acute myocardial injury and, thus, allows identification of early myocardial injury at the cellular level.³³

Echocardiography is currently the first choice for CTRCD diagnosis, which is non-invasive and convenient.^3,4,10 However, echocardiography has a large variability in measuring LVEF and overall longitudinal strain. Several studies showed that echocardiography results were only reduced in the presence of critical myocardial injury and cardiac dysfunction.³⁴ After echocardiographic diagnosis of CTRCD and therapeutic interventions, up to 58% of patients still failed to recover.³⁵ Meanwhile, electrocardiogram (ECG) is a non-invasive technique and has advantages in determining the presence of arrhythmias (such as atrial fibrillation, long QT, and premature beats) in CTRCD patients. However, patients often have severe myocardial injury by the time they present with ST-segment changes, and its diagnosis of myocardial injury has a delay. Myocardial nuclear imaging can show myocardial perfusion abnormalities, myocardial apoptosis, and cardiac function for early CTRCD diagnosis in cancer patients after treatment.⁴ However, this technique has been performed less frequently and there is less evidence from clinical studies. Galán-Arriola et al. used cardiac MRI for the early diagnosis of cardiotoxicity in pigs treated with anthracyclines and found that cardiac MRI could be used to monitor the stage of myocardial oedema in pigs treated with doxorubicin, and the progression of this stage could be stopped after withdrawing doxorubicin treatment. This observation indicated that cardiac MRI could monitor the early reversible stage of cardiotoxicity caused by treatment with anthracyclines.⁶ In a prospective study, hs-cTnT was reportedly increased from 4.6 ± 1.4 ng/dL before treatment to 21.3 ± 14.4 ng/dL (P < 0.01) in women with breast cancer in the early phase of treatment with anthracyclines (during Days 79–146), whereas cardiac MRI showed a decrease in LVEF from 69.4 ± 3.6% before treatment to 58.0 ± 6.0% (P < 0.001). The 12% decrease suggested that cardiac MRI can monitor CTRCD patients with elevated hs-cTnT and no significant changes in cardiac function.⁵ Although cardiac MRI has the advantage of early monitoring of CTRCD, its clinical application is somewhat limited due to its high cost and difficulty in obtaining continuous monitoring. In conclusion, for CTRCD diagnosis, echocardiography and electrocardiography have a lag, and myocardial nuclear imaging and cardiac MRI can monitor CTRCD at an early stage, but with limitations in clinical implementation. Therefore, hs-cTnT is more advantageous than other diagnostic tools for early CTRCD diagnosis.

This study had some limitations. First, the number of studies and patients included in the meta-analysis was small, and all tumour types and antitumor treatments were included in a generalized manner, whereas the effects of different tumour types and treatments on the results were not examined in detail. Second, there was a risk of bias in terms of literature quality, and the included studies had a high risk of bias in the Index Test and Flow and Timing domains, mainly because some studies did not provide blinded information on hs-cTnT and did not pre-define hs-cTnT thresholds or failed to detect hs-cTnT and LVEF at appropriate intervals, nor did they examine patients under the same gold standard, nor did they include all patients in statistical analyses. Therefore, the quality of the included studies must be improved. Third, the included patients were free of cardiovascular disease before anticancer therapy, which might have caused the underestimation of the cardiotoxicity risk resulting from anticancer therapy. Fourth, the instruments used to test hs-cTnT in the eight studies were from different distributors, and the instrument models were not identical, which may have impacted the test results. Fifth, we failed to determine the optimal cut-off value for diagnosis, which is quite difficult in clinical practice and needs to be determined by future studies. Finally, because of the limited number of relevant studies, this meta-analysis evaluated the diagnostic accuracy of only one cardiac-specific biomarker (hs-cTnT) and did not compare it with other markers of myocardial injury, such as high-sensitivity cardiac troponin I (hs-cTnI) and traditional cardiac troponin T (cTnT), which would also have a certain impact on its clinical applicability.

In future research, we will focus on studies of hs-cTnI for CTRCD diagnosis and compare the ability of hs-cTnT vs. hs-cTnI for early CTRCD diagnosis. In addition, we hope to further compare the sensitivity and specificity of hs-cTnT with conventional cTnT in the diagnosis of CTRCD in order to screen for more sensitive diagnostic indicators. However, because the numbers of patients in their two study groups were significantly unequal, these results need to be further validated. Therefore, we hope to further compare the sensitivity and specificity of hs-cTnT with those of conventional cTnT for CTRCD diagnosis for the development of a more sensitive diagnostic index. Finally, analysis of the simultaneous use of hs-cTnT and hs-cTnI for CTRCD diagnosis may be more beneficial for our next research study.

In conclusion, early (3–6 months) detection of hs-cTnT showed a high diagnostic value for CTRCD. However, more prospective multicentre studies with rigorous designs and large sample sizes are needed to determine the optimal cut-off value for CTRCD diagnosis. We hope to combine serum biomarkers with other diagnostic methods for more accurate early CTRCD diagnosis and improve the quality of life and prognosis of cancer survivors.

Acknowledgements

We thank the researchers working in cardiac oncology.

Conflict of interest

None declared.

Funding

This work was supported by the National Natural Science Foundation of China (Grant Numbers 81873132 and 81860786), the Special Open Project of Gansu Research Center of Traditional Chinese Medicine (No. zyzx-2020-zx8), the Lanzhou Science and Technology Plan Project (No. 2022-3-27), and the 2018 Chinese Medicine Research Projects to Prevent and Treat Major Diseases (No. GZKZD-2018-02).

References

Miller KD, Nogueira L, Mariotto AB, Rowland JH, Yabroff KR, Alfano CM, Jemal A, Kramer JL, Siegel RL. Cancer treatment and survivorship statistics, 2019. CA Cancer J Clin. 2019; 69: 363–385.

Adhikari BB, Shi S, Dimond EP, Shelburne N, Desvigne‐Nickens P, Minasian LM. Spectrum of National Institutes of Health‐funded research in cardio‐oncology: a basic, clinical, and observational science perspective. Heart Fail Clin. 2022; 18: 515–528.

Leong DP, Lenihan DJ. Clinical practice guidelines in cardio‐oncology. Heart Fail Clin. 2022; 18: 489–501.

Zhang P, Hu X, Yue J, Meng X, Han D, Sun X, Yang G, Wang S, Wang X, Yu J. Early detection of radiation‐induced heart disease using ^99mTc‐MIBI SPECT gated myocardial perfusion imaging in patients with oesophageal cancer during radiotherapy. Radiother Oncol. 2015; 115: 171–178.

Ferreira de Souza T, Quinaglia ACST, Osorio Costa F, Shah R, Neilan TG, Velloso L, Nadruz W, Brenelli F, Sposito AC, Matos‐Souza JR, Cendes F, Coelho OR, Jerosch‐Herold M, Coelho‐Filho OR. Anthracycline therapy is associated with cardiomyocyte atrophy and preclinical manifestations of heart disease. JACC Cardiovasc Imaging. 2018; 11: 1045–1055.

Galán‐Arriola C, Lobo M, Vilchez‐Tschischke JP, Lopez GJ, de Molina‐Iracheta A, Perez‐Martinez C, Aguero J, Fernandez‐Jimenez R, Martin‐Garcia A, Oliver E, Villena‐Gutierrez R, Pizarro G, Sánchez PL, Fuster V, Sánchez‐González J, Ibanez B. Serial magnetic resonance imaging to identify early stages of anthracycline‐induced cardiotoxicity. J Am Coll Cardiol. 2019; 73: 779–791.

Pudil R, Mueller C, Čelutkienė J, Henriksen PA, Lenihan D, Dent S, Barac A, Stanway S, Moslehi J, Suter TM, Ky B, Štěrba M, Cardinale D, Cohen‐Solal A, Tocchetti CG, Farmakis D, Bergler‐Klein J, Anker MS, von Haehling S, Belenkov Y, Iakobishvili Z, Maack C, Ciardiello F, Ruschitzka F, Coats AJS, Seferovic P, Lainscak M, Piepoli MF, Chioncel O, Bax J, Hulot JS, Skouri H, Hägler‐Laube ES, Asteggiano R, Fernandez TL, Boer RA, Lyon AR. Role of serum biomarkers in cancer patients receiving cardiotoxic cancer therapies: a position statement from the Cardio‐Oncology Study Group of the Heart Failure Association and the Cardio‐Oncology Council of the European Society of Cardiology. Eur J Heart Fail. 2020; 22: 1966–1983.

Zamorano JL, Lancellotti P, Rodriguez Munoz D, Aboyans V, Asteggiano R, Galderisi M, Habib G, Lenihan DJ, Lip GYH, Lyon AR, Lopez Fernandez T, Mohty D, Piepoli MF, Tamargo J, Torbicki A, Suter TM, ESC Scientific Document Group. 2016 ESC Position Paper on cancer treatments and cardiovascular toxicity developed under the auspices of the ESC Committee for Practice Guidelines: the Task Force for cancer treatments and cardiovascular toxicity of the European Society of Cardiology (ESC). Eur Heart J. 2016; 37: 2768–2801.

Pudil RMC, Čelutkienė J, Henriksen PA, Lenihan D, Dent S, Barac A, Stanway S, Moslehi J, Suter TM, Ky B, Štěrba M, Cardinale D, Cohen‐Solal A, Tocchetti CG, Farmakis D, Bergler‐Klein J, Anker MS, Von Haehling S, Belenkov Y, Iakobishvili Z, Maack C, Ciardiello F, Ruschitzka F, Coats AJS, Seferovic P, Lainscak M, Piepoli MF, Chioncel O, Bax J, Hulot JS, Skouri H, Hägler‐Laube ES, Asteggiano R, Fernandez TL, de Boer RA, Lyon AR. Role of serum biomarkers in cancer patients receiving cardiotoxic cancer therapies: a position statement from the Cardio‐Oncology Study Group of the Heart Failure Association and the Cardio‐Oncology Council of the European Society of Cardiology. Eur J Heart Fail. 2020; 22: 1966–1983.

Curigliano G, Lenihan D, Fradley M, Ganatra S, Barac A, Blaes A, Herrmann J, Porter C, Lyon AR, Lancellotti P, Patel A, DeCara J, Mitchell J, Harrison E, Moslehi J, Witteles R, Calabro MG, Orecchia R, de Azambuja E, Zamorano JL, Krone R, Iakobishvili Z, Carver J, Armenian S, Ky B, Cardinale D, Cipolla CM, Dent S, Jordan K, ESMO Guidelines Committee. Electronic address: [email protected]. Management of cardiac disease in cancer patients throughout oncological treatment: ESMO consensus recommendations. Ann Oncol. 2020; 31: 171–190.

Michel L, Mincu RI, Mahabadi AA, Settelmeier S, Al‐Rashid F, Rassaf T, Totzeck M. Troponins and brain natriuretic peptides for the prediction of cardiotoxicity in cancer patients: a meta‐analysis. Eur J Heart Fail. 2020; 22: 350–361.

Katsurada KIM, Sakuragi M, Takehara M, Hozumi Y, Kario K. High‐sensitivity troponin T as a marker to predict cardiotoxicity in breast cancer patients with adjuvant trastuzumab therapy. Springerplus. 2014; 3: 620.

Cheung YF, Li VW, Lai CT, Shin VY, Keung W, Cheuk DK, Kwong A, Li RA, Chan GC. Circulating high‐sensitivity troponin T and microRNAs as markers of myocardial damage during childhood leukaemia treatment. Pediatr Res. 2021; 89: 1245–1252.

Diaz‐Anton B, Madurga R, Zorita B, Wasniewski S, Moreno‐Arciniegas A, Lopez‐Melgar B, Ramirez Merino N, Martin‐Asenjo R, Barrio P, Amado Escanuela MG, Solís J, Parra Jiménez FJ, Ciruelos E, Castellano JM, Fernández‐Friera L. Early detection of anthracycline‐ and trastuzumab‐induced cardiotoxicity: value and optimal timing of serum biomarkers and echocardiographic parameters. ESC Heart Fail. 2022; 9: 1127–1137.

Prayogo AA, Suryantoro SD, Savitri M, Hendrata WM, Wijaya AY, Pikir BS. High sensitivity troponin T as complementary modality for determining doxorubicin regimen cardiotoxicity in non‐Hodgkin lymphoma patients. Adv Pharm Bull. 2022; 12: 163–168.

Demissei BG, Hubbard RA, Zhang L, Smith AM, Sheline K, McDonald C, Narayan V, Domchek SM, DeMichele A, Shah P, Clark AS, Fox K, Matro J, Bradbury AR, Knollman H, Getz KD, Armenian SH, Januzzi JL, Tang WHW, Liu P, Ky B. Changes in cardiovascular biomarkers with breast cancer therapy and associations with cardiac dysfunction. J Am Heart Assoc. 2020; 9: [eLocator: e014708].

Zhang CJ, Pei XL, Song FY, Guo Y, Zhang QL, Shu XH, Hsi DH, Cheng LL. Early anthracycline‐induced cardiotoxicity monitored by echocardiographic Doppler parameters combined with serum hs‐cTnT. Echocardiography. 2017; 34: 1593–1600.

Bisoc A, Ciurescu D, Rădoi M, Tântu MM, Rogozea L, Sweidan AJ, Bota DA. Elevations in high‐sensitive cardiac troponin T and N‐terminal prohormone brain natriuretic peptide levels in the serum can predict the development of anthracycline‐induced cardiomyopathy. Am J Ther. 2020; 27: e142–e150.

Katsurada K, Ichida M, Sakuragi M, Takehara M, Hozumi Y, Kario K. High‐sensitivity troponin T as a marker to predict cardiotoxicity in breast cancer patients with adjuvant trastuzumab therapy. Springerplus. 2014; 3: 620.

Kitayama H, Kondo T, Sugiyama J, Kurimoto K, Nishino Y, Kawada M, Hirayama M, Tsuji Y. High‐sensitive troponin T assay can predict anthracycline‐ and trastuzumab‐induced cardiotoxicity in breast cancer patients. Breast Cancer. 2017; 24: 774–782.

Finke D, Romann SW, Heckmann MB, Hund H, Bougatf N, Kantharajah A, Katus HA, Müller OJ, Frey N, Giannitsis E, Lehmann LH. High‐sensitivity cardiac troponin T determines all‐cause mortality in cancer patients: a single‐centre cohort study. ESC Heart Fail. 2021; 8: 3709–3719.

Petricciuolo S, Delle Donne MG, Aimo A, Chella A, De Caterina R. Pre‐treatment high‐sensitivity troponin T for the short‐term prediction of cardiac outcomes in patients on immune checkpoint inhibitors. Eur J Clin Invest. 2021; 51: [eLocator: e13400].

Kang Y, Xu X, Cheng L, Li L, Sun M, Chen H, Pan C, Shu X. Two‐dimensional speckle tracking echocardiography combined with high‐sensitive cardiac troponin T in early detection and prediction of cardiotoxicity during epirubicine‐based chemotherapy. Eur J Heart Fail. 2014; 16: 300–308.

Zardavas D, Suter TM, Van Veldhuisen DJ, Steinseifer J, Noe J, Lauer S, Al‐Sakaff N, Piccart‐Gebhart MJ, de Azambuja E. Role of troponins I and T and N‐terminal prohormone of brain natriuretic peptide in monitoring cardiac safety of patients with early‐stage human epidermal growth factor receptor 2‐positive breast cancer receiving trastuzumab: a herceptin adjuvant study cardiac marker substudy. J Clin Oncol. 2017; 35: 878–884.

GS. HJ: Cochrane Handbook for Systematic Reviews of Interventions, Version 5.1.0. https://trainingcochraneorg/handbook 2018. Accessed March 2011.

McInnes MDF, Moher D, Thombs BD, McGrath TA, Bossuyt PM, the PRISMA‐DTA Group, Clifford T, Cohen JF, Deeks JJ, Gatsonis C, Hooft L, Hunt HA, Hyde CJ, Korevaar DA, Leeflang MMG, Macaskill P, Reitsma JB, Rodin R, Rutjes AWS, Salameh JP, Stevens A, Takwoingi Y, Tonelli M, Weeks L, Whiting P, Willis BH. Preferred reporting items for a systematic review and meta‐analysis of diagnostic test accuracy studies: the PRISMA‐DTA statement. JAMA. 2018; 319: 388–396.

Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, Leeflang MM, Sterne JA, Bossuyt PM, QUADAS‐2 Group. QUADAS‐2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011; 155: 529–536.

Díaz‐Antón B, Madurga R, Zorita B, Wasniewski S, Moreno‐Arciniegas A, López‐Melgar B, Ramírez Merino N, Martín‐Asenjo R, Barrio P, Amado Escañuela MG, Solís J, Parra Jiménez FJ, Ciruelos E, Castellano JM, Fernández‐Friera L. Early detection of anthracycline‐ and trastuzumab‐induced cardiotoxicity: value and optimal timing of serum biomarkers and echocardiographic parameters. ESC Heart Fail. 2022; 9: 1127–1137.

Xu TMQ, Gilchrist SC, Lin SH, Lin R, Xu T, Milgrom SA, Gandhi SJ, Wu H, Zhao Y, Lopez‐Mattei JC, Mohan R, Liao Z. Assessment of prognostic value of high‐sensitivity cardiac troponin T for early prediction of chemoradiation therapy‐induced cardiotoxicity in patients with non‐small cell lung cancer: a secondary analysis of a prospective randomized trial. Int J Radiat Oncol Biol Phys. 2021; 111: 907–916.

Ferdinandy P, Baczko I, Bencsik P, Giricz Z, Gorbe A, Pacher P, Varga ZV, Varro A, Schulz R. Definition of hidden drug cardiotoxicity: paradigm change in cardiac safety testing and its clinical implications. Eur Heart J. 2019; 40: 1771–1777.

Simões R, Silva LM, Cruz A, Fraga VG, de Paula Sabino A, Gomes KB. Troponin as a cardiotoxicity marker in breast cancer patients receiving anthracycline‐based chemotherapy: a narrative review. Biomed Pharmacother. 2018; 107: 989–996.

Cardinale DM, Zaninotto M, Cipolla CM, Passino C, Plebani M, Clerico A. Cardiotoxic effects and myocardial injury: the search for a more precise definition of drug cardiotoxicity. Clin Chem Lab Med. 2020; 59: 51–57.

Askin L, Tanriverdi O, Turkmen S. Clinical importance of high‐sensitivity troponin T in patients without coronary artery disease. North Clin Istanb. 2020; 7: 305–310.

Curigliano G, Cardinale D, Dent S, Criscitiello C, Aseyev O, Lenihan D, Cipolla CM. Cardiotoxicity of anticancer treatments: epidemiology, detection, and management. CA Cancer J Clin. 2016; 66: 309–325.

Moudgil R, Hassan S, Palaskas N, Lopez‐Mattei J, Banchs J, Yusuf SW. Evolution of echocardiography in subclinical detection of cancer therapy‐related cardiac dysfunction. Echocardiography. 2018; 35: 860–868.