Abstract
Purpose
This study aimed to systematically assess the efficacy of cardioprotective agents in preventing anthracycline-induced cardiotoxicity in patients with breast cancer using a comprehensive network meta-analysis (NMA).
Methods
This study included patients with breast cancer undergoing anthracycline-based chemotherapy. Randomized controlled trials (RCTs) published before March 2020 were identified through systematic searches in MEDLINE, Cochrane CENTRAL, Web of Science, and CINAHL. The primary outcome was left ventricular ejection fraction (LVEF), assessed using cardiac magnetic resonance imaging, multigated radionuclide angiography, or echocardiography. The NMA integrated direct and indirect comparisons to estimate the relative effectiveness of pharmacological interventions.
Results
The systematic review included 31 RCTs with 3,228 participants, whereas the NMA synthesized 25 effect sizes from 15 RCTs. Mineralocorticoid receptor antagonists (MRAs) [standardized mean difference (SMD): −1.78, 95% confidence interval (CI): −2.81 to −0.75] and trimetazidine (SMD: −1.12, 95%CI: −2.32 to −0.09) exhibited the most substantial cardioprotective effects. Dexrazoxane (SMD: −0.53, 95%CI: −1.90 to −0.02) and β-blockers (SMD: −0.34, 95%CI: −0.70 to 0.02) showed potential benefits, albeit with greater uncertainty. Direct comparisons showed that dexrazoxane was more effective than β-blockers (SMD: −1.25, 95%CI: −2.22 to −0.48), with mineralocorticoid receptor antagonists (MRAs) outperforming both. Despite heterogeneity and potential publication bias, mineralocorticoid receptor antagonists (MRAs) and trimetazidine consistently ranked as the most effective interventions. LVEF findings confirmed the cardioprotective benefits of β-blockers, ARBs, ACE inhibitors, and dexrazoxane.
Conclusions
RCT evidence suggested that cardioprotective drugs effectively mitigate anthracycline-induced LVEF decline. However, the lack of direct head-to-head trials limits definitive conclusions on comparative efficacy, warranting trials in patients with lower baseline LVEF to optimize cardioprotective strategies.
Supplementary Information
The online version contains supplementary material available at 10.1007/s10549-025-07791-7.
Keywords: Anthracyclines, Chemotherapy, Cardioprotective interventions, Breast cancer, Randomized clinical trials, Network meta-analysis
Introduction
Cardiotoxicity and heart failure are critical complications of anthracycline-based chemotherapy, significantly increasing morbidity and mortality among cancer patients [1, 2]. The adverse effects of these agents are primarily evaluated using left ventricular ejection fraction (LVEF) and cardiac biomarkers, which constitute indicators of myocardial injury and systolic dysfunction [3]. A reduction in LVEF is widely acknowledged as an early sign of anthracycline-induced cardiotoxicity. However, the long-term impact of these agents on cardiac function remains unclear [4]. Addressing anthracycline-induced cardiotoxicity is a major challenge in breast cancer survivorship, necessitating effective prophylactic strategies, including early detection of subclinical toxicity, modified anthracycline formulations, and cardioprotective pharmacological interventions [5].
Several classes of cardioprotective agents, including β-blockers, angiotensin-converting enzyme inhibitors (ACEIs), angiotensin receptor blockers (ARBs), mineralocorticoid receptor antagonists (MRAs), and dexrazoxane, have been investigated for mitigating anthracycline-induced cardiac toxicity [6, 7]. While previous studies have demonstrated their potential benefits, direct head-to-head comparisons remain lacking, causing difficulty in establishing the most effective intervention [8].
Considering the existing uncertainty, a network meta-analysis (NMA) is imperative for synthesizing direct and indirect evidence to facilitate a robust comparative assessment of available pharmacological interventions. This study aimed to provide a comprehensive and methodologically rigorous evaluation of prophylactic pharmacological strategies for patients with breast cancer undergoing anthracycline-based chemotherapy, thereby supporting evidence-based clinical decision-making. The primary objective was to assess the efficacy of cardioprotective agents in preventing LVEF decline and chemotherapy-induced cardiotoxicity. Furthermore, an NMA enables indirect comparisons between interventions not directly evaluated in randomized trials, enhancing clinical evidence robustness.
Methods
Protocol and registration
The network meta-analysis was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses for Network Meta-Analyses (PRISMA-NMA) guidelines [9] and methodological principles for evaluating multiple treatments, as described in the Cochrane Handbook for Systematic Reviews of Interventions. The review protocol was prospectively registered in the PROSPERO database (CRD42020168609) (https://www.york.ac.uk/crd/) to ensure methodological transparency and rigor.
Search strategy
A thorough systematic literature search was performed across multiple electronic databases, including MEDLINE (PubMed), Cochrane CENTRAL, Web of Science, and CINAHL, encompassing studies published from the inception of each database up to June 15, 2022. Advanced text mining techniques were applied to optimize the selection of search terms regarding the target population (breast cancer), intervention (cardioprotective agents), and study design (randomized controlled trials). Specifically, Python’s Natural Language Toolkit (NLTK) was used for text processing, whereas term frequency-inverse document frequency (TF-IDF) and word embedding models (such as Word2Vec and GloVe) facilitated the identification of relevant keywords. The final search strategy incorporated both MeSH and free-text terms, systematically combined using Boolean operators to enhance retrieval precision. The search and screening process was independently conducted by two reviewers, with any discrepancies resolved through discussion with a third reviewer. A comprehensive list of search terms and a representative search strategy are presented in Table 1.
Table 1.
PubMed search strategy
| ((((((((((((((((RCT[Title/Abstract]) OR Experimental study[Title/Abstract]) OR Randomis*[Title/Abstract]) OR randomized control*[Title/Abstract]) OR randomised[Title/Abstract]) OR randomized*[Title/Abstract]) OR randomised control trial[Title/Abstract]) OR Randomised clinical[Title/Abstract]) OR Randomized clinical[Title/Abstract])) AND ( "0001/01/01"[PDat]: "2022/06/15"[PDat]) AND Humans[Mesh] AND English[lang])) AND Humans[Mesh] AND English[lang])) AND ((((((breast cancer[MeSH Terms] AND ( "0001/01/01"[PDat]: "2022/06/15"[PDat]) AND Humans[Mesh] AND English[lang]) AND Humans[Mesh] AND English[lang])) AND (((((((((((((Adrenergic beta-Antagonists) OR Angiotensin-Converting Enzyme Inhibitors) OR Angiotensin Receptor Antagonists) OR Diuretics) AND ( "0001/01/01"[PDat]: "2022/06/15"[PDat]) AND Humans[Mesh] AND English[lang])) OR (((((((Cardioxane) OR Cardioxan) OR Dexrazoxan*) OR Zinecard) OR Razoxane) OR Dexrazoxane) AND ( "0001/01/01"[PDat]: "2022/06/15"[PDat]) AND Humans[Mesh] AND English[lang])) AND ( "0001/01/01"[PDat]: "2022/06/15"[PDat]) AND Humans[Mesh] AND English[lang])) AND ( "0001/01/01"[PDat]: "2022/06/15"[PDat]) AND Humans[Mesh] AND English[lang]))) AND Humans[Mesh] AND English[lang])) AND ((((((((((((anthracycline) OR aclarubicin) OR daunorubicin) OR daunomycin) OR carubicin) OR plicamycin) OR doxorubicin) OR epirubicin) OR idarubicin) OR nogalamycin) OR menogaril) AND ( "0001/01/01"[PDat]: "2022/06/15"[PDat]) AND Humans[Mesh] AND English[lang])) AND Humans[Mesh] AND English[lang])) AND Humans[Mesh] AND English[lang]) |
Inclusion criteria
Studies were eligible for the NMA if they met the following criteria: (1) utilized a randomized controlled trial (RCT) design; (2) included female participants aged ≥ 18 years; (3) included patients with confirmed breast cancer diagnosis; (4) administered anthracycline-based chemotherapy as part of primary tumor treatment; (5) optionally included concomitant radiation therapy; (6) evaluated a single cardioprotective pharmacological agent, such as ACEIs, ARBs, antimitotics, β−1 adrenergic receptor blockers, trimetazidine, or MRAs; (7) assessed LVEF as the primary outcome, alongside secondary measures such as cardiac biomarkers, documented cardiac events (e.g., heart failure), treatment-related adverse events, and cancer-related mortality; (8) published in a peer-reviewed, English-language medical journal; and (9) trial completion before the COVID-19 pandemic lockdown (March 22, 2020).
Data collection and quality assessment
The study selection and data extraction process were conducted independently by two investigators (PV and KJM) using a standardized data collection form. During initial screening, the titles and abstracts of all retrieved articles were assessed according to predefined inclusion and exclusion criteria. Studies meeting these criteria underwent a full-text review for further evaluation. Any discrepancies were resolved through discussion and consensus with a third investigator (FG). When necessary, the corresponding authors of the included RCTs were contacted—at most twice—to obtain additional study details.
A standardized Microsoft Excel data extraction form was employed to systematically collect key study information, including study identification details, publication year, country, application of the intention-to-treat principle, patient demographics, cardioprotective agents, intervention details (prior and concurrent therapies), treatment context (metastatic or adjuvant setting), cardiotoxicity management, chemotherapy regimens, anthracycline type, cumulative anthracycline dose, history of prior anthracycline exposure, sample size, mean patient age, treatment duration, follow-up period, comorbidities, radiotherapy specifics (frequency, dose, cumulative exposure), and risk of bias.
Each included study’s risk of bias was evaluated using the Cochrane Collaboration’s risk of bias assessment tool, examining seven key domains: random sequence generation, allocation concealment, participant and study personnel blinding, outcome assessment blinding, outcome data completeness, selective outcome reporting, and other potential sources of bias. Two investigators (PV and KJM) independently conducted the evaluation, and any discrepancies were resolved through discussion with a third investigator (FG).
Primary and secondary outcomes
The primary outcome was the change in LVEF from baseline to follow-up in patients receiving cardioprotective interventions relative to those in the control group. Mean LVEF changes were determined using the reported mean (M), standard deviation (SD), and sample size (n). Effect sizes were calculated or converted into standardized mean differences (SMD) using pooled standard deviations or established transformation methods, including Hedges’ g correction to adjust for small-sample bias. If essential statistical data were unavailable, alternative estimations were applied where feasible; otherwise, studies were omitted from the quantitative analysis.
Secondary outcomes included the incidence of chemotherapy-induced cardiotoxicity, cancer-related mortality, and safety outcomes, such as serious adverse events, treatment-related adverse effects, and known toxicities associated with anthracycline therapy.
The NMA framework operates under the assumption that effect estimates are derived from a collection of RCTs with comparable effect modifiers [10]. However, the transitivity assumption may be violated if the distribution of these effect modifiers substantially differs among studies. To address this potential issue, comparator groups were categorized based on the type of cardioprotective agents, including ACEIs, ARBs, antimitotics, β−1 blockers, trimetazidine, and MRAs.
Statistical analysis
Data analysis was conducted using Stata software (Stata Statistical Software: Release 15, StataCorp LLC, College Station, TX) to perform a multivariate random-effects meta-analysis [11]. The change in LVEF from baseline to post-chemotherapy was treated as a continuous variable, reported as the SMD with corresponding 95% confidence intervals (CI). Hedges’ g correction was applied to estimate SMD values, and effect sizes were interpreted based on Cohen’s criteria, where 0.2 represents a small, 0.5 a medium, and 0.8 a large effect [12].
The degree of heterogeneity was evaluated using the I2 statistic, classified as follows: low (0–25%), moderate (25–50%), substantial (50–75%), and high (> 75%). Cochran’s Q test was applied to assess the statistical significance, with a threshold of p < 0.10. Between-study heterogeneity and inconsistency were examined in the NMA using global inconsistency models based on the design-by-treatment interaction approach and local inconsistency tests employing node-splitting analysis. A random-effects model using either a Bayesian hierarchical framework or the frequentist DerSimonian and Laird method was employed to account for between-study variability. The consistency of direct and indirect evidence was examined by comparing estimates from both sources. Sensitivity analyses were conducted by excluding studies with a high risk of bias or exhibiting extreme effect sizes.
Publication bias and small-study effects were examined through comparison-adjusted funnel plots and Egger’s test. The certainty of the evidence supporting the network estimates was assessed using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) framework. Statistical significance was set at a two-sided p-value < 0.05 [12].
A network plot was generated using the ‘network plot’ command to display direct treatment comparisons. The contribution plot (‘net weight’) was utilized to demonstrate the relative influence of direct and indirect comparisons within the network model, whereas the interval plot (‘interval plot’) was used to visualize ranking uncertainty. Treatment efficacy was evaluated using the surface under the cumulative ranking curve (SUCRA) and mean ranks, with analyses performed using the ‘sucra’ and ‘network rank’ commands. Additionally, inconsistency was examined through the ‘ifplot’ command and inconsistency models, which assessed loop-specific and global consistency. Publication bias and small-study effects were examined through comparison-adjusted funnel plots using the ‘netfunnel’ command.
Data synthesis
A systematic approach was employed to integrate findings across studies for primary and secondary outcomes. Pairwise meta-analyses were conducted for direct comparisons using a random-effects model to account for inter-study variability, whereas an NMA was conducted for indirect comparisons using either a Bayesian hierarchical model or a frequentist approach, ensuring coherence between direct and indirect estimates.
Missing or inconsistent data were managed through multiple imputation techniques when applicable. Sensitivity analyses were conducted to evaluate the potential influence of missing data on the overall findings. If studies reported incomplete outcome measures, effect sizes were estimated using standard conversion methods based on the available data. The analysis adhered to the assumption of consistency, and any deviations were explored through node-splitting and inconsistency models. This rigorous methodological framework ensured a comprehensive and reliable synthesis of evidence across the network.
Results
The systematic literature search identified 766 potentially relevant records. After removing duplicates (n = 146) (Fig. 1), the remaining RCTs underwent title and abstract screening by two independent reviewers, achieving a 96% agreement rate. After the initial screening, 77 full-text articles were evaluated for eligibility, of which 46 studies were excluded based on predefined criteria (Fig. 1). The full-text review was independently conducted by two researchers, achieving a 95% agreement rate. The systematic review encompassed 31 RCTs with 3,228 participants. Among these, 15 RCTs (n = 1,545 participants, 25 effect sizes) met the inclusion criteria for the NMA.
Fig. 1.
Flowchart of the study selection process
The NMA encompassed 15 RCTs with 1,545 participants, including 892 in the treatment group and 653 in the control group (Fig. 2). All participants were female patients diagnosed with breast cancer who underwent anthracycline-based chemotherapy, receiving either doxorubicin or epirubicin. The RCTs investigated anthracyclines administered alone or combined with other pharmacological agents and/or radiotherapy. The cumulative anthracycline dose varied across studies, ranging from 160.8 to 960 mg/m2. The included studies evaluated various prophylactic cardioprotective agents, including β−1 blockers (8 RCTs), ARBs (2), ACEIs (1), MRAs (1), trimetazidine (1), and dexrazoxane (3) (Table 2).
Fig. 2.
Network plot of included RCTs and predictive interval plot for cardioprotective agents. Black diamonds indicate the estimated effect size (Hedges’ g), with narrower horizontal lines representing confidence intervals (CI) and wider lines depicting predictive intervals (PrI). The blue vertical line marks the null hypothesis (Hedges’ g = 0), while negative values indicate a stronger cardioprotective effect in the comparator (left) group. Abbreviations: ACEI angiotensin-converting enzyme inhibitors, ARB angiotensin II receptor blockers, BB β−1 adrenergic receptor blockers, MRA mineralocorticoid receptor antagonists, DEX dexrazoxane, TMZ trimetazidine, CON control
Table 2.
Characteristics of participants, interventions, and methodologies in randomized controlled trials included in the systematic review
| Author (year) | Treatment Mage (SD) | Control Mage (SD), | Intervention sample size (n) | Control sample size (n) | Anthracycline cumulative drug dose (mg/m2) Treatment. Control |
Type of protective drug intervention | Intervention Dose mg/day (Dex: mg/m2) | Length of intervention (weeks) | Outcomes measure(s) | |
|---|---|---|---|---|---|---|---|---|---|---|
| Akpek (2015) [13] | 50(10.8) | 50.6(10.1) | 43 | 40 | X | X | Spironolactone | 25 | 24 | Transthoracic echocardiography LVEF/Cardiac and oxidative biomarkers/Troponin-I |
| Avila (2018) [14, 15] | 50.8(0.1) | 52.9(9.05) | 96 | 96 | 240 | 240 | Beta-blocker (carvedilol) | 5 | 24 | A 10% LVEF decline at six months, with elevated troponin I, increased brain natriuretic peptide (BNP), and diastolic dysfunction |
| Cochera (2018) [16] | 53(13) | 52 (11) | 30 | 30 | 521 | 519 | Beta-blocker (nebivolol) | 50 | 24 | Echocardiography, LVEF; tissue Doppler and speckle tracking echocardiographic imaging; left ventricular systolic and diastolic function; strain |
| Davis (2019) [17] | 53.9(2) | 49.1(12.8) | 22 | 22 | X | X | Spironolactone (eplerenone) | 12.5 | 24 | Transthoracic echocardiograms, LVEF, systolic and diastolic function; cardiac biomarkers |
| Elitok (2014) [18] | 54.3(9.3) | 52.9(11.2) | 40 | 40 | 535.6 | 523.3 | Beta-blocker (carvedilol) | 25 | 24 | Cardiac evaluation, a standard ECG, a conventional echocardiography, and SI. LVEF; fractional shortening; M-mode echocardiography; strain imaging |
| Gulati (2016) [19] | 50.0 | X | 30 | X | X | X | Candesartan–metoprolol | 32/100 | X | Cardiac MRI, blood samples, physical examinations, and electrocardiograms/LVEF/Cardiac troponin I |
| 51.7(10.7) | 50.8(9.2) | 32 | 32 | 297.5 | 301.3 | Candesartan | 32 | X | ||
| 50.5(9.1) | 32 | 32 | 301.3 | 301.3 | Metoprolol | 100 | X | |||
| Gulati (2017) [20] |
51* (42.0, 59.0) |
X | 27 | X | 400 | X | Metoprolol with placebo (n = 6), candesartan with placebo (n = 7), metoprolol with candesartan (n = 7), and placebo with placebo (n = 7) | Cardiovascular magnetic resonance imaging, echocardiographic assessments, and circulating biomarker levels | ||
|
48.5* (43.8, 58.0) |
X | 94 | X | < 400 | X | Metoprolol and placebo (n = 23), candesartan and placebo (n = 25), metoprolol and candesartan (n = 23), placebo and placebo (n = 23) | ||||
| Heck (2018) [21] | 49.8 | 50.3(9.6) | 18 | 160.8*(160.8,241.2) | Candesartan–metoprolol | 13* (12.0–16.5) | Cardiovascular magnetic resonance | |||
| 52.6(10.2) | 20 | 160.8*(160.8,241.2) | 201.0*(160.8,268.0) | Candesartan | 32 | |||||
| 49.2(8.1) | 13 | 160.8*(160.8,201.0) | Metoprolol | 100 | ||||||
| Kaya (2013) [22] | 51.4(9.4) | 50.5(11.1) | 27 | 18 | 361 | 348 | Beta-blocker (nebivolol) | 5 | 24 | Echocardiography (echocardiographic variables) and NT-pro-BNP levels |
| Lee (2021) [23] | 47.8(8.7) | 48.5(10.4) | 82 | 43 |
240 (240.0– 240.0) |
240 (240.0– 240.0) |
Candesartan | 16 | 24 | Transthoracic echocardiogram |
| 46.6(7.6) | 48.5(10.4) | 70 | 43 |
240 (240.0– 240.0) |
240 (240.0– 240.0) |
Carvedilol | 12.5 | |||
| Marty (2006) [24] | 50*(31–76) | 52* (30–71) | 85 | 79 | 669(247–936) | 608(244–900) | Dexrazoxane | 500 | 24 | Cardiac assessment included a physical examination, MUGA scan, or echocardiography, along with evaluations of cardiac event incidence, congestive heart failure occurrence, and tumor response rate |
| Mross (1993) [25] | 53*(32–66) | 51*(31–67) | 26 | 25 | X | X | Verapamil | 480 | 8 | Response rate and overall survival |
| Nabati (2017) [26] | 47.57(8.75) | 47.1(12.17) | 46 | 45 | 359.91 | 348.56 | Beta-blocker (carvedilol) | 3.125 mg orally twice a day > > 6.125 mg twice daily | 24 | Echocardiographic evaluation of LVEF, end-diastolic, and end-systolic volumes |
| Słowik (2020) [27] | 237.5 | 240 | 48 | 48 |
DOX 237,5 (12,2), EPI 383,3 (102,32) |
DOX 240 (0.0), EPI356.2 (166.30) | Ramipril | 10 | 24 | Echocardiography, LVEF changes, and serum NT-proBNP levels |
| Speyer (1988) [28] | 57*(29–71) | 58.3*(32–76) | 47 | 45 | 471.3 | 376.5 | Dexrazoxane | 1000 | 9.9 cycles(intervention), 8.1 cycles (control) | Radionuclide cardiac scan/nuclear scan/endometrial biopsy |
| Speyer (1990) [29] | X | X | 65 | 70 | 300–1400 | 300–899 | Dexrazoxane | 1000 | X | Clinical exam/MUGA (serial resting and exercise gated pool at baseline and fixed time point) |
| Speyer (1992) [30] | 55.5 | 56.2 | 76 | 74 | 558 | 407.4 | Dexrazoxane | 1000 | 12.3cycles(intervention), 8.8cycles (control) | A clinical score for chronic heart failure to these events and analysis of MUGA scans/LVEF |
| Sun (2015) [31] | 53.47(5.45) | 55.11(2.36) | 40 | 40 | 480 | 480 | Dexrazoxane | 800 | 18 | Echocardiographic assessment of systolic and diastolic function using conventional methods and tissue Doppler imaging ECG |
| Sun (2016) [32] | 53.8(4.99) | 55.25(3.75) | 51 | 52 | 480 | 480 | Dexrazoxane | 800 | 18 | Standard ECG and 24-h Holter monitoring, analyzing low-frequency (LF, 0.04–0.15 Hz), high-frequency (HF, 0.15–0.40 Hz), and the LF/HF ratio to assess autonomic balance |
| Swain (1997) [33] | 57*(33–81) | 56*(25–77) | 102 | 99 | ≥ 300 | ≥ 300 | Dexrazoxane | 500–1000 | ≥ 7 cycles | Ejection fractions were evaluated using MUGA scans. A cardiac event was defined as the onset of congestive heart failure or a decline in left ventricular ejection fraction exceeding 20 percentage points from baseline |
|
study 1: 088001 |
58*(26–84) | 56*(25–82) | 168 | 181 | ≥ 300 | ≥ 300 | Dexrazoxane | 1000 | 18 | Cardiac evaluation included a baseline physical examination, electrocardiography, and assessment of resting left ventricular ejection fraction via multigated acquisition nuclear scanning |
| study2: 088006 | 56*(35–76) | 59 5*(23–79) | 81 | 104 | ≥ 300 | ≥ 300 | 500 | |||
| Tallarico (2003) [35] | 55.87(12.4) | X | 36 | X |
DOX 180 mg/mL, EPI 960 mg/m |
x | Dexrazoxane and TMZ | TMZ, 60 DEX 100 mg/m2(IV) | Doppler echocardiography is utilized to evaluate the diastolic function by measuring key parameters such as E-wave velocity, A-wave velocity, and isovolumetric relaxation time | |
| 61.86(10.09) | 38 | TMZ | 60 | |||||||
| 47.25(12.31) | 38 | Dexrazoxane | DEX 100 mg/m2(IV) | |||||||
| Tashakori (2016) [36] | 42*(29–54) | 39.9*(29–54) | 45 | 45 | 240 | 240 | Carvedilol (non-selective beta-blocker) | 6.25 mg twice daily | Cardiac Monitoring and ECG | |
| Venturini (1996) [37] | 57*(32–73) | 57*(34–74) | 84 | 78 | 702 | 713 | Dexrazoxane | 500 | Cardiac assessment (ECG, chest x-rays, and MUGA scan | |
| Vici (1998) [38] | 55*(26–71) | 58*(33–72) | 45 | 50 | 960 | 960 | Dexrazoxane | 1000 | Cardiac monitoring (MUGA Scan and RIS), LVEF/Survival rate | |
Data are presented as mean ± standard deviation unless stated otherwise, with * indicating median (interquartile range)
SD standard deviation, ECG electrocardiogram, MUGA multiple-gated acquisition, RNCS radionuclide cardiac scan, MRI magnetic resonance imaging, LVEF left ventricular ejection fraction, BNP brain natriuretic peptide, X data not available, TMZ trimetazidine
Risk of bias
Overall, the RCTs demonstrated a low risk of selective reporting bias and other potential sources of bias. Most studies demonstrated a low overall risk of bias, with only two reporting a moderate risk across three of the seven assessed domains [26, 30]. One notable limitation was the insufficient reporting of outcome assessment blinding (detection bias) in most studies. Moreover, no RCT explicitly described the process of allocation concealment (selection bias), resulting in uncertainty regarding the possible impact of this bias. A detailed assessment of the risk of bias for each study is provided in Fig. 3.
Fig. 3.
Risk of bias evaluation for randomized controlled trials included in the quantitative analysis
Results of individual studies and synthesis
Figure 2 presents the complete NMA results. Direct and indirect comparisons were integrated to assess the comparative effectiveness of cardioprotective agents. The network plot highlighted that control and β-blockers were the most frequently investigated interventions. Mineralocorticoid receptor antagonists (MRAs) (−1.78, 95%CI: −2.81 to −0.75) and trimetazidine (−1.12, 95%CI: −2.32 to −0.09) exhibited the most pronounced cardioprotective effects relative to the control group. Dexrazoxane (−0.53, 95%CI: −1.90 to −0.02) and β-blockers (−0.34, 95%CI: −0.70 to 0.02) showed potential benefits; however, their wide credible intervals indicated greater uncertainty in effect estimates. Direct comparisons indicated that dexrazoxane demonstrated significantly greater efficacy than β-blockers (SMD: −1.25, 95%CI: −2.22 to −0.48). Additionally, MRAs were found to be more effective than both β-blockers and ACEIs.
However, the wide predictive intervals in several comparisons suggest substantial between-study heterogeneity, necessitating careful interpretation of findings. Indirect comparisons showed no statistically significant differences between ARBs and β-blockers (0.11, 95%CI:−0.51 to 0.74), whereas MRAs appeared more effective than dexrazoxane (−1.25, 95%CI: −2.42 to −0.08). Additionally, the comparison-adjusted funnel plot indicated potential publication bias, suggesting that smaller or negative studies may be underrepresented in the literature.
Table 3 presents a summary of the changes in LVEF across RCTs included in this NMA (primary outcome), evaluated using echocardiography, cardiac magnetic resonance imaging (MRI), and multigated radionuclide angiography (MUGA).
Table 3.
Primary outcomes: left ventricular ejection fraction in randomized controlled trials incorporated in the network meta-analysis
| Author (year) | Treatment Group | Pre-Treatment Mean | Pre-Treatment SD | Post-Treatment Mean | Post-Treatment SD |
|---|---|---|---|---|---|
| Echocardiography | LVEF | LVEF | LVEF | LVEF | |
| Akpek (2015) [13] (Transthoracic) | MRAs | 67 | 6.1 | 65.7 | 7.4 |
| Control | 67.7 | 6.3 | 53.6 | 6.8 | |
| Avila (2018) [14, 15] (Transthoracic) | Beta-blocker | 64.8 | 4.7 | 63.9 | 3.8 |
| Control | 65.2 | 3.6 | 63.9 | 5.2 | |
| Cochera (2018) [16] | Beta-blocker | 62 | 4 | 61 | 3 |
| Control | 61 | 2 | 60 | 3 | |
| Elitok (2014) [16] | Beta-blocker | 66 | 6.1 | 64.1 | 5.1 |
| Control | 65 | 4.5 | 63.3 | 4.8 | |
| Kaya (2013) [22] | Beta-blocker | 65.6 | 4.8 | 63.8 | 3.9 |
| Control | 66.6 | 5.5 | 57.5 | 5.6 | |
| Lee (2021) [23] | ARB | 64.5 | (63.7,65.2) * | 63.4 | (61.1–65.9) |
| Beta-blocker | 65.0 | (63.0–66.1) * | 62.9 | (61.0–64.0) * | |
| Control | 64.0 | 62.8–66.7* | 60.2 | 58.4–63.0* | |
| Nabati (2017) [26] | Beta-blocker | 58.7 | 4.7 | 57.4 | 7.52 |
| Control | 61.13 | 4.9 | 51.7 | 6.0 | |
| Slowik (2020) [27] | ACEI | 66.5 | 3.8 | 63.9 | 3.8 |
| Control | 67.8 | 3.8 | 64.5 | 3.8 | |
| Sun (2015) [31] | Dexrazoxane | 63 | 18 | 64 | 16 |
| Control | 65 | 8 | 65 | 13 | |
| Tallarico (2003) [35] | TMZ | 63.6 | 11.2 | 61.3 | 11.6 |
| Dexrazoxane | 60.5 | 9.3 | 50.8 | 13.6 | |
| Tashakori (2016) [36] | Beta-blocker | 61.3 | 3.2 | 61.1 | 3.4 |
| Control | 59.4 | 4.2 | 59.3 | 4.3 | |
| Cardiac MRI | |||||
|
Gulati (2016) [19] Heck (2018) [21] |
ARB | 62.5 | 4.4 | 61.6 | 4.3 |
| Beta-blocker | 63.3 | 4.3 | 60.8 | 4.4 | |
| Control | 63.1 | 4.4 | 60.3 | 4.7 | |
| MUGA scan | |||||
| Vici (1998) [38] | Dexrazoxane | 63.4 | 9.6 | ||
| Control | 58.1 | 8.6 |
Data are reported as mean ± standard deviation unless otherwise specified. * = median (interquartile range)
SD standard deviation, MUGA multiple-gated acquisition, MRI magnetic resonance imaging, LVEF left ventricular ejection fraction
The included studies reported varying degrees of LVEF preservation among different cardioprotective agents, including β-blockers (e.g., carvedilol, nebivolol), ARBs (e.g., candesartan), ACEIs, MRAs (e.g., spironolactone), and dexrazoxane. β-blockers and ARBs were associated with significant cardioprotective effects, as observed in studies by Avila et al. [14, 15], Kaya et al. [22], and Lee et al. [23]. Similarly, spironolactone [13] and ACEIs [27] showed protective effects by attenuating LVEF decline compared to controls. Dexrazoxane, evaluated in multiple trials [31, 35, 38], exhibited robust protective effects, effectively preventing substantial LVEF reductions.
The PRADA trial [19, 21] further reinforced the role of ARBs and β-blockers in preserving cardiac function. Collectively, these findings highlight the efficacy of cardioprotective agents in reducing chemotherapy-induced cardiotoxicity, underscoring the clinical significance of prophylactic interventions in high-risk patients.
Secondary outcomes
Secondary outcomes of this NMA included adverse events associated with cardioprotective interventions, functional cardiac outcomes, and epidemiological endpoints (Table 4). Adverse event analysis showed that ACEIs, ARBs, β-blockers, and dexrazoxane were generally well tolerated. However, dexrazoxane was associated with hematologic toxicity, including neutropenia and myelosuppression, whereas ACEIs frequently caused hypotension and dry cough. Functional cardiac assessments indicated that β-blockers and ARBs effectively preserved hemodynamic stability and autonomic function, as reflected in improved heart rate, blood pressure, and electrocardiographic parameters. Epidemiological outcomes further highlighted the cardioprotective benefits of dexrazoxane, including significantly reduced heart failure incidence and cardiac adverse events with improved overall survival.
Table 4.
Secondary outcomes of randomized controlled trials incorporated in the network meta-analysis
| a. Reported Adverse Events in the Included Studies | ||
|---|---|---|
| Treatment group | Study | Adverse event(s) |
| ACEI | Mross (1993) [25] | Hair loss, moderate nausea/vomiting, stomatitis/mucositis, hypotension |
| Słowik (2020) [27] | Dry cough | |
| ARB | Lee (2021) [23] | GI trouble, Dizziness, palpitation, Symptomatic hypotension |
| BB | Avila (2018) [14, 15] | Dizziness, Symptomatic hypotension |
| Kaya (2013) [22] | Hypotension and bradycardia | |
| Nabati (2017) [26] | Hypotension | |
| Lee (2021) [23] | GI trouble, Symptomatic hypotension | |
| DEX | Marty (2006) [24] | Alopecia, nausea, neutropenia, vomiting, leukopenia, anemia, and febrile neutropenia |
| Speyer (1988) [28] | Decreased WBC/cell count/platelet/hematocrit, and stomatitis/n/v/infection/alopecia (Myelosuppression slightly more in the intervention group) | |
| Speyer (1990) [29] | Hematologic toxicity (Neutropenia) | |
| Speyer (1992) [30] | Alopecia, n/v, stomatitis, Hematologic toxicity (Decreased PLT/and WBC) | |
| Venturini (1996) [37] | mild hematologic toxicity | |
| Vici (1998) [38] | Neutropenic fever/Cardiac outcomes/N/V | |
| b Functional Outcomes as Secondary Outcomes in Studies Included in the Meta-Analysis | |||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author (year) | HR | SBP | DBP | ECG (RHR) | ECG (LF/HF) | ||||||||||||||||
| Before | After | Before | After | Before | After | Before | After | before | After | Before | After | ||||||||||
| Akpek (2015) [13] | intervention | 75 | 10 | 76 | 13 | 113 | 14 | 113 | 13 | 75 | 8 | 76 | 7 | ||||||||
| Control | 75 | 11 | 77 | 10 | 111 | 15 | 112 | 15 | 75 | 8 | 75 | 8 | |||||||||
| Avila (2018) [14, 15] | intervention | 80 | 14.1 | 78 | 0 | 120.3 | 16.6 | 110 | 0 | 77.9 | 11.9 | 71 | 0 | ||||||||
| Control | 82.4 | 12.6 | 86 | 0 | 124.8 | 17.2 | 120 | 0 | 78.4 | 10.2 | 75 | 0 | |||||||||
|
Sun (2016) [32] |
intervention | 131.6 | 5.45 | 129.7 | 5.1 | 79.4 | 5.6 | 79.6 | 5.3 | 77.9 | 14.4 | 84.4 | 12.2 | 2.1 | 0.4 | 2.6 | 0.4 | ||||
| Control | 128.8 | 6.25 | 130.0 | 5.4 | 81.2 | 5.85 | 81.3 | 6.1 | 75.0 | 12.6 | 92.5 | 14.6 | 2.0 | 0.3 | 3.3 | 1.2 | |||||
| c Secondary Outcomes Reported in the Included Meta-Analysis Studies | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Author(s) | Completed intervention size | Completed control size | All sample (after) | All sample (before) | Cardiac adverse events% | HF (%) | The median survival times (months) | A survival analysis (hazard ratio) | |||
| Intervention | Control | Intervention | Control | Intervention | Control | ||||||
| Marty et al. (2006) [24] | 36 | 43 | 79 | 164 | 13% (95% CI 6% to 22%) | 39%, (95% CI 28% to 51%) | 1%, (95% CI 0.032% to 7%) | 11%, (95% CI 5% to 20%) | 13.5 (95% CI 0.2 to 27.8 +) | 16.0 (95% CI 0.5 to 25.3) | |
| Mross et al. (1993) [25] | 26 | 25 | 51 | 51 | 8.9 (95% CI 7.5 to 10.5) | 7.4 (95% CI 5.9 to 12.6) | |||||
| Speyer et al. (1990) [29] | 65 | 70 | 135 | 135 | 6% | 47% | |||||
| Speyer et al. (1992) [30] | 0 | 150 | 8% | 50% | 18.3 | 16.7 | |||||
| Swain (1997) [33, 34] | 102 | 99 | 201 | 211 | 25% | 60% | 3% | 22% | 2.2 (95% CI, 1.4 to 3.4) | ||
| Swain (1997) [33, 34] | 416 | 534 | |||||||||
| Study 1: 088001 | 141 | 152 | 293 | 15% | 31% | 0% | 8% | 1.02 (95% CI, 0.80 to 1.31) | |||
| Study 2: 088006 | 54 | 69 | 123 | 14% | 31% | 2% | 7% | 0.82 (95% CI. 0.59 to 1.14) | |||
| Venturini (1996) [37] | 82 | 78 | 160 | 162 | 2.40% | 5.10% | |||||
ACEI angiotensin-converting enzyme inhibitors, ARB angiotensin II receptor blockers, BB β−1 adrenergic receptor blockers, DEX dexrazoxane, HR heart rate, SBP systolic blood pressure, DBP diastolic blood pressure, ECG electrocardiography
Assessment of inconsistency
Inconsistency network models were applied to assess the agreement between direct and indirect estimates in pairwise and multi-arm comparisons. The assumption of overall consistency was upheld for all treatments (p > 0.05), except for MRAs, which displayed significant inconsistency (p < 0.01). The comparison of direct and indirect estimates for ARBs versus control and ARBs versus β-blockers demonstrated no significant inconsistency (p > 0.05), indicating that indirect estimates were consistent with direct evidence. However, owing to limited data availability, direct–indirect comparisons were not feasible for other treatment comparisons and were instead considered within the GRADE assessment framework.
Loop-specific heterogeneity was assessed using an inconsistency plot. Inconsistency factors (IFs) were reported exclusively for the ARB-β-blockers-Control loop, yielding a calculated IF of 0.29, suggesting no significant inconsistency. Data for assessment in other loops was insufficient, and the consistency assumption was maintained and incorporated into the GRADE evaluation. Comprehensive results from the inconsistency analyses and corresponding plots are presented in Supplementary Fig. S1a, whereas contribution plots depicting the network of cardioprotective effects are shown in Supplementary Fig. S1b.
Assessment of publication bias and sensitivity analyses
Figure 4 presents funnel plots assessing publication bias in the primary outcome. A slight indication of publication bias was observed, with one outlier suggesting small-study effects [22]. However, a sensitivity analysis excluding this outlier yielded a symmetrical funnel plot, indicating that small-study effects were unlikely to have significantly influenced the overall findings. The comparison-adjusted funnel plot further evaluated publication bias and heterogeneity among cardioprotective drug studies. While most studies clustered around the pooled effect, asymmetry suggested potential bias or variability, particularly in ARBs and MRAs. In contrast, comparisons involving control versus dexrazoxane, ACEIs, and β-blockers appeared more consistent. The presence of small-study effects suggested some variability in study precision. Despite minor asymmetry, the overall distribution supports the robustness of the meta-analysis, warranting further statistical tests to confirm and adjust for bias.
Fig.4.
Comparison-Adjusted Funnel Plot for Cardioprotective Agents. The red vertical line denotes the null hypothesis, signifying that individual effect size estimates do not significantly deviate from the pooled estimates for each comparison. Abbreviations: ACEI angiotensin-converting enzyme inhibitors, ARB angiotensin II receptor blockers, BB β1-adrenergic receptor blockers, MRA Mineralocorticoid receptor antagonists, DEX dexrazoxane, TMZ trimetazidine, CON control
Evaluation of certainty of evidence using the GRADE approach
The GRADE assessment (Table 5) evaluated the certainty of evidence for cardioprotective interventions, considering key factors such as risk of bias, inconsistency, imprecision, and indirectness. Comparisons of ARBs, β-blockers, and MRAs against control were classified as moderate certainty evidence, with β-blockers and ARBs consistently maintaining a moderate confidence level.
Table 5.
Summary of the grade assessment for certainty of evidence
| Comparison effect | Number of participants | Number of direct comparisons | Nature of evidence | Certainty | Reason for downgrading |
|---|---|---|---|---|---|
| ACEI vs. Control | 48 vs. 48 | 1 | Mixed | Very Low | Risk of biasa, inconsistencyb, imprecisionc |
| ARB vs. CON | 112 vs. 73 | 2 | Mixed | Moderate | Imprecision |
| BB vs. CON | 369 vs. 342 | 8 | Mixed | Moderate | Imprecision |
| DEX vs. CON | 272 vs. 223 | 3 | Mixed | Very Low | Risk of biasa, inconsistencyb, imprecisionc |
| DIU vs. CON | 43 vs. 40 | 1 | Mixed | Moderate | inconsistencyb |
| TMZ vs. CON | 0 vs. 0 | 0 | Indirect | Very Low | inconsistencyb, imprecisiond |
| ARB vs. ACEI | 0 vs. 0 | 0 | Indirect | Very Low | inconsistencyb, imprecisiond |
| BB vs. ACEI | 0 vs. 0 | 0 | Indirect | Very Low | inconsistencyb, imprecisiond |
| DEX vs. ACEI | 0 vs. 0 | 0 | Indirect | Very Low | inconsistencyb, imprecisiond |
| DIU vs. ACEI | 0 vs. 0 | 0 | Indirect | Very Low | inconsistencyb, imprecisiond |
| TMZ vs. ACEI | 0 vs. 0 | 0 | Indirect | Very Low | inconsistencyb, imprecisiond |
| BB vs. ARB | 100 vs. 112 | 2 | Mixed | Moderate | imprecisionc |
| DEX vs. ARB | 0 vs. 0 | 0 | Indirect | Very Low | inconsistencyb, imprecisiond |
| DIU vs. ARB | 0 vs. 0 | 0 | Indirect | Very Low | inconsistencyb, imprecisiond |
| TMZ vs. ARB | 0 vs. 0 | 0 | Indirect | Very Low | inconsistencyb, imprecisiond |
| DEX vs. BB | 0 vs. 0 | 0 | Indirect | Very Low | inconsistencyb, imprecisiond |
| DIU vs. BB | 0 vs. 0 | 0 | Indirect | Very Low | inconsistencyb, imprecisiond |
| TMZ vs. BB | 0 vs. 0 | 0 | Indirect | Very Low | inconsistencyb, imprecisiond |
| DIU vs. DEX | 0 vs. 0 | 0 | Indirect | Very Low | inconsistencyb, imprecisiond |
| TMZ vs. DEX | 22 vs. 24 | 1 | Mixed | Very Low | Risk of biasa, inconsistencyb, imprecisionc |
| TMZ vs. DIU | 0 vs. 0 | 0 | Indirect | Very Low | inconsistencyb, imprecisiond |
aPotential risk of bias due to performance bias. bPotential inconsistency due to not enough data
cConfidence intervals encompass values supporting either treatment
dConfidence intervals encompass values supporting either treatment; limited sample size
ACEI angiotensin-converting enzyme inhibitors, ARB = angiotensin II receptor blockers; BB β−1 adrenergic receptor blockers; MRA, DEX dexrazoxane, TMZ trimetazidine, CON control
However, certain comparisons including ACEIs and dexrazoxane versus control were downgraded to very low certainty. This downgrade was primarily attributed to methodological limitations, substantial heterogeneity, and wide confidence intervals. Indirect comparisons, particularly those without direct supporting studies (e.g., ARB vs. ACEI, dexrazoxane vs. β-blockers, MRAs vs. β-blockers), were further constrained by inconsistency and imprecision.
Direct head-to-head comparisons between cardioprotective agents generally exhibited very low certainty, except for the β-blockers vs. ARBs comparison, which demonstrated moderate certainty based on data from 100 and 112 participants, respectively. The GRADE framework identified significant gaps in current evidence, warranting RCTs to rigorously determine the efficacy and safety of cardioprotective strategies in reducing chemotherapy-induced cardiotoxicity [12].
Discussion
This NMA constitutes the most comprehensive evaluation of pharmacological interventions aimed at preventing anthracycline-induced cardiotoxicity in patients with breast cancer to date. Cardiotoxicity remains a major contributor to morbidity and mortality among cancer survivors, particularly following treatment with anthracycline-based chemotherapy [39]. The persistent risk of long-term cardiac dysfunction underscores the need for effective cardioprotective interventions. The findings of this analysis suggested that ACEIs, ARBs, β-blockers, trimetazidine, dexrazoxane, and MRAs provide significant cardioprotective effects compared to control treatments. However, the limited availability of direct head-to-head comparisons among these interventions impacts the overall certainty of the evidence. Notably, β-blockers and ARBs were exceptions, demonstrating evidence of moderate certainty. Furthermore, the generalizability of these results is constrained by the predominance of participants with normal LVEF in included studies, limiting their applicability to broader breast cancer populations, particularly those with pre-existing cardiac impairment.
Despite these promising results, substantial between-study heterogeneity was observed, as reflected in the wide predictive intervals across several comparisons, likely attributed to differences in study design, patient populations, chemotherapy regimens, and follow-up durations. Furthermore, indirect comparisons indicated no significant differences between ARBs and β-blockers, whereas MRAs appeared to be more effective than dexrazoxane, suggesting that specific interventions may provide greater benefits in certain patient subgroups.
Additionally, the comparison-adjusted funnel plot revealed potential publication bias, indicating that smaller or negative studies may be underrepresented in the literature. Publication bias can distort treatment rankings, underscoring the need for large-scale RCTs with standardized methodologies. Future research should minimize heterogeneity by implementing standardized patient selection criteria, anthracycline regimens, and cardiotoxicity assessment methods to enhance reliability and reproducibility.
Safety and secondary outcomes
The secondary outcome analysis provided valuable insights into drug safety profiles and functional outcomes. While most treatments were well tolerated, dexrazoxane was associated with hematologic toxicity, including neutropenia and myelosuppression, whereas ACEIs Cardioprotective Efficacy of Pharmacological Interventions.
Among the evaluated interventions in the NMA, MRAs exhibited the greatest efficacy, demonstrating a significant advantage over control treatments in LVEF in patients with breast cancer receiving anthracycline-based chemotherapy. Similarly, trimetazidine exhibited notable cardioprotective effects, although its role in routine clinical practice remains underexplored [40]. The protective effects of β-blockers and dexrazoxane in preserving LVEF align with the findings from prior observational studies examining the cardioprotective role of β-blockers in mitigating chemotherapy-induced cardiotoxicity [22, 41], contributing further evidence supporting the effectiveness of prophylactic β-blockers and dexrazoxane in preventing LVEF decline during chemotherapy.
Trimetazidine was included in this analysis based on a single small randomized controlled trial [40], and its apparent cardioprotective effect should be interpreted with caution. The limited sample size and absence of independent validation restrict the generalizability of these findings. Moreover, its primary mechanism—enhancing myocardial energy efficiency—does not directly target the established pathways of anthracycline-induced cardiotoxicity, such as oxidative stress and mitochondrial dysfunction. As such, the current evidence remains hypothesis-generating, and robust, mechanistically driven trials are warranted to clarify its potential role in cardio-oncology. In contrast, angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin receptor blockers (ARBs) demonstrated significant cardioprotective effects, aligning with prior evidence supporting their role in mitigating anthracycline-associated cardiac injury [14]. β-blockers and ARBs were among the most extensively studied agents; however, variability in efficacy across trials highlights the need for additional well-powered studies to refine optimal dosing strategies and treatment durations. Continued collaborative efforts in this area are essential to inform evidence-based cardioprotective protocols for patients undergoing anthracycline-based chemotherapy.
Notable findings include the strong protective effects of β-blockers and dexrazoxane, which significantly outperformed control treatments in preserving LVEF and reducing cardiac dysfunction risk. These findings suggest that early initiation of β-blockers, particularly carvedilol and nebivolol, may substantially benefit high-risk patients receiving anthracyclines [6]. Similarly, dexrazoxane remains one of the most effective agents in reducing cardiotoxicity; however, concerns regarding its hematologic toxicity have limited widespread clinical adoption [42].
frequently caused hypotension and dry cough [25, 27]. Nevertheless, β-blockers and ARBs effectively maintained hemodynamic stability and autonomic function, providing additional cardiovascular benefits beyond LVEF preservation [14]. Importantly, DEX significantly reduced heart failure incidence and cardiac adverse events while improving overall survival, reinforcing its role as a key prophylactic strategy despite its hematologic risks [24, 37].
Strengths and limitations
A key strength of this NMA is the high methodological quality of the included studies, minimizing the risk of bias and enhancing result reliability.
Additionally, this analysis reduces gender-related confounding effects and ensures homogeneity in treatment protocols by focusing exclusively on patients with breast cancer. However, several limitations must be acknowledged. Despite an extensive and comprehensive search strategy, only 15 RCTs fulfilled the eligibility criteria, which may constrain the breadth of available evidence and highlight the need for further high-quality trials in this area. In addition, the relatively small sample sizes and variability in study designs across the included trials may influence the overall strength and certainty of the evidence, potentially limiting the conclusiveness of some comparisons. The limited sample sizes in individual studies may contribute to bias, reducing statistical power and generalizability. Moreover, variations in anthracycline dosing, concurrent cancer treatments, and follow-up durations introduce potential confounding factors. Differences in treatment regimens, baseline cardiovascular risk, and cardiotoxicity assessment methods contribute to the observed heterogeneity. Furthermore, publication bias and methodological inconsistencies across RCTs cannot be ruled out, as studies with negative or inconclusive results may be underrepresented. While most cardioprotective interventions demonstrated efficacy, β-blockers, ACEIs, ARBs, and antimitotics did not consistently show protective effects, likely owing to heterogeneity in study populations and unmeasured confounders. Nonetheless, despite these limitations, consistent trends toward cardioprotection—particularly in preserving left ventricular (LV) function—were observed across multiple interventions, reinforcing the clinical relevance of these findings.
Clinical and research implications
These findings underscored the importance of tailoring cardioprotective strategies based on efficacy, safety profiles, and patient-specific risk factors. The consistent cardioprotective effects demonstrated by angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin receptor blockers (ARBs) reinforce their established role in mitigating anthracycline-induced cardiotoxicity [14]. While β-blockers and ARBs have been extensively investigated, the observed variability in efficacy across trials highlights the necessity for further well-powered studies to elucidate the optimal dosing regimens and treatment durations.
Intriguingly, agents such as MRAs and trimetazidine exhibited notable efficacy, suggesting that metabolic and fluid-regulating interventions may hold a more prominent role in preventing chemotherapy-induced cardiotoxicity than previously recognized [43]. However, the inclusion of trimetazidine in this analysis was based on a single, small randomized controlled trial (RCT) [40]. Despite a seemingly substantial effect size, the limited evidence base and lack of clear mechanistic relevance to anthracycline-induced cardiotoxicity render these specific findings hypothesis-generating. This warrants rigorous validation through robust, independent clinical trials. Furthermore, the overall observed heterogeneity across studies and the potential for publication bias necessitate cautious interpretation, particularly when considering application to populations with pre-existing cardiovascular risk factors [6].
These results, while informative, should be considered largely hypothesis-generating and underscore an urgent need for future well-designed, adequately powered randomized controlled trials (RCTs). Such trials are imperative to validate these preliminary findings, refine optimal treatment rankings, and inform the development of evidence-based clinical guidelines. Future direct head-to-head RCTs are warranted to validate these findings, refine treatment rankings, and optimize cardioprotective strategies for clinical practice. Additionally, future research should explore optimal dosing regimens, combination therapies, and long-term cardiac outcomes to ensure sustained cardioprotection in cancer survivors. Standardized definitions of cardiotoxicity, echocardiographic assessment methods, and long-term follow-up are critical for improving comparability across studies and strengthening the evidence base for clinical decision-making.
Conclusion
The study findings support the superiority of cardioprotective drugs over control treatments in mitigating anthracycline-induced cardiotoxicity. However, the absence of direct head-to-head comparisons among these cardioprotective interventions in RCTs limits the ability to establish a definitive hierarchical ranking of their therapeutic efficacy.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
Pinyadapat Vacharanukrauh received support from an Australian Government Research Training Program (RTP) Fee Offset Scholarship through Federation University Australia. The sponsors did not participate in the study design, data analysis, interpretation, manuscript writing, or the decision to submit the manuscript for publication.
Author contributions
The study was initiated by PV and FG, who came up with the idea. PV, FG, and KJM planned the study. During the PROSPERO registration process, NK and FG handled the registration of the meta-analysis study [Registration No: CRD42020168609]. PV and KJM conducted screening, data extraction, and data analysis, while FG was responsible for resolving any discrepancies found in the articles. NK conducted the data analysis, and FG and KJM confirmed the findings. PV and FG were responsible for interpreting the findings and data. KJM supervised PV’s manuscript drafting, KJM and AR contributed to the final draft. All co-authors contributed to the intellectual content and approved the final manuscript.
Funding
Open Access funding enabled and organized by CAUL and its Member Institutions. Financial interests: Pinyadapat Vacharanukrauh received support from an Australian Government Research Training Program (RTP) Fee Offset Scholarship through Federation University Australia.
Data availability
The data utilized in this network meta-analysis were derived exclusively from previously published studies identified through a comprehensive and systematic literature search. All relevant summary data extracted from the included studies are presented in the Supplementary Materials, where applicable, to promote transparency and facilitate reproducibility. No individual participant data or proprietary datasets were used in this analysis. The datasets supporting the findings of this study are available from the corresponding author upon reasonable request. As this study is based solely on published data, ethical approval was not required.
Declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Ethical approval
This study is a systematic review and meta-analysis and did not involve human participants or animals.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Lyon AR, López-Fernández T, Couch LS, Asteggiano R, Aznar MC, Bergler-Klein J, Boriani G, Cardinale D, Cordoba R, Cosyns B, Cutter DJ, de Azambuja E, de Boer RA, Dent SF, Farmakis D, Gevaert SA, Gorog DA, Herrmann J, Lenihan D, Moslehi J, Moura B, Salinger SS, Stephens R, Suter TM, Szmit S, Tamargo J, Thavendiranathan P, Tocchetti CG, van der Meer P, van der Pal HJH, ESC Scientific Document Group (2022) 2022 ESC guidelines on cardio-oncology developed in collaboration with the european hematology association (EHA), the European society for therapeutic radiology and oncology (ESTRO) and the international cardio-oncology society (IC-OS). Eur Heart J Cardiovasc Imaging 23(10):e333–e465. 10.1093/ehjci/jeac106 [DOI] [PubMed] [Google Scholar]
- 2.Zamorano JL, Lancellotti P, Rodriguez Muñoz 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) 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 37(36):2768–2801. 10.1093/eurheartj/ehw211 [DOI] [PubMed] [Google Scholar]
- 3.Plana JC, Galderisi M, Barac A, Ewer MS, Ky B, Scherrer-Crosbie M, Ganame J, Sebag IA, Agler DA, Badano LP, Banchs J, Cardinale D, Carver J, Cerqueira M, DeCara JM, Edvardsen T, Flamm SD, Force T, Griffin BP, Jerusalem G, Liu JE, Magalhães A, Marwick T, Sanchez LY, Sicari R, Villarraga HR, Lancellotti P (2014) Expert consensus for multimodality imaging evaluation of adult patients during and after cancer therapy: a report from the American society of echocardiography and the European association of cardiovascular imaging. Eur Heart J Cardiovasc Imaging 15(10):1063–1093. 10.1093/ehjci/jeu192 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.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: clinicalguidelines@esmo.org (2020) Management of cardiac disease in cancer patients throughout oncological treatment: ESMO consensus recommendations. Ann Oncol 31(2):171–190. 10.1016/j.annonc.2019.10.023 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Huang S, Zhao Q, Yang Z-G, Diao K-Y, He Y, Shi K, Shen M-T, Fu H, Guo Y-K (2019) Protective role of beta-blockers in chemotherapy-induced cardiotoxicity—A systematic review and meta-analysis of carvedilol. Heart Fail Rev 24(3):325–333. 10.1007/s10741-018-9755-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Armenian SH, Lacchetti C, Barac A, Carver J, Constine LS, Denduluri N, Dent S, Douglas PS, Durand J-B, Ewer M, Fabian C, Hudson M, Jessup M, Jones LW, Ky B, Mayer EL, Moslehi J, Oeffinger K, Ray K, Ruddy K, Lenihan D (2017) Prevention and monitoring of cardiac dysfunction in survivors of adult cancers: American society of clinical oncology clinical practice guideline. J Clin Oncol 35(8):893–911. 10.1200/JCO.2016.70.5400 [DOI] [PubMed] [Google Scholar]
- 7.Kalam K, Marwick TH (2013) Role of cardioprotective therapy for prevention of cardiotoxicity with chemotherapy: a systematic review and meta-analysis. Eur J Cancer 49(13):2900–2909. 10.1016/j.ejca.2013.04.030 [DOI] [PubMed] [Google Scholar]
- 8.Nebigil CG, Désaubry LJF (2018) Updates in anthracycline-mediated cardiotoxicity. Front Pharmacol 9:1262. 10.3389/fphar.2018.01262 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Hutton B, Catalá-López F, Moher DJMC (2016) The PRISMA statement extension for systematic reviews incorporating network meta-analysis: PRISMA-NMA. Med Clin (Barc) 147(6):262–266. 10.1016/j.medcli.2016.02.025 [DOI] [PubMed] [Google Scholar]
- 10.Jansen JP, Naci HJB m (2013) Is network meta-analysis as valid as standard pairwise meta-analysis? It All Depends Distrib Eff Mod 11(1):1–8 [DOI] [PMC free article] [PubMed]
- 11.StataCorp, L (2017) Stata statistical software, release 15. College Station, TX
- 12.Salanti G, Del Giovane C, Chaimani A, Caldwell DM, Higgins JPT (2014) Evaluating the quality of evidence from a network meta-analysis. PLoS ONE 9(7):e99682. 10.1371/journal.pone.0099682 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Akpek M, Ozdogru I, Sahin O, Inanc M, Dogan A, Yazici C, Berk V, Karaca H, Kalay N, Oguzhan A, Ergin A (2015) Protective effects of spironolactone against anthracycline-induced cardiomyopathy. Eur J Heart Fail 17(1):81–89. 10.1002/ejhf.196 [DOI] [PubMed] [Google Scholar]
- 14.Avila MS, Ayub-Ferreira SM, de Barros Wanderley MR, das Dores Cruz F, Goncalves Brandao SM, Rigaud VO, Higuchi-dos-Santos MH, Hajjar LA, Kalil Filho R, Hoff PM, Sahade M (2018) Carvedilol for prevention of chemotherapy-related cardiotoxicity: the CECCY trial. J Am Coll Cardiol 71(20):2281–2290 [DOI] [PubMed] [Google Scholar]
- 15.Avila MS, Ayub-Ferreira SM, de Barros Wanderley MR, Jr., das Dores Cruz F, Gonçalves Brandão SM, Rigaud VOC, Higuchi-Dos-Santos MH, Hajjar LA, Kalil Filho R, Hoff PM, Sahade M, Ferrari MSM, de Paula Costa RL, Mano MS, Bittencourt Viana Cruz CB, Abduch MC, Lofrano Alves MS, Guimaraes GV, Issa VS, Bittencourt MS, Bocchi EA (2018) Carvedilol for prevention of chemotherapy-related cardiotoxicity: The CECCY trial. J Am Coll Cardiol 71(20):2281–2290. 10.1016/j.jacc.2018.02.049 [DOI] [PubMed]
- 16.Cochera F, Dinca D, Bordejevic DA, Citu IM, Mavrea AM, Andor M, Trofenciuc M, Tomescu MCJC (2018) Nebivolol effect on doxorubicin-induced cardiotoxicity in breast cancer. Cancer Manag Res 10:2071–2081. 10.2147/CMAR.S166481 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Davis MK, Villa D, Tsang TSM, Starovoytov A, Gelmon K, Virani SA (2019) Effect of eplerenone on diastolic function in women receiving anthracycline-based chemotherapy for breast cancer. JACC CardioOncol 1(2):295–298. 10.1016/j.jaccao.2019.10.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Elitok A, Oz F, Cizgici AY, Kilic L, Ciftci R, Sen F, Bugra Z, Mercanoglu F, Oncul A, Oflaz HJC (2014) Effect of carvedilol on silent anthracycline-induced cardiotoxicity assessed by strain imaging: a prospective randomized controlled study with six-month follow-up. Cardiol J 21(5):509–515. 10.5603/CJ.a2013.0150 [DOI] [PubMed] [Google Scholar]
- 19.Gulati G, Heck SL, Ree AH, Hoffmann P, Schulz-Menger J, Fagerland MW, Gravdehaug B, von Knobelsdorff-Brenkenhoff F, Bratland Å, Storås TH, Hagve T-A, Røsjø H, Steine K, Geisler J, Omland T (2016) Prevention of cardiac dysfunction during adjuvant breast cancer therapy (PRADA): A 2 × 2 factorial, randomized, placebo-controlled, double-blind clinical trial of candesartan and metoprolol. Eur Heart J 37(21):1671–1680. 10.1093/eurheartj/ehw022 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Gulati G, Heck SL, Røsjø H, Ree AH, Hoffmann P, Hagve T-A, Norseth J, Gravdehaug B, Steine K, Geisler J, Omland T (2017) Neurohormonal blockade and circulating cardiovascular biomarkers during anthracycline therapy in breast cancer patients: results from the PRADA (prevention of cardiac dysfunction during adjuvant breast cancer therapy) study. J Am Heart Assoc 6(11):e006513. 10.1161/JAHA.117.006513 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Heck SL, Gulati G, Hoffmann P, von Knobelsdorff-Brenkenhoff F, Storås TH, Ree AH, Gravdehaug B, Røsjø H, Steine K, Geisler J, Schulz-Menger J, Omland T (2018) Effect of candesartan and metoprolol on myocardial tissue composition during anthracycline treatment: the PRADA trial. Eur Heart J Cardiovasc Imaging 19(5):544–552. 10.1093/ehjci/jex159 [DOI] [PubMed] [Google Scholar]
- 22.Kaya MG, Ozkan M, Gunebakmaz O, Akkaya H, Kaya EG, Akpek M, Kalay N, Dikilitas M, Yarlioglues M, Karaca H, Berk V, Ardic I, Ergin A, Lam YY (2013) Protective effects of nebivolol against anthracycline-induced cardiomyopathy: a randomized control study. Int J Cardiol 167(5):2306–2310. 10.1016/J.IJCARD.2012.06.023 [DOI] [PubMed] [Google Scholar]
- 23.Lee M, Chung WB, Lee JE, Park CS, Park WC, Song BJ, Youn HJ (2021) Candesartan and carvedilol for primary prevention of subclinical cardiotoxicity in breast cancer patients without a cardiovascular risk treated with doxorubicin. Cancer Med 10(12):3964–3973. 10.1002/cam4.3956 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Marty M, Espie M, Llombart A, Monnier A, Rapoport B, Stahalova V, Oncology (2006) Multicenter randomized phase III study of the cardioprotective effect of dexrazoxane (Cardioxane®) in advanced/metastatic breast cancer patients treated with anthracycline-based chemotherapy. Ann Oncol 17(4):614–622 [DOI] [PubMed] [Google Scholar]
- 25.Mross K, Bohn C, Edler L, Jonat W, Queisser W, Heidemann E, Goebel M, Hossfeld DK (1993) Randomized phase II study of single-agent epirubicin +/− verapamil in patients with advanced metastatic breast cancer. An AIO clinical trial. Arbeitsgemeinschaft Internistische Onkologie of the German cancer society. Ann Oncol 4(1):45–50. 10.1093/oxfordjournals.annonc.a058356 [DOI] [PubMed] [Google Scholar]
- 26.Nabati M, Janbabai G, Baghyari S, Esmaili K, Yazdani J (2017) Cardioprotective effects of carvedilol in inhibiting doxorubicin-induced cardiotoxicity. J Cardiovasc Pharmacol 69(5):279–285. 10.1097/FJC.0000000000000470 [DOI] [PubMed] [Google Scholar]
- 27.Słowik AJ, Jagielski P, Potocki P, Streb J, Ochenduszko S, Wysocki P, Gajos G, Konduracka E (2020) Anthracycline-induced cardiotoxicity prevention with angiotensin-converting enzyme inhibitor ramipril in women with low-risk breast cancer: Results of a prospective randomized study. Polish Heart J 78(2):131–137. 10.33963/KP.15163 [DOI] [PubMed] [Google Scholar]
- 28.Speyer JL, Green MD, Kramer E, Rey M, Sanger J, Ward C, Dubin N, Ferrans V, Stecy P, Zeleniuch-Jacquotte A (1988) Protective effect of the bispiperazinedione ICRF-187 against doxorubicin-induced cardiac toxicity in women with advanced breast cancer. N Engl J Med 319(12):745–752. 10.1056/NEJM198809223191203 [DOI] [PubMed] [Google Scholar]
- 29.Speyer JL, Green MD, Sanger J, Zeleniuch-Jacquotte A, Kramer E, Rey M, Wernz JC, Blum RH, Hochester H, Meyers M (1990) A prospective randomized trial of ICRF-187 for prevention of cumulative doxorubicin-induced cardiac toxicity in women with breast cancer. Cancer Treat Rev 17(2–3):161–163. 10.1016/0305-7372(90)90041-d [DOI] [PubMed] [Google Scholar]
- 30.Speyer JL, Green MD, Zeleniuch-Jacquotte A, Wernz JC, Rey M, Sanger J, Kramer E, Ferrans V, Hochster H, Meyers M (1992) ICRF-187 permits longer treatment with doxorubicin in women with breast cancer. J Clin Oncol 10(1):117–127. 10.1200/JCO.1992.10.1.117 [DOI] [PubMed] [Google Scholar]
- 31.Sun F, Qi X, Geng C, Li X (2015) Dexrazoxane protects breast cancer patients with diabetes from chemotherapy-induced cardiotoxicity [Journal Article; Randomized Controlled Trial]. Am J Med Sci 349(5):406–412. 10.1097/MAJ.0000000000000432 [DOI] [PubMed] [Google Scholar]
- 32.Sun F, Shi J, Geng C (2016) Dexrazoxane improves cardiac autonomic function in epirubicin-treated breast cancer patients with type 2 diabetes. Medicine 95(44):e5228. 10.1097/MD.0000000000005228 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Swain SM, Whaley FS, Gerber MC, Ewer MS, Bianchine JR, Gams RA (1997) Delayed administration of dexrazoxane provides cardioprotection for patients with advanced breast cancer treated with doxorubicin-containing therapy. J Clin Oncol 15(4):1333–1340. 10.1200/JCO.1997.15.4.1333 [DOI] [PubMed] [Google Scholar]
- 34.Swain SM, Whaley FS, Gerber MC, Weisberg S, York M, Spicer D, Jones SE, Wadler S, Desai A, Vogel C, Speyer J, Mittelman A, Reddy S, Pendergrass K, Velez-Garcia E, Ewer MS, Bianchine JR, Gams RA (1997) Cardioprotection with dexrazoxane for doxorubicin-containing therapy in advanced breast cancer. J Clin Oncol 15(4):1318–1332. 10.1200/JCO.1997.15.4.1318 [DOI] [PubMed] [Google Scholar]
- 35.Tallarico D, Rizzo V, Di Maio F, Petretto F, Bianco G, Placanica G, Marziali M, Paravati V, Gueli N, Meloni F, Campbell SV (2003) Myocardial cytoprotection by trimetazidine against anthracycline-induced cardiotoxicity in anticancer chemotherapy. Angiology 54(2):219–227. 10.1177/000331970305400212 [DOI] [PubMed] [Google Scholar]
- 36.Tashakori Beheshti A, Mostafavi Toroghi H, Hosseini G, Zarifian A, Homaei Shandiz F, Fazlinezhad A (2016) Carvedilol administration can prevent doxorubicin-induced cardiotoxicity: a double-blind randomized trial. Cardiology 134(1):47–53. 10.1159/000442722 [DOI] [PubMed] [Google Scholar]
- 37.Venturini M, Michelotti A, Del Mastro L, Gallo L, Carnino F, Garrone O, Tibaldi C, Molea N, Bellina RC, Pronzato P, Cyrus P, Vinke J, Testore F, Guelfi M, Lionetto R, Bruzzi P, Conte PF, Rosso R (1996) Multicenter randomized controlled clinical trial to evaluate cardioprotection of dexrazoxane versus no cardioprotection in women receiving epirubicin chemotherapy for advanced breast cancer. J Clin Oncol 14(12):3112–3120. 10.1200/JCO.1996.14.12.3112 [DOI] [PubMed] [Google Scholar]
- 38.Vici P, Ferraironi A, Di Lauro L, Carpano S, Conti F, Belli F, Paoletti G, Maini CL, Lopez M (1998) Dexrazoxane cardioprotection in advanced breast cancer patients undergoing high-dose epirubicin treatment. Clin Ter 149(921):15–20 [PubMed] [Google Scholar]
- 39.Virani SA, Dent S, Brezden-Masley C, Clarke B, Davis MK, Jassal DS, Johnson C, Lemieux J, Paterson I, Sebag IA, Simmons C, Sulpher J, Thain K, Thavendiranathan P, Wentzell JR, Wurtele N, Côté MA, Fine NM, Haddad H, Hayley BD, Hopkins S, Joy AA, Rayson D, Stadnick E, Straatman L (2016) Canadian Cardiovascular Society guidelines for evaluation and management of cardiovascular complications of cancer therapy. Can J Cardiol 32(7):831–841. 10.1016/j.cjca.2016.02.078 [DOI] [PubMed] [Google Scholar]
- 40.Cardinale D, Iacopo F, Cipolla CM (2020) Cardiotoxicity of anthracyclines. Front Cardiovasc Med 7:26. 10.3389/fcvm.2020.00026 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Liu L, Liu ZZ, Liu YY, Zheng ZD, Liang XF, Han YL, Xie XD (2013) Preventive effect of low-dose carvedilol combined with candesartan on the cardiotoxicity of anthracycline drugs in the adjuvant chemotherapy of breast cancer. Zhonghua Zhong Liu Za Zhi [Zhonghua Zhong Liu Za Zhi] 35(12):936–940 [PubMed] [Google Scholar]
- 42.Abdel-Qadir H, Ong G, Fazelzad R, Amir E, Lee DS, Thavendiranathan P, Tomlinson G (2017) Interventions for preventing cardiomyopathy due to anthracyclines: a Bayesian network meta-analysis. Ann Oncol 28(3):628–633. 10.1093/annonc/mdw671 [DOI] [PubMed] [Google Scholar]
- 43.Cardinale D, Ciceri F, Latini R, Franzosi MG, Sandri MT, Civelli M, Cucchi G, Menatti E, Mangiavacchi M, Cavina R et al (2018) Anthracycline-induced cardiotoxicity: a multicenter randomised trial comparing two strategies for guiding prevention with enalapril: The International CardioOncology Society-one trial. Eur J Cancer 94:126–137. 10.1016/j.ejca.2018.02.005 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data utilized in this network meta-analysis were derived exclusively from previously published studies identified through a comprehensive and systematic literature search. All relevant summary data extracted from the included studies are presented in the Supplementary Materials, where applicable, to promote transparency and facilitate reproducibility. No individual participant data or proprietary datasets were used in this analysis. The datasets supporting the findings of this study are available from the corresponding author upon reasonable request. As this study is based solely on published data, ethical approval was not required.