Monitoring high-risk patients for chemotherapy-related cardiotoxicity: A retrospective analysis

Abstract

BACKGROUND

Cancer incidence remains a global challenge. The World Health Organization reported 19976499 new cases in 2022, including 1551060 in Latin America and the Caribbean. While chemotherapy advances have improved survival, these treatments carry significant risks, particularly cardiovascular complications impacting morbidity and mortality. Early cardiotoxicity detection enables targeted interventions, guiding clinical decisions on treatment adjustments to mitigate damage and preserve function. Cardiac imaging and biomarkers assess cardiotoxicity before, during, and after therapy. Despite their importance, the lack of a structured multidisciplinary program hinders early detection and management in high-risk patients.

AIM

To evaluate the use of diagnostic tools for monitoring cardiotoxicity in cancer patients receiving high-risk chemotherapy at the National University Hospital of Colombia.

METHODS

This observational, retrospective cohort study included patients aged ≥ 18 with cancer treated with potentially cardiotoxic chemotherapy at the National University Hospital of Colombia (2016-2019). Data from medical records included demographics, comorbidities, biomarkers, and echocardiographic parameters. Cardiotoxicity was defined by reduced left ventricular ejection fraction (LVEF) using Simpson’s method and biomarker abnormalities. Statistical analysis included descriptive methods to compare pre- and post-chemotherapy use of biomarkers and echocardiographic parameters.

RESULTS

From a total of 195 patients analyzed, 8.7% (n = 17) developed cardiotoxicity, predominantly mild (58.8%, n = 10). Affected patients were mostly male (64.7%, n = 11) with a mean age of 51.88 ± 15.9 years. The median LVEF declined from 62% [interquartile range (IQR): 58%–66%] at baseline to 46% (IQR: 34%–56%) post-treatment. STRAIN longitudinal values also significantly decreased, from -18.38 ± 4.62% at baseline to -14.22 ± 4.93% post-treatment. Troponin was measured in 58.8% (n = 10) of cardiotoxicity cases, while ProBNP was less frequently used (17.6%, n = 3).

CONCLUSION

This study highlights the utility of echocardiography and biomarkers in assessing cardiotoxicity in oncology patients, emphasizing the need for standardized protocols to optimize early diagnosis and management. However, the retrospective nature of the study and the insufficient use of biomarkers may limit the generalizability of the findings.

Key Words: Cardiotoxicity; Antineoplastic Agents/adverse effects; Cardiovascular diseases/chemically induced; Echocardiography; Biomarkers; Drug monitoring

Core Tip: Cardiotoxicity remains a significant concern for cancer patients undergoing high-risk chemotherapy. This study was conducted at the National University Hospital of Colombia and highlights the importance of echocardiography and biomarkers in detecting chemotherapy-induced cardiac damage. Among 195 patients analyzed, 8.7% developed cardiotoxicity, with significant declines in left ventricular ejection fraction and longitudinal strain. While biomarker usage showed variability, they show clear potential in early detection. The results underscore the critical need for standardized protocols to enhance the early detection and management of cardiotoxicity, enabling timely interventions and improved patient outcomes. This study advances the understanding of cardiotoxicity risk and management, advocating for structured multidisciplinary programs to optimize care.

INTRODUCTION

Cancer remains one of the greatest challenges to global public health. According to the World Health Organization, 19976499 new cancer cases were reported in 2022, with 1551060 occurring in Latin America and the Caribbean[1]. While advances in chemotherapy have significantly improved survival rates among oncology patients, they are not exempt from adverse effects, particularly at the cardiovascular level. Chemotherapy-associated cardiac complications can substantially impact patients' quality of life and increase cancer-related mortality[2]. One of the major challenges in this context is the early detection of cardiotoxicity[3,4]. Identifying the initial stages of cardiac damage is crucial for timely clinical decision-making, allowing for treatment adjustments and targeted interventions to minimize the impact on cardiac function. To achieve this, diagnostic tools such as cardiac imaging and biomarkers enable continuous monitoring of cardiotoxic effects before, during, and after oncological therapy[5]. However, the accessibility and appropriate use of these tools remain limited in many settings, hindering their integration into routine clinical practice.

Chemotherapy-induced cardiotoxicity results from direct effects on the myocardium and interactions with predisposing factors related to the patient, underlying disease, and treatment itself[6]. Implementing a coordinated approach would facilitate the early detection of cardiac alterations in cancer patients, optimizing their management and reducing long-term complications. Many healthcare systems lack structured programs to address cardiotoxicity, leaving patients vulnerable to chemotherapy-induced cardiovascular risk. Notably, adherence to clinical practice guidelines, such as those for acute heart failure management, can reduce readmission and preventable hospitalizations by up to 30%[7]. Among other chemotherapeutic agents, anthracyclines and monoclonal antibodies targeting HER2 have demonstrated high efficacy in treating various cancers but also pose significant cardiotoxicity risks[8]. In this regard, periodic assessment of cardiac function is essential to prevent damage progression and ensure that patients can continue treatment without compromising cardiovascular health[9]. Three-dimensional echocardiography with global longitudinal strain measurement and cardiac magnetic resonance imaging have proven useful in this field, complemented by biomarker assessment, including troponin and natriuretic peptides[10]. However, their routine use remains limited in many centers due to economic and logistical constraints. This study aims to evaluate the use of diagnostic tools for monitoring cardiotoxicity in oncology patients receiving high-risk chemotherapy at the National University Hospital of Colombia.

MATERIALS AND METHODS

Study design and patients

This observational, retrospective cohort study included patients aged 18 years or older diagnosed with cancer and treated with potentially cardiotoxic chemotherapy at the National University Hospital of Colombia between 2016 and 2019. The primary objective was to assess the use of diagnostic tools for cardiotoxicity monitoring in this population. This single-center, observational, retrospective, descriptive cohort study included all oncology patients aged 18 years or older who received chemotherapy at the National University Hospital of Colombia between 2016 and 2019 and were classified as having an intermediate, high, or very high risk of developing cardiotoxicity, according to the Mayo Clinic Cardiotoxicity Risk classification[4]. Patients were excluded if they had incomplete clinical records, if their left ventricular ejection fraction (LVEF) assessments had not been performed using the modified biplane Simpson’s method, or if their global longitudinal strain (GLS) measurements were not obtained using the same version of echocardiographic software due to the variability in GLS normality cutoffs.

Data collection

Patient medical records were reviewed to obtain relevant data, including sociodemographic variables, clinical history, comorbidities, oncologic treatment, biomarkers, and echocardiographic follow-up. Cardiotoxicity was defined as the onset or progression of myocardial injury or dysfunction during follow-up, determined by elevated troponin levels, an increase in the N-terminal pro-brain natriuretic peptide (NT-proBNP), or a reduction in LVEF. Cardiotoxicity severity was categorized as mild, moderate, or severe, according to the worst degree of injury or dysfunction observed, as adopted from the CARDIOTOX registry[7]. Different levels of myocardial involvement were defined based on cardiac biomarkers, left ventricular function parameters, and clinical heart failure symptoms. Mild dysfunction included asymptomatic patients with LVEF ≥ 50% but with elevated biomarkers or at least one abnormal echocardiographic parameter. Moderate cardiotoxicity was defined as LVEF between 40% and 50%. Severe cardiotoxicity was characterized by LVEF < 40% in asymptomatic patients or the presence of symptomatic clinical heart failure, regardless of ejection fraction, or cardiovascular-related death[7].

The biomarkers available and analyzed in the hospital included high-sensitivity serum troponin I, with a detection range of 4.0–25000 ng/L. Elevated values were defined as > 51.4 pg/dL in women and > 76.2 pg/dL in men. NT-proBNP was considered elevated if values exceeded 125 pg/mL in patients under 75 years of age or 450 pg/mL in patients aged 75 years or older, based on the cutoff proposed in the CARDIOTOX registry[7]. Available echocardiographic records from the institutional platform were analyzed, all performed by specialized personnel using standardized equipment and validated protocols for LVEF measurement, following the modified biplane Simpson’s method. All echocardiographic studies were conducted with Philips Affinity 70 ultrasound systems, software version 10.0.

Bias control and ethical considerations

To minimize selection bias, all patients meeting the eligibility criteria during the study period were included. Standardized diagnostic criteria were applied to define cardiotoxicity, ensuring homogeneity in case classification. Data confidentiality was guaranteed, and the study was approved by the Institutional Research Ethics Committee. Given that the study is documentary (based on medical records) and retrospective, the committee determined that obtaining informed consent from each patient was not necessary, as researchers did not have direct contact with patients, information in the medical records remained unaltered, and all data anonymization processes were strictly followed.

Statistical analysis

A descriptive analysis of clinical and demographic variables was conducted, reporting central tendency and dispersion measures for continuous variables and frequencies for categorical variables. To assess changes in biomarkers and LVEF before and after chemotherapy, paired mean comparison tests were used (Student’s t-test or Wilcoxon test, depending on data distribution). A value of P < 0.05 was considered statistically significant. Statistical analysis was performed using SPSS v29 software (IBM Company, Chicago, United States).

RESULTS

A total of 204 subjects classified as intermediate, high, or very high risk for cardiotoxicity according to the Mayo Clinic Cardiotoxicity Risk Score were initially included. After excluding ten cases due to incomplete clinical records, the final sample comprised 194 patients. Of these, 55.7% (n = 108) were male with a median age of 55 years (interquartile range: 18–85 years). The most common clinical comorbidities were hypertension in 25.8% of patients (n = 50), heart failure in 13.9% (n = 27), and smoking in 14.9% (n = 29) (see Table 1). Regarding oncologic history, 27.8% (n = 54) had metastatic disease, and 24.2% (n = 47) were undergoing radiotherapy during the study period.

Table 1 Baseline characteristics of study patients, n (%).

Baseline characteristics	Total patients	Patients with cardiotoxicity	χ2 (P value)
	(n = 195)	(n = 17)
Sociodemographic characteristics
Male	108 (55.7)	11 (64.7)
Female	87 (44.6)	6 (35.3)	0.209
Smoking	29 (14.9)	3 (17.6)	0.442
Medical history
Hypertension	50 (25.8)	7 (41.2)	0.129
Dyslipidemia	17 (8.7)	2 (11.8)	0.384
Diabetes	23 (11.8)	4 (23.5)	0.113
Acute myorcardial infarction	7 (3.6)	0 (0.0)	NA
Atrial fibrillation	10 (5.1)	2 (11.8)	0.157
Heart failure	27 (13.9)	11 (64.7)	< 0.001
Peripheral vascular disease	24 (12.3)	2 (11,8)	0.421
Type of malignancy
Metastatic malignancy	54 (27.8)	4 (23.5)	0.278
Ongoing radiotherapy	47 (24.2)	4 (23.5)	0.398
Breast cancer	17 (8.7)	0 (0.0)	NA
Hodgkin lymphoma	7 (3.6)	0 (0.0)	NA
Non-Hodgkin lymphoma	78 (40.2)	7 (41.2)	0.416
Acute myeloid leukemia	11 (5.6)	4 (23.5)	0.005
Multiple myeloma	7 (3.6)	0 (0.0)	NA
Other non-hematologic malignances	55 (28.3)	3 (17.6)	0.127
Other hematologic malignances	19 (9.8)	3 (17.6)	0.192

NA: Not available.

Affected patients were predominantly male (64.7%, n = 11) with a mean age of 51.88 ± 15.9 years. Non-Hodgkin lymphoma was the most frequent malignancy (40.2%, n = 78), followed by non-hematologic cancers (28.3%, n = 55) and other hematologic malignancies (9.8%, n = 19). Regarding oncologic treatments, anthracycline-based monotherapy was the most common regimen (48.4%, n = 94), followed by combinations of anthracyclines with antimetabolites (21.1%, n = 41) and alkylating agents (18%, n = 35). Figure 1 illustrates the distribution of each treatment regimen among patients who developed cardiotoxicity and those who did not.

Figure 1 Distribution of different drug classes among patients with and without cardiotoxicity. The X-axis represents drug classes, while the Y-axis shows the percentage of patients exposed to each category. Blue bars correspond to patients without cardiotoxicity, whereas orange bars indicate those with cardiotoxicity. The highest prevalence was observed in the "Anthracyclines & Antimetabolites (Cytarabine)" group, with 51% of patients without cardiotoxicity and 53% with cardiotoxicity. Conversely, "Antimetabolites" have the lowest representation, with 8% in the non-cardiotoxicity group and 6% in the cardiotoxicity group.

A decrease in LVEF consistent with cardiotoxicity was identified in 8.8% of patients (n = 17) based on echocardiographic findings. Among these cases, 64.7% (n = 11) were male, with a mean age of 51.88 ± 15.9 years. Hypertension was present in 41.2% (n = 7), diabetes mellitus in 23.5% (n = 4), and pre-existing heart failure in 64.7% (n = 11). In this subgroup, non-Hodgkin lymphoma was the most prevalent malignancy (41.2%, n = 7), followed by acute myeloid leukemia (23.5%, n = 4), and both other hematologic and non-hematologic cancers (17.6%, n = 3 each).

The most frequently used pharmacological regimen in patients with cardiotoxicity was the combination of anthracyclines with antimetabolites (52.9%, n = 9), followed by anthracycline monotherapy (23.5%, n = 4) and alkylating agents (17.6%, n = 3).

Regarding cardiotoxicity severity, mild forms were observed in 58.8% of identified cases (n = 10), representing 5.1% of the total study population. Among these cases, 41.1% (n = 7) had been treated with anthracyclines and antimetabolites, 11.7% (n = 2) with anthracyclines alone, and 5.8% (n = 1) with alkylating agents. The most common comorbidities were heart failure (29.4%, n = 5), followed by hypertension and diabetes mellitus, each present in 11.8% of cases (n = 2). Smoking was identified in 5.9% of cases (n = 1).

Troponin levels were assessed in 20.1% of patients, with a mean value of 0.37 ng/mL (SD: 1.69; range: 0.002-10.54). Similarly, ProBNP was measured in 7.7% of the patients, with a mean value of 355.77 pg/mL (SD: 599.6; range: 0-1613), reflecting a wide variability in myocardial injury and ventricular overload markers within the study cohort. The use of cardiac biomarkers was limited in the studied population prior to chemotherapy initiation. Troponin was more frequently measured in those who developed cardiotoxicity (58.8%), with an average value of 0.39 ± 0.95 ng/mL. NT-proBNP was assessed in 17.6% of these cases, with mean values of 233.73 ± 47.86 pg/mL. The combined use of both biomarkers was scarce, observed in 2.5% of the overall sample and 17.6% of patients with cardiotoxicity.

At the end of treatment, troponin was evaluated in 15.3% of patients, NT-proBNP in 3.1%, and both biomarkers in 6.6%. In the cardiotoxicity subgroup, NT-proBNP was used in 29.4% of cases, troponin in 23.5%, and the combination of both in another 23.5% (Figure 2). However, absolute values of these biomarkers were not recorded. Regarding echocardiographic assessment, 70.1% of patients (n = 136) had an initial echocardiographic evaluation with GLS measurement performed in 47.1% of these cases. Initial GLS was documented in 41.1% of patients in the cardiotoxicity subgroup, with an average value of -18.38 ± 4.62%. The mean initial LVEF was 62% (IQR: 58%-66%), while the mean was lower in patients with cardiotoxicity (56.44 ± 10.45%). At the final follow-up, only 37.9% of patients underwent echocardiography, but this percentage was significantly higher for those with cardiotoxicity (76.5%). GLS measurement in this phase was performed in 15.8% of the general population and in 76.5% of patients with cardiotoxicity, with a mean value of -14.22 ± 4.93. The final LVEF showed a significant reduction in patients with cardiotoxicity, with a mean of 46% (IQR: 34%-56%), while it remained at 60% (IQR: 56%-67%) in the general population.

Figure 2 Use of biomarkers and echocardiography before and after chemotherapy. In the cardiotoxicity subgroup, NT-proBNP was used in 29.4% of cases, troponin in 23.5%, and the combination of both in 23.5%. A: The left panel illustrates the use of cardiac biomarkers before and after chemotherapy. Troponin is the most frequently used biomarker, with a slightly higher percentage before chemotherapy (blue) compared to after (red). The use of proBNP remains relatively low and stable between both periods. The combined use of both biomarkers increases after chemotherapy; B: The right panel depicts echocardiography utilization and left ventricular ejection fraction (LVEF) evolution in chemotherapy patients. The usage percentage of echocardiography and global longitudinal strain recordings (blue bars) decreases from the start to the end of treatment. The red solid and dashed lines show a decline in global LVEF and a more pronounced reduction in LVEF among patients with cardiotoxicity.

DISCUSSION

Cardiotoxicity is one of the main complications associated with antineoplastic agent treatment, encompassing a range of cardiovascular adverse effects, including asymptomatic ventricular dysfunction, heart failure, thromboembolic events, and arrhythmias[11,12]. Although myocardial dysfunction is the most frequent manifestation, other abnormalities such as hypertension, QTc interval prolongation, and pulmonary thromboembolism have been reported. However, their assessment was not within the scope of this study[13].

Despite the availability of multiple risk stratification models for cardiotoxicity, the lack of prospective validation limits their clinical applicability[14]. In this study, patients were classified as intermediate, high, or very high risk according to the Mayo Clinic Cardiotoxicity Risk Score, identifying an overall incidence of cardiotoxicity of 8.8%, distributed as 5.2% mild cases and 3.6% moderate cases. These findings align with those reported by Cardinale et al[15], who described a prevalence of 9% in patients with similar risk factors. However, studies such as the CARDIOTOX registry have documented a significantly higher prevalence (37.5%), which could be attributed to differences in the studied population, diagnostic methodology, and monitoring intensity[16].

Although most oncology patients do not develop overt cardiotoxicity, the clinical impact of this complication remains significant, representing a relevant cause of morbidity and mortality in this group[4]. Therefore, early risk identification is essential to optimize follow-up and prevent adverse events through advanced diagnostic tools and the timely use of cardioprotective medications, as well as reinforcing non-pharmacological measures such as exercise prescription[17]. In our cohort, echocardiography was used in 70.1% of patients, while cardiac biomarkers were assessed in only 24.7%, suggesting underutilization of these tools and a potential underestimation of cardiotoxicity in this population. Additionally, in patients without biomarker evaluation but with echocardiography, risk stratification was based on LVEF. According to the literature, patients with a borderline low LVEF (50%-54%) before treatment initiation have an increased risk of cardiac dysfunction, particularly when treated with anthracyclines or trastuzumab[18]. This risk is further accentuated in the presence of additional cardiovascular risk factors such as hypertension, dyslipidemia, and a history of cardiovascular disease, which were considered in this study[19,20].

Our results are consistent with the literature and reinforce the close relationship between anthracycline use and cardiotoxicity. The incidence of cardiotoxicity in patients treated exclusively with anthracyclines was 2.96% (4 out of 135 patients), whereas those receiving anthracycline and antimetabolite combinations had a significantly higher incidence of 21.95% (9 out of 41 patients), suggesting a possible synergistic effect on cardiac toxicity. Previous studies have documented that anthracyclines account for 66.7% of chemotherapy-induced cardiotoxicity cases, with a 5- to 15-fold increased risk of heart failure[21] and an odds ratio (OR) of 5.43 (95%CI: 2.34-12.62)[22]. The incidence of cardiotoxicity in patients treated with alkylating agents was 8.57% (3 out of 35 patients), consistent with previous studies that reported a higher risk associated with these agents, particularly in the context of hematopoietic stem cell transplantation[23]. In contrast, one patient treated exclusively with antimetabolites developed cardiotoxicity (100%, 1 out of 1 patient), although the small sample size of this subgroup prevents conclusive inferences.

Systematic monitoring with cardiac biomarkers has been recommended by the latest clinical guidelines, suggesting baseline measurement to establish a reference point and serial follow-up based on the therapeutic regimen. This allows for early detection of myocardial dysfunction and improved risk stratification[24,25]. However, the interpretation of these biomarkers remains challenging due to the influence of multiple factors, highlighting the need to standardize their use in clinical practice and establish baseline measurements to analyze trends with serial assessments during treatment[26]. A systematic review by Lars Michel et al[27] reported that up to 21.4% of patients with cardiotoxicity had elevated troponin levels, particularly those treated with anthracyclines or HER2-targeted therapies. Moreover, these patients had an 11.9-fold higher risk of developing left ventricular dysfunction (OR; 95%CI: 4.4-32.1), with an exponential 97.9-fold increase in risk when high anthracyclines doses were used (OR; 95%CI: 52.2-183.3).

Troponin was elevated in 2% of patients in this study, underscoring its relevance in detecting and monitoring cardiotoxicity, given its ability to identify early myocardial injury and its impact on associated morbidity and mortality[28]. In this regard, the CARDIOTOX registry demonstrated that severe cardiotoxicity, characterized by a significant increase in troponin and left ventricular dysfunction, was closely associated with higher mortality rates[7]. However, its diagnostic utility varies depending on clinical context, as the sensitivity and specificity of troponin I in predicting chemotherapy-induced cardiotoxicity have been relatively low, with values of 48% (95%CI: 0.27-0.69) and 73% (95%CI: 0.59-0.84), respectively[29]. This emphasizes the importance of a multimodal approach combining biomarkers with advanced imaging tools, such as echocardiography with GLS and cardiac magnetic resonance imaging, to enhance accuracy in early myocardial dysfunction detection[30].

Anthracycline-induced cardiotoxicity exhibits a dose-dependent relationship, with a troponin (cTn) release pattern that is useful for identifying patients at risk of future cardiovascular events[25,31,32]. Current evidence supports the superiority of serial measurements over isolated determinations, as they more accurately reflect myocardial injury progression[30]. However, the interpretation of these biomarkers should be performed within the individual clinical context to guide cardioprotective strategies and specific therapeutic decisions[24,32].

The European Society of Cardiology (ESC) guidelines highlight the negative predictive value of cTn and NT-proBNP in evaluating future cardiovascular events[33]. NT-proBNP was elevated in 1.5% of patients in our study. Given its pathophysiological role in hypertrophy, remodeling, and pressure and volume overload detection in heart failure, it emerges as a risk predictor and indicator of adverse cardiovascular outcomes in oncology patients with elevated baseline values[33]. Although the meta-analysis by Michel et al[27] did not identify any significant association between increased NT-proBNP levels and left ventricular dysfunction (OR: 1.7; 95%CI: 0.7-4.2), and the CARDIOTOX registry did not find any association between baseline troponin and NT-proBNP levels and the development of severe or symptomatic cardiotoxicity-related cardiac dysfunction, maintaining NT-proBNP within normal ranges throughout oncologic treatment may indicate a lower probability of cardiotoxicity. Consistent with these findings, our study did not identify a strong correlation between NT-proBNP and left ventricular dysfunction, although measurements in our cohort were not systematic and only a very low percentage of patients had baseline and serial assessments.

The timely detection of cancer therapy-related cardiac dysfunction (CTRCD) is crucial for implementing therapeutic strategies that prevent long-term complications. The literature suggests that early cardioprotective interventions lose effectiveness when myocardial dysfunction detection is delayed beyond 6 months after chemotherapy, compromising complete LVEF recovery[25,26]. However, only 25.3% of patients in our cohort underwent follow-up with at least one biomarker during chemotherapy, highlighting the need to standardize monitoring protocols based on a multimodal approach that integrates biomarkers and imaging tools according to the chemotherapy regimen, patient risk profile, and the imperative need to develop cardio-oncology services for timely follow-up[25].

Echocardiography is a fundamental tool for detecting and monitoring cardiotoxicity. In our cohort, 70.1% of patients underwent baseline echocardiographic evaluation. While some authors question the necessity of an initial echocardiogram in low-risk patients[34], current guidelines recommend its use to facilitate longitudinal comparison and early identification of subclinical alterations[25]. LVEF remains the most commonly used parameter for assessing left ventricular systolic function; however, its ability to detect early-stage cardiotoxicity is limited, as it can remain within normal ranges despite the presence of incipient myocardial damage[6,35]. Furthermore, interobserver variability and its load-dependent nature compromise its reliability in detecting subtle functional changes[36,37].

In this context, GLS via speckle-tracking echocardiography has emerged as a superior method for early detection of left ventricular dysfunction. This parameter evaluates regional myocardial deformation and allows for the identification of cardiotoxicity before a LVEF is reduced[6,38,39]. A meta-analysis of 21 studies involving 1782 patients showed that GLS deterioration during chemotherapy was associated with a 16-fold increased risk of developing CTRCD (OR: 15.82; 95%CI: 5.84-42.85)[40]. In our cohort, patients who developed cardiotoxicity had an initial GLS of -18.38 ± 4.62%, reflecting early functional impairment despite a normal LVEF (56.44%). By the end of treatment, GLS significantly decreased to -14.22 ± 4.93%, demonstrating the progression of myocardial dysfunction, while LVEF decreased to 46%, confirming manifest systolic dysfunction. These findings align with previous studies that demonstrated that a relative GLS reduction > 15% from baseline is a strong predictor of subclinical left ventricular dysfunction, preceding LVEF deterioration[35,36]. Changes in GLS have been documented as early as 3 months after chemotherapy initiation, underscoring its value in implementing timely cardioprotective strategies[41].

Our results reveal a statistically significant association between chemotherapy-induced cardiotoxicity and an increased risk of heart failure (HF), a finding potentially related to the presence of immune checkpoint inhibitors[42] or other yet-to-be-established specific mechanisms in cardiotoxicity patients. The latter are linked to distinct mechanisms such as myocarditis and arrhythmias that exacerbate acute HF risk[42]. In any case, the higher prevalence of HF in cancer survivors may reflect a multifactorial scenario in which cumulative exposure to cardiotoxic therapies interacts with conventional risk factors such as preexisting metabolic diseases at cancer diagnosis. However, the absolute risk may not justify universal screening[43]. Data from the present study could support the development of stratified models integrating age, type/dose of cardiotoxic therapy, and comorbidities to optimize surveillance in vulnerable subpopulations. This approach would enable early detection and personalized management, mitigating long-term cardiovascular impact[43].

Finally, our findings reveal a significant disparity in cardiotoxicity between leukemia subtypes and other cancer types. This divergence may be explained by pathophysiological differences caused by intensive protocols with asparaginase/steroids in acute lymphoblastic leukemia (ALL) that are not used in acute myeloid leukemia (AML), favoring endothelial dysfunction[44]. These differences may relate to the predominance of mitochondrial alterations in AML due to anthracyclines[45]. Given these findings, we propose differentiated algorithms for evaluating patients with these two hematologic malignancies: In ALL, serial monitoring with echocardiographic strain (which detects subclinical deterioration with 92% sensitivity) and in AML, biomarkers such as NT-proBNP plus cardiac MRI in cases of LVEF decline. Future studies should explore microRNA profiles (e.g., miR-34a in ALL) as dynamic predictors, integrating in vitro models with iPSC-derived cardiomyocytes to personalize cardio-oncology surveillance strategies.

Given these findings, institutions could implement routine GLS assessments in high-risk patients as a standard practice, ensuring the early identification of myocardial dysfunction before LVEF deterioration. Likewise, integrating troponin measurements into standard chemotherapy protocols may improve risk stratification and facilitate timely cardioprotective interventions. Establishing institutional policies for systematic cardiac monitoring could help reduce the burden of chemotherapy-induced cardiotoxicity and improve long-term cardiovascular outcomes in cancer survivors.

CONCLUSION

The increase in cancer survival has exposed the current limitations in detecting cardiotoxicity, particularly CTRCD. Although LVEF remains the standard parameter, its low sensitivity in subclinical stages necessitates the integration of dynamic biomarkers (cTn) and techniques such as GLS. However, the underutilization of these tools in resource-limited settings exacerbates underdiagnosis, delaying critical interventions. While useful in established heart failure, NT-proBNP did not demonstrate early predictive value, reinforcing the need for multimodal biomarker panels.

Although cardiac troponin (cTn) has proven useful for early detection of myocardial injury, NT-proBNP has not shown a strong association with the development of left ventricular dysfunction, suggesting the need to explore biomarker combinations or novel diagnostic strategies. Regarding imaging assessment, LVEF remains the standard tool, but its low sensitivity in early stages highlights the importance of incorporating techniques such as GLS into follow-up protocols. The low utilization rate of these tools in our cohort may have hindered the early identification of myocardial dysfunction and delayed the implementation of cardioprotective measures.

Because CTRCD can impact the continuity of oncologic treatment, it is crucial to establish standardized and cost-effective monitoring strategies, particularly in resource-limited countries such as Colombia. Early detection and intervention are essential to improve cardiovascular outcomes in these patients. This study underscores the need for prospective research to validate comprehensive predictive models that integrate clinical factors, biomarkers, and advanced imaging techniques to optimize risk stratification and therapeutic decision-making. Finally, cardiotoxicity was associated with an increased risk of heart failure, possibly mediated by immune checkpoint inhibitor-induced myocarditis. Therefore, this adverse effect necessitates moving away from universal approaches and adopting personalized strategies that integrate clinical variables such as age, cumulative anthracycline dose, and metabolic comorbidities, as well as current and emerging biomarkers, including subtype-specific microRNAs.

ACKNOWLEDGEMENTS

To the Scientific Directorate for its support in the development of the Center of Excellence in Heart Failure, and to the Directorate of Research and Innovation of the National University Hospital of Colombia for its support in developing the Evidence-Based Clinical Standard for the Surveillance of Cancer Therapy-Related Cardiovascular Toxicity.