Echocardiography Core Laboratory Reproducibility of Cardiac Safety Assessments in Cardio-Oncology

Machine Learning and Bias in Medical Imaging: Opportunities and Challenges

Bias in health care has been well documented and results in disparate and worsened outcomes for at-risk groups. Medical imaging plays a critical role in facilitating patient diagnoses but involves multiple sources of bias including factors related to access to imaging modalities, acquisition of images, and assessment (ie, interpretation) of imaging data. Machine learning (ML) applied to diagnostic imaging has demonstrated the potential to improve the quality of imaging-based diagnosis and the precision of measuring imaging-based traits. Algorithms can leverage subtle information not visible to the human eye to detect underdiagnosed conditions or derive new disease phenotypes by linking imaging features with clinical outcomes, all while mitigating cognitive bias in interpretation. Importantly, however, the application of ML to diagnostic imaging has the potential to either reduce or propagate bias. Understanding the potential gain as well as the potential risks requires an understanding of how and what ML models learn. Common risks of propagating bias can arise from unbalanced training, suboptimal architecture design or selection, and uneven application of models. Notwithstanding these risks, ML may yet be applied to improve gain from imaging across all 3A's (access, acquisition, and assessment) for all patients. In this review, we present a framework for understanding the balance of opportunities and challenges for minimizing bias in medical imaging, how ML may improve current approaches to imaging, and what specific design considerations should be made as part of efforts to maximize the quality of health care for all.

Temporal trends in the prevalence and severity of aortic stenosis within a contemporary and diverse community-based cohort

BACKGROUND

Data on the epidemiology of aortic stenosis (AS) are primarily derived from single center experiences and administrative claims data that do not delineate by degree of disease severity.

METHODS

An observational cohort study of adults with echocardiographic AS was conducted January 1st, 2013-December 31st, 2019 at an integrated health system. The presence/grade of AS was based on physician interpretation of echocardiograms.

RESULTS

A total of 66,992 echocardiogram reports for 37,228 individuals were identified. The mean ± standard deviation (SD) age was 77.5 ± 10.5, 50.5% (N = 18,816) were women, and 67.2% (N = 25,016) were non-Hispanic whites. The age-standardized AS prevalence increased from 589 (95% Confidence Interval [CI] 580-598) to 754 (95% CI 744-764) cases per 100,000 during the study period. The age-standardized AS prevalences were similar in magnitude among non-Hispanic whites (820, 95% CI 806-834), non-Hispanic blacks (728, 95% CI 687-769), and Hispanics (789, 95% CI 759-819) and substantially lower for Asian/Pacific Islanders (511, 95% CI 489-533). Finally, the distribution of AS by degree of severity remained relatively unchanged over time.

CONCLUSIONS AND RELEVANCE

The population prevalence of AS has grown considerably over a short timeframe although the distribution of AS severity has remained stable.

Challenges and Burdens in the Coronary Artery Disease Care Pathway for Patients Undergoing Percutaneous Coronary Intervention: A Contemporary Narrative Review

Clinical and economic burdens exist within the coronary artery disease (CAD) care pathway despite advances in diagnosis and treatment and the increasing utilization of percutaneous coronary intervention (PCI). However, research presenting a comprehensive assessment of the challenges across this pathway is scarce. This contemporary review identifies relevant studies related to inefficiencies in the diagnosis, treatment, and management of CAD, including clinician, patient, and economic burdens. Studies demonstrating the benefits of integration and automation within the catheterization laboratory and across the CAD care pathway were also included. Most studies were published in the last 5-10 years and focused on North America and Europe. The review demonstrated multiple potentially avoidable inefficiencies, with a focus on access, appropriate use, conduct, and follow-up related to PCI. Inefficiencies included misdiagnosis, delays in emergency care, suboptimal testing, longer procedure times, risk of recurrent cardiac events, incomplete treatment, and challenges accessing and adhering to post-acute care. Across the CAD pathway, this review revealed that high clinician burnout, complex technologies, radiation, and contrast media exposure, amongst others, negatively impact workflow and patient care. Potential solutions include greater integration and interoperability between technologies and systems, improved standardization, and increased automation to reduce burdens in CAD and improve patient outcomes.

Global Longitudinal Strain Monitoring to Guide Cardioprotective Medications During Anthracycline Treatment

Purpose of the Review

Anthracycline chemotherapy carries a risk of myocardial dysfunction and heart failure even at relatively low doses, and the clinical prediction of cancer treatment-related cardiac dysfunction (CTRCD) is inexact. Careful imaging or biomarker surveillance during chemotherapy can identify CTRCD before the development of heart failure. Currently, this surveillance is performed using ejection fraction (EF). While this is a reliable and reproducible test with three-dimensional techniques, the most widely used imaging technique is two-dimensional echocardiography, for which EF measurements have broad confidence intervals.

Recent Findings

The use of global myocardial strain (GLS) provides a more reliable and reproducible means of assessing global cardiac function and shows meaningful changes before a significant change of EF. Observational studies have shown that although absolute measurements of GLS, both at baseline and during therapy, are predictive of CTRCD risk, the most reliable approach is to assess the change of GLS with therapy — a meaningful relative change of 10–15% being significant. A clinical trial comparing GLS to EF surveillance did not show a significant change of EF in the overall study group, but did show that patients managed with a the GLS-guided approach were less likely to develop a meaningful change of cardiac function to an abnormal level. In at-risk patients, there is good evidence for the protective value of neurohormonal antagonists and statins: the use of GLS enables these benefits to be directed to those most likely to benefit, while minimizing their use in the majority of people, who do not need them.

Summary

Although GLS requires an element of training and efforts to ensure uniformity, it has proven to be a feasible, robust, and reproducible technique, ready for wide adoption.

Harnessing Machine Intelligence in Automatic Echocardiogram Analysis: Current Status, Limitations, and Future Directions

Echocardiography (echo) is a critical tool in diagnosing various cardiovascular diseases. Despite its diagnostic and prognostic value, interpretation and analysis of echo images are still widely performed manually by echocardiographers. A plethora of algorithms has been proposed to analyze medical ultrasound data using signal processing and machine learning techniques. These algorithms provided opportunities for developing automated echo analysis and interpretation systems. The automated approach can significantly assist in decreasing the variability and burden associated with manual image measurements. In this paper, we review the state-of-the-art automatic methods for analyzing echocardiography data. Particularly, we comprehensively and systematically review existing methods of four major tasks: echo quality assessment, view classification, boundary segmentation, and disease diagnosis. Our review covers three echo imaging modes, which are B-mode, M-mode, and Doppler. We also discuss the challenges and limitations of current methods and outline the most pressing directions for future research. In summary, this review presents the current status of automatic echo analysis and discusses the challenges that need to be addressed to obtain robust systems suitable for efficient use in clinical settings or point-of-care testing.

Artificial Intelligence Aids Cardiac Image Quality Assessment for Improving Precision in Strain Measurements

OBJECTIVES

The aim of this study was to develop an artificial intelligence tool to assess echocardiographic image quality objectively.

BACKGROUND

Left ventricular global longitudinal strain (LVGLS) has recently been used to monitor cancer therapeutics-related cardiac dysfunction (CTRCD) but image quality limits its reliability.

METHODS

A DenseNet-121 convolutional neural network was developed for view identification from an athlete's echocardiographic dataset. To prove the concept that classification confidence (CC) can serve as a quality marker, values of longitudinal strain derived from feature tracking of cardiac magnetic resonance (CMR) imaging and strain analysis of echocardiography were compared. The CC was then applied to patients with breast cancer free from CTRCD to investigate the effects of image quality on the reliability of strain analysis.

RESULTS

CC of the apical 4-chamber view (A4C) was significantly correlated with the endocardial border delineation index. CC of A4C >900 significantly predicted a <15% relative difference in longitudinal strain between CMR feature tracking and automated echocardiographic analysis. Echocardiographic studies (n =752) of 102 patients with breast cancer without CTRCD were investigated. The strain analysis showed higher parallel forms, inter-rater, and test-retest reliabilities in patients with CC of A4C >900. During sequential comparisons of automated LVGLS in individual patients, those with CC of A4C >900 had a lower false positive detection rate of CTRCD.

CONCLUSIONS

CC of A4C was associated with the reliability of automated LVGLS and could also potentially be used as a filter to select comparable images from sequential echocardiographic studies in individual patients and reduce the false positive detection rate of CTRCD.

Improving reproducibility of left ventricular ejection fraction in pediatric oncology patients: less is more

Reproducible measurement of left ventricular (LV) systolic function by echocardiography is important to detect cancer therapy-related cardiac dysfunction (CTRCD). We hypothesized that limiting the number of imaging operators and use of a single vendor would improve reproducibility of these measures. A standard operating procedure (SOP) whereby LV measurements were standardized and a cardio-oncology imaging team (5 sonographers, 6 cardiologists) was established. All pediatric oncology patient echocardiograms were acquired on a single vendor platform. In total, 100 consecutive pre-SOP and 100 post-SOP studies were reviewed. LV end-diastolic dimension (LVEDD), posterior wall thickness (PW), shortening fraction (SF), and ejection fraction by Simpson's biplane (EF) were re-measured by 2 blinded readers, and compared to what was originally reported. Image quality was scored by number of LV segments imaged (grades 1-4). Inter-observer reproducibility pre/post-SOP was assessed with intra-class coefficient (α). Reducing the number of imaging operators improved image quality (Grade ≥ 3: 13% vs. 46%, p < 0.001). Reproducibility of PW and LVEDD marginally improved (PW: 0.78 to 0.82; LVEDD: 0.96 to 0.97), and SF improved significantly (α = 0.65 vs. 0.79, p < 0.001). Pre-SOP reproducibility of LV EF was poor (α = 0.65), but improved significantly post-SOP (α = 0.83, p < 0.001). Reproducibility of LV EF improved with higher image quality score. Limiting imaging operators and vendor platform for pediatric oncology echocardiograms improves image quality and reproducibility of LV EF. Establishing an SOP and a cardio-oncology echocardiography team may improve precision of measurements used to detect CTRCD.

Impact of a training program incorporating cardiac magnetic resonance imaging on the accuracy and reproducibility of two-dimensional echocardiographic measurements of left ventricular volumes and ejection fraction

Background

Left ventricular (LV) ejection fraction (LVEF) assessed by two-dimensional echocardiography (2DE) is the most widely used parameter for clinical decision-making, but reproducibility and accuracy problems remain. We evaluated the usefulness of a novel training program based on cardiac magnetic resonance (CMR) imaging to obtain more reliable values of 2DE-derived LVEF and LV volumes.

Methods

Fifty-four sonographers from five hospitals independently measured LV volumes and LVEF using the same 2DE images from 15 patients who underwent CMR and 2DE. After receiving a lecture from an expert on how to properly trace the LV endocardium, each sonographer voluntary performed the measurements using the same datasets, and was invited to perform the same analysis for additional patients. The effect of the training intervention was evaluated using the coefficient of variation (CV) and coverage probability (CP).

Results

Before the intervention, the LV volumes were significantly underestimated and the LVEF was significantly overestimated compared to the CMR results; however, these differences were reduced after the intervention. In particular, the CP (0.52 vs. 0.76, p < 0.001) for the LVEF showed significant improvement. However, the degree of improvement differed among institutions, and the CV actually became worse in two hospitals after the intervention. Level of experience and self-practice was associated with the reproducibility after the intervention.

Conclusions

A training program using CMR as a reference improved the accuracy of 2DE-determined LV measurements. Since the degree of improvements differed among hospitals, individualization of training programs and periodical objective evaluation may be required to reduce inter-institutional variability.

The Evolving Design of NIH-Funded Cardio-Oncology Studies to Address Cancer Treatment-Related Cardiovascular Toxicity

Cardiovascular (CV) toxicity from cancer therapy is a significant and growing concern. Conventional oncology clinical trial designs focused solely on cancer treatment efficacy have not provided sufficient information on both CV risk factors and outcomes. Similarly, traditional CV trials evaluating standard interventions typically exclude cancer patients, particularly those actively receiving cancer therapy. Neither trial type simultaneously evaluates the balance between CV toxicity and cancer outcomes; however, there is increasing collaboration among oncologists and cardiologists to design new cardio-oncology trials that address this important need. In this review, we detail 5 ongoing, oncology-based trials with integrated CV endpoints. Key design features include: 1) a careful assessment of CV risk factors and disease before, during, and after cancer therapy with standardized collection of clinical imaging, functional, and biomarker data; 2) an introduction of cardioprotective interventions at various timepoints in cancer therapy; 3) a balance of the risk of subclinical CV injury with the need for ongoing cancer treatment; and 4) an understanding of the time profile for development of clinically apparent CV toxicity. Additional critical priorities in cardio-oncology clinical research include harmonization of data collection and definitions for all physician- and patient-reported exposures and outcomes.

•

Prospective assessment of CV risk factors before, during, and after cancer treatment.

•

Longitudinal monitoring of CV function with standardized review of CV imaging and functional and biomarker endpoints for evidence of subclinical cardiotoxicity.

•

Consideration for the timing of the introduction of the cardioprotective strategy.

•

Need to balance the delivery of cancer treatment with the risk of CV injury.

•

Long-term follow-up beyond cancer treatment intervention to determine clinical cardiotoxicity outcomes.

•

Rigorous collection of cancer and CV endpoints to answer questions about the impact of CV events on the delivery of cancer treatment and the long-term patient outcomes.

Accuracy and reliability of novel semi-automated two-dimensional layer specific speckle tracking software for quantifying left ventricular volumes and function

Purpose

To determine whether the semi-automated two-dimensional echocardiography (2DE) layer strain software, compared to cardiac magnetic resonance (CMR), is reliable for left ventricular (LV) volume quantification.

Methods and results

We retrospectively selected 84 patients who underwent CMR and 2DE on the same day. Novel 2DE layer strain software automatically provides LV contour in 3 myocardial layers and performs layer specific speckle tracking analysis, which calculates LV volumes, ejection fraction (LVEF), and global longitudinal strain (GLS) in each layer. These values were compared with reference values from CMR disk-area summation and feature tracking methods. Coverage probability (CP) was determined using predefined cut-off values and absolute differences in LV volumes of 30 mL, those in LVEF of 10%, and those in GLS of 4%. The software did not work in 3 patients (feasibility: 96%). Different layers resulted in different degrees of under- or over-estimation of LV volumes. Epicardial tracking significantly overestimated the LV volumes and significantly underestimated LVEF and GLS. Mid-myocardial tracking had less underestimation of LV volumes and equivalent CP values of LVEF (0.77 vs. 0.75 using the disk-area summation method and 0.90 vs. 0.94 using the feature tracking method) and GLS (0.95 vs. 0.92) compared with endocardial tracking. The new software showed excellent reproducibility, especially endocardial and mid-myocardial tracking.

Conclusions

Mid-myocardial tracking with the novel 2DE strain software provided less bias of LV volumes with high CP values of LVEF and GLS, which suggests that mid-myocardial layer speckle tracking analysis approximates CMR derived LV functional parameters.

Prospective evaluation of the cardiac safety of HER2-targeted therapies in patients with HER2-positive breast cancer and compromised heart function: the SAFE-HEaRt study

Purpose

HER2-targeted therapies have substantially improved the outcome of patients with breast cancer, however, they can be associated with cardiac toxicity. Guidelines recommend holding HER2-targeted therapies until resolution of cardiac dysfunction. SAFE-HEaRt is the first trial that prospectively tests whether these therapies can be safely administered without interruptions in patients with cardiac dysfunction.

Methods

Patients with stage I–IV HER2-positive breast cancer candidates for trastuzumab, pertuzumab or ado-trastuzumab emtansine (TDM-1), with left ventricular ejection fraction (LVEF) 40–49% and no symptoms of heart failure (HF) were enrolled. All patients underwent cardiology visits, serial echocardiograms and received beta blockers and ACE inhibitors unless contraindicated. The primary endpoint was completion of the planned HER2-targeted therapies without developing either a cardiac event (CE) defined as HF, myocardial infarction, arrhythmia or cardiac death or significant asymptomatic worsening of LVEF. The study was considered successful if planned oncology therapy completion rate was at least 30%.

Results

Of 31 enrolled patients, 30 were evaluable. Fifteen patients were treated with trastuzumab, 14 with trastuzumab and pertuzumab, and 2 with TDM-1. Mean LVEF was 45% at baseline and 46% at the end of treatment. Twenty-seven patients (90%) completed the planned HER2-targeted therapies. Two patients experienced a CE and 1 had an asymptomatic worsening of LVEF to ≤ 35%.

Conclusion

This study provides safety data of HER2-targeted therapies in patients with breast cancer and reduced LVEF while receiving cardioprotective medications and close cardiac monitoring. Our results demonstrate the importance of collaboration between cardiology and oncology providers to allow for delivery of optimal oncologic care to this unique population.

Electronic supplementary material

The online version of this article (10.1007/s10549-019-05191-2) contains supplementary material, which is available to authorized users.

Optimal Number of Heartbeats Required for Representing Left Chamber Volumes and Function in Patients with Rate-Controlled Atrial Fibrillation

BACKGROUND

The optimal number of heartbeats required for representing left heart chamber function in patients with atrial fibrillation (AFib) has not been extensively studied.

METHODS

To determine the optimal number, we performed an automated quantification analysis of three-dimensional echocardiography (3DE) data sets in 93 patients with AFib for whom 10-20 consecutive one-beat full-volume 3DE data sets were acquired twice. We measured left ventricular end-diastolic volume (LVEDV), left ventricular ejection fraction (LVEF), and maximal left atrial volume (LAVmax) in each heartbeat; each parameter was averaged using a serial number of heartbeats randomly selected, and these values were compared with the averaged value obtained from the entire set of heartbeats. Coverage probability was determined using predefined cutoff values, the relative percentage differences in LVEDV and LAVmax of 5%, and the absolute percentage differences in LVEF of 5%. The optimal number of heartbeats was defined as the minimum number of heartbeats showing coverage probability ≥95%.

RESULTS

Out of 93 patients, 73 patients had acceptable left ventricular contour casts (feasibility, 78%), and 79 patients had acceptable left atrial contour casts (feasibility, 85%). Using the aforementioned criteria, the minimum optimal number of heartbeats was nine for LVEDV and six for LAVmax. The corresponding minimum optimal number of heartbeats for LVEF was eight. However, the results varied as a function of the size of the chamber, the left ventricular function, and whether the AFib ventricular rate was controlled.

CONCLUSIONS

In patients with AFib, the optimal number of heartbeats required to obtain representative chamber volumes and function was six to nine heartbeats randomly selected using 3DE automated quantification software.

Optimal threshold of three-dimensional echocardiographic fully automated software for quantification of left ventricular volumes and ejection fraction: Comparison with cardiac magnetic resonance disk-area summation method and feature tracking method

AIMS

Novel fully automated left chamber quantification software for three-dimensional echocardiography (3DE) has a potential for reliable measurement of left ventricular (LV) volumes and ejection fraction (LVEF). However, the optimal setting of global LV endocardial border threshold has not been settled.

METHODS AND RESULTS

We performed LV volumes and LVEF analysis using fully automated left chamber quantification software (Dynamic HeartModelA.I., Philips Medical Systems) in 65 patients who had undergone both 3DE and cardiac magnetic resonance (CMR) examinations on the same day. We recorded LV end-diastolic volume (LVEDV) and LV end-systolic volume (LVESV) according to the change in LV global border threshold settings from 0-point to 100-point with each increment of 10-point. These values were compared to the corresponding values of CMR with disk-area summation method and feature tracking (FT) method. Coverage probability (CP) was calculated as an index of accuracy and reliability. Fully automated software provided LV volumes and LVEF in 57 patients (Feasibility: 88%). LVEDV and LVESV increased steadily according to the increase in border threshold and reached minimal bias when border threshold setting was 80 against CMR disk-summation method and 90 against CMR FT method. Corresponding CP of LVEF was 0.74 and 0.84 against disk-area summation method and FT method.

CONCLUSIONS

With CMR values as a reference, LV endocardial border threshold value can be set around 80 to 90 with the same number of LV end-diastole and end-systole threshold to approximate LVEDV, LVESV and LVEF with clinically acceptable CP values of LVEF.

Echocardiographic Assessment for the Detection of Cardiotoxicity Due to Vascular Endothelial Growth Factor Inhibitor Therapy in Metastatic Renal Cell and Colorectal Cancers

BACKGROUND

Cardio-oncology is a recently established discipline that focuses on the management of patients with cancer who are at risk for developing cardiovascular complications as a result of their underlying oncologic treatment. In metastatic colorectal cancer (mCRC) and metastatic renal cell carcinoma (mRCC), vascular endothelial growth factor inhibitor (VEGF-i) therapy is commonly used to improve overall survival. Although these novel anticancer drugs may lead to the development of cardiotoxicity, whether early detection of cardiac dysfunction using serial echocardiography could potentially prevent the development of heart failure in this patient population requires further study. The aim of this study was to investigate the role of two-dimensional speckle-tracking echocardiography in the detection of cardiotoxicity due to VEGF-i therapy in patients with mCRC or mRCC.

METHODS

Patients with mRCC or mCRC were evaluated using serial echocardiography at baseline and 1, 3, and 6 months following VEGF-i treatment.

RESULTS

A total of 40 patients (34 men; mean age, 63 ± 9 years) receiving VEGF-i therapy were prospectively recruited at two academic centers: 26 (65%) were receiving sunitinib, eight (20%) pazopanib, and six (15%) bevacizumab. The following observations were made: (1) 8% of patients developed clinically asymptomatic cancer therapeutics-related cardiac dysfunction; (2) 30% of patients developed clinically significant decreases in global longitudinal strain, a marker for early subclinical cardiac dysfunction; (3) baseline abnormalities in global longitudinal strain may identify a subset of patients at higher risk for developing cancer therapeutics-related cardiac dysfunction; and (4) new or worsening hypertension was the most common adverse cardiovascular event, afflicting nearly one third of the study population.

CONCLUSIONS

Cardiac dysfunction defined by serial changes in myocardial strain assessed using two-dimensional speckle-tracking echocardiography occurs in patients undergoing treatment with VEGF-i for mCRC or mRCC, which may provide an opportunity for preventive interventions.

Cardiotoxicity of cancer chemotherapy in clinical practice

Several anticancer agents are associated with significant cardiotoxicity. The list of cardiotoxic cancer therapeutic agents includes anthracyclines, trastuzumab, alkylating agents, antimetabolites, which have been in use for decades; and recently introduced anticancer therapies such as tyrosine kinase inhibitors, angiogenesis inhibitors, checkpoint inhibitors and proteasome inhibitors. Cardiac imaging using echocardiography, nuclear imaging techniques, and magnetic resonance (MR) imaging can help in the early detection of chemotherapy-related cardiotoxicity. This can prevent the morbidity and mortality resulting from the cardiotoxicity of these agents. Further research is needed to improve our understanding of the underlying mechanism of their cardiotoxicity and to develop newer preventive and therapeutic strategies for chemotherapy related cardiotoxicity.

Cardiac safety evaluation in cancer clinical trials

Identification and quantification of the cardiac adverse effects of new cancer therapeutics is important when comparing treatment arms in clinical trials. Heart failure and left ventricular dysfunction are some of the most common adverse cardiac effects of a range of cancer treatments, including anthracyclines, trastuzumab and other targeted agents. Using the example of trastuzumab-induced cardiac dysfunction, we evaluated phase III clinical trials performed over the past decade to establish the methods used to identify heart failure and impairment of left ventricular function. Both these adverse events are difficult to accurately quantify. A clinical diagnosis of heart failure is subjective, and measurement of left ventricular ejection fraction has high interobserver variability depending on the method used to measure it. We found there was heterogeneity in methods used to diagnose both these adverse events. In addition, the use of quality assurance techniques to reduce measurement variability was low. We discuss and propose methods to standardise and reduce variability of cardiac event assessment in cancer clinical trials. This will allow true comparison of cardiac events between arms and trials with the aim of ensuring cardiac safety data are accurate.

Definition of hidden drug cardiotoxicity: paradigm change in cardiac safety testing and its clinical implications

Unexpected cardiac adverse effects are the leading causes of discontinuation of clinical trials and withdrawal of drugs from the market. Since the original observations in the mid-90s, it has been well established that cardiovascular risk factors and comorbidities (such as ageing, hyperlipidaemia, and diabetes) and their medications (e.g. nitrate tolerance, adenosine triphosphate-dependent potassium inhibitor antidiabetic drugs, statins, etc.) may interfere with cardiac ischaemic tolerance and endogenous cardioprotective signalling pathways. Indeed drugs may exert unwanted effects on the diseased and treated heart that is hidden in the healthy myocardium. Hidden cardiotoxic effects may be due to (i) drug-induced enhancement of deleterious signalling due to ischaemia/reperfusion injury and/or the presence of risk factors and/or (ii) inhibition of cardioprotective survival signalling pathways, both of which may lead to ischaemia-related cell death and/or pro-arrhythmic effects. This led to a novel concept of ‘hidden cardiotoxicity’, defined as cardiotoxity of a drug that manifests only in the diseased heart with e.g. ischaemia/reperfusion injury and/or in the presence of its major comorbidities. Little is known on the mechanism of hidden cardiotoxocity, moreover, hidden cardiotoxicity cannot be revealed by the routinely used non-clinical cardiac safety testing methods on healthy animals or tissues. Therefore, here, we emphasize the need for development of novel cardiac safety testing platform involving combined experimental models of cardiac diseases (especially myocardial ischaemia/reperfusion and ischaemic conditioning) in the presence and absence of major cardiovascular comorbidities and/or cotreatments.

Biomarker Discovery in Cardio-Oncology

PURPOSE OF REVIEW

In this article, we review current and emerging approaches to biomarker discovery to facilitate early diagnosis of cancer therapy-associated cardiovascular toxicity.

RECENT FINDINGS

Although small studies have demonstrated an association between established biomarkers of cardiac injury (troponins and brain natriuretic peptide) and acute or subacute cardiotoxicity, there is insufficient evidence to support their use in routine clinical care. Preclinical studies to define the molecular mechanisms of cardiotoxicity, as well as the use of unbiased "omics" techniques in small patient cohorts, have yielded promising candidate biomarkers that have the potential to enrich current risk stratification algorithms. New biomarkers of cardiotoxicity have the potential to improve patient outcomes in cardio-oncology. Further studies are needed to assess the clinical relevance of molecular mechanisms described in animal models. Similarly, findings from "omics" platforms require validation in large patient cohorts before they can be incorporated into everyday practice.