A dual in silico framework for cardiac drug evaluation to assess electrophysiological and mechanical responses

Cardiac therapeutics act through well-characterized mechanisms, most commonly by altering electrophysiological properties or myocardial contractility. However the response to pharmacological interventions often differs between healthy and diseased myocardium. It is therefore essential to characterize how medications influence both electrophysiological properties and contractile performance under baseline conditions as well as in the setting of cardiac disease. Doing so ensures that differences in safety and efficacy are not overlooked. In the context of electrophysiology, drugs that target ion channels, receptors or autonomic regulation can alter action potential duration, conduction velocity and excitability (Schleifer & Srivathsan, 2015). Even in a healthy heart such changes may lead to arrhythmias if not carefully understood. Importantly, medications not primarily intended for cardiovascular disease, such as certain antibiotics, antidepressants or antipsychotics, can still exert proarrhythmic effects by prolonging repolarization or altering ion channel function (Schleifer & Srivathsan, 2015). This means that drug-induced arrhythmias are not confined to cardiac therapeutics but may also arise as unintended consequences of treatments for unrelated conditions (Schleifer & Srivathsan, 2015). Diseased hearts further amplify this risk, as remodelling of ion channels and gap junctions slows down conduction, alters repolarization and increases arrhythmia susceptibility (Coronel et al., 2013). The surprising scale of this concern is reflected in surveillance resources such as CredibleMeds, which categorizes drugs based on their potential to prolong the QT interval and induce Torsades de Pointes. Currently the database lists 318 drugs that should be avoided in patients with congenital long QT syndrome across three risk categories: (1) known, (2) possible and (3) conditional (CredibleMeds, 2025). Also within these categories 277 medications are recognized as a risk for inducing Torsades de Pointes, even outside of congenital disease (CredibleMeds, 2025). The length of these lists truly highlights how proarrhythmic risk extends across many commonly prescribed therapies, including those not specifically developed for heart disease. This highlights the importance of evaluating electrophysiological safety both in healthy individuals and in patients with underlying heart disease. In addition to electrophysiology, contractile performance can be affected by drug action. Many drugs influence calcium handling and myocardial contractility (Schleifer & Srivathsan, 2015). Establishing drug effects in a healthy heart provides a baseline for expected systolic and diastolic responses. In the context of disease states such as heart failure, where contractility is already impaired, even mild depression of contractile function could worsen dysfunction. Conversely, agents that enhance contractility may be therapeutically valuable, but they can also increase the risk of adverse electrophysiological events, such as arrhythmia susceptibility (Schleifer & Srivathsan, 2015). Careful evaluation in both healthy and diseased myocardium is therefore needed to balance therapeutic benefits with potential risks. Building on this, it is important to consider how structural and molecular remodelling of the heart, including fibrosis, hypertrophy and altered calcium handling, affects drug responses (Azevedo et al., 2016). Such remodelling results in effects that are rarely uniform across patient populations. Comprehensive characterization of drug effects on both electrophysiology and contractility in health and disease would allow researchers to distinguish on-target benefits from off-target risks. One approach to investigating these effects is through in silico modelling. This approach is used in the Comprehensive in vitro Proarrhythmia assay (CiPA) initiative (Colatsky et al., 2016). The CiPA initiative is an international effort led by regulatory, industry and academic scientists to create a new paradigm for assessing proarrhythmic risk, by replacing the traditional reliance on hERG block and QT prolongation as primary predictors of this risk (Colatsky et al., 2016). By integrating mechanistic insights with human ECG data, CiPA improves upon existing safety assessment approaches (Colatsky et al., 2016). Earlier guidelines were effective in reducing torsadogenic risk but often resulted in the rejection of therapeutically valuable drugs (Colatsky et al., 2016). The framework consists of three complementary pillars: (1) assessment of drug effects on human ventricular ion channels, (2) in silico integration of these ion channel data to simulate cardiac electrophysiology and (3) validation in human stem-cell–derived cardiomyocytes and human ECG studies (Colatsky et al., 2016). This multi-tiered structure forms the foundation of predictive cardiac safety assessment under physiological conditions. Mora and colleagues (Mora et al., 2025) extend this foundation through a coupled electromechanical model capable of simulating both electrophysiological and contractile behaviour in healthy and diseased myocardium. This advancement bridges a critical translational gap by allowing the assessment of drug efficacy and safety across disease states, which is a major step towards more comprehensive, clinically translatable cardiac modelling. To achieve this Mora and colleagues (Mora et al., 2025) used the ToR-ORd model, which reproduces key electrophysiological properties of human cardiomyocytes, including action potential morphology, calcium handling and conduction velocity. Additionally, it has been independently validated against experimental datasets (Tomek et al., 2019). Mora et al. (2025) coupled this electrophysiological model with a computational framework of cardiomyocyte contractile function developed by Land et al. (2017). The resulting integration, the ToR-Land model, represents an important advance in the field, as it allows simultaneous assessment of electrical and mechanical drug responses within a unified modelling environment, thus offering new opportunities to investigate therapeutic efficacy and safety under various physiological conditions (Fig. 1). Using an in silico simulation to investigate the electromechanical effects of omecamtiv mecarbil (OM) on cardiomyocytes, Mora and colleagues (Mora et al., 2025) capture the dose–response relationship in a disease-relevant context, providing a reliable and accessible framework for predicting both therapeutic and adverse effects. The model accurately reproduces electromechanical behaviour at the cellular level, mimicking baseline physiology before being scaled up to tissue simulations (Mora et al., 2025). The results show a consistent prediction of dose-dependent increases in contractile force across both single-cell (0-D) and tissue-level (3-D) simulations, reinforcing confidence that these effects arise from underlying mechanistic processes (Mora et al., 2025). The 3-D ventricular model demonstrates greater wedge shortening and longer time-to-peak contractions, highlighting the added value of spatially detailed simulations for evaluating the integrated effects of OM on cardiac tissue (Mora et al., 2025). Although replicating the full complexity of human cardiac physiology remains challenging, these findings show how multiscale models that combine biophysics, tissue geometry and dose–response data bring us closer to that goal. A demonstration of the model's utility is provided by the Heart Failure with reduced Ejection Fraction (HFrEF) simulations. In this experiment, in silico administration of OM improves systolic contraction but also reveals important trade-offs (Mora et al., 2025). The simulations show that contraction duration is prolonged, accompanied by reduced fractional shortening and incomplete relaxation at higher OM concentrations (Mora et al., 2025). These findings indicate that the model can detect aspects of impaired mechanical performance under physiological and pharmacological stress. Importantly, this approach enables non-invasive assessment of contractile parameters and provides insight into electromechanical responses that would otherwise be difficult to detect prior to clinical evaluation. Frequency-dependent simulations further demonstrate the sensitivity of the model. When pacing frequencies are increased to 2 Hz or higher, elevated OM concentrations enhance cardiomyocyte contractile force but also prolong the time to reach peak contraction (Mora et al., 2025). At higher doses (1 µM), the model predicts impaired diastolic relaxation, increased left ventricular end-diastolic diameter and an overall reduction in fractional shortening (Mora et al., 2025). These findings show that the model can predict contractile performance under different pacing and dosing conditions, demonstrating its ability to assess the complex interplay between systolic and diastolic function in HFrEF. This study demonstrates that this in silico model can reproduce both the electrophysiological and mechanical aspects of human cardiomyocyte physiology in response to pharmacological interventions. Such work addresses a critical gap in the field: the need for models that are clinically translatable, cost-effective and capable of safely predicting the impact of therapeutic interventions in humans. By providing a controlled and reproducible framework for evaluating drug actions, this study advances the development of alternatives to traditional experimental approaches. The importance of this advance is emphasized by recent changes in research priorities. In July 2025, the National Institutes of Health announced that it will no longer create funding opportunities that are exclusively reliant on animal models, instead encouraging studies that incorporate artificial intelligence and computational modelling. This policy reflects a broader transition in biomedical research, one that is likely to be adopted by other funding agencies in the near future. Accordingly, the coupled in silico model demonstrated by Mora et al. (2025) represents a significant step towards improving clinical translatability, reducing development costs and minimizing risks in early-phase clinical trials. The integration of electrophysiological and contractile modelling also provides a foundation for future expansion into additional dimensions of cardiac physiology. One particularly important next step is the incorporation of cardiomyocyte metabolism. The heart is one of the most energy-demanding organs in the body, and metabolic dysfunction is a hallmark of heart disease progression. Healthy and diseased cardiomyocytes differ markedly in substrate utilization and mitochondrial function, and drugs that alter calcium cycling, ion fluxes or contractility may impose additional stress on mitochondria in the diseased state. In recent modelling work, metabolism has been coupled to electromechanical simulations by explicitly linking ATP consumption to both cross-bridge cycling and ion transport (Forouzandehmehr et al., 2022). Using this approach, investigators have simulated how pharmacological agents that affect calcium handling or myofilament sensitivity alter cellular energetic demand and ATP turnover, revealing potential metabolic side effects not evident from electrophysiology or contractile mechanics alone (Forouzandehmehr et al., 2022). Incorporating similar metabolic coupling into the framework developed by Mora et al. (2025), would therefore provide an even greater understanding of drug actions. Another area for future development is the ability to investigate dose- and time-dependent effects of pharmacological interventions. Modelling different doses in both healthy and diseased states over short-term versus chronic administration would substantially enhance the predictive value of these tools. Such refinements would allow researchers to anticipate not only therapeutic efficacy but also the potential for adverse effects under different clinical scenarios. Taken together, the coupled in silico model developed by Mora et al. (2025) offers an innovative approach to predicting the risks and benefits of heart medications prior to costly clinical trials. By bridging gaps in safety assessment, clinical translation and resource allocation, such models have the potential to reshape preclinical cardiovascular research and accelerate the development of safer, more effective therapies. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article. No competing interests declared. Amelia Malicki: Conception or design of the work; drafting the work or revising it critically for important intellectual content; final approval of the version to be published; agreement to be accountable for all aspects of the work. Patrick Sanosa: Conception or design of the work; drafting the work or revising it critically for important intellectual content; final approval of the version to be published; agreement to be accountable for all aspects of the work. Ngoc Le: Conception or design of the work; drafting the work or revising it critically for important intellectual content; final approval of the version to be published; agreement to be accountable for all aspects of the work. None.