The functional outcomes of cardiotoxicity are cardiomyopathy, arrhythmias and heart failure. The latter occurs when the function of the heart is compromised to the extent that it can no longer pump enough blood to oxygenate the organs of the body. Ultimately, this can lead to death. These functional outcomes cannot be measured in vitro. Instead, biomarkers of key events at the cellular and tissue level, which are associated with these outcomes, are measured [1]. The challenge is to understand how the resultant in vitro concentration-response curves can be interpreted to give meaningful predictions of adverse outcomes in humans, both on the individual and the population level.

Physiologically-based kinetic (PBK) models are deterministic computational models that use differential equations to capture and describe the biokinetics of chemical compounds. Their typical application is to predict the concentration of chemical compounds at their site of action at specified times following known exposures. PBK models range from comprehensive whole-body models to more detailed models of specific organs and tissues.

There is rising recognition of the value of PBK models within the new approach methodology (NAM) framework for predicting toxicity [2]. For example, they are used to translate the concentration-response curves that are measured in vitro to in vivo dose-response curves (iterative forward dosimetry). This is an essential step towards determining a Benchmark Dose (BMD) or a No Observable Adverse Effect Level (NOAEL) for regulatory purposes. They are also used to reconstruct in vivo exposures by combining biomonitoring data and in vitro concentration-response curves (reverse dosimetry).

In this presentation, we will examine how PBK models have been used in chemical risk assessment for cardiotoxicity. We will then proceed to discuss how these models can be refined given current knowledge of the adverse outcome pathways for cardiotoxicity, and the biomarkers that can be measured in advanced in vitro models.