Regulatory agencies require safety and toxicity studies to be performed in rodent and non-rodent species for registration of chemicals, such as pharmaceuticals, industrial chemicals, food additives, cosmetic ingredients, and pesticides (Van der Jagt et al., 2004). Toxicokinetics (TK) is an important component of these studies which describes the absorption, distribution, metabolism, and excretion (ADME) of the toxic substance in the body, and relates its exposure to toxicological findings and relevance to clinical safety (Van der Merwe D and Buur, 2018). Traditional approaches used for the calculation of TK parameters and systemic exposure are one- or two-compartmental analysis, noncompartmental analysis using statistical moment theory or population toxicokinetic analysis. However, these methods do not consider physiological characteristics and thus, can only reflect concentration changes over time in specific tissues or organs, and lack the ability to analyze the distribution and metabolism of a chemical simultaneously. As a result, these methods fail to predict and extrapolate results across different species, exposure routes, and doses (Hu et al., 2022; Sarigiannis et al., 2019).

Physiologically based pharmacokinetic or toxicokinetic (PBPK/PBTK) models have gained increasing regulatory acceptance in recent years for applications such as interspecies scaling and chemical risk assessment. These models facilitate quantitative descriptions of the temporal change in the concentrations of chemicals and/or their metabolites in biological matrices (e.g., blood, tissue, urine, alveolar air) of the exposed organism. Toxicokinetic models often describe the organism as a set of compartments that are characterized physiologically or empirically (Gerlowski and Jain, 1983). TK studies generally involve testing higher doses of the drug than those used in pharmacokinetic/pharmacology studies, resulting in saturation of drug dissolution, absorption, metabolism and excretion processes, leading to nonlinear kinetics (Li et al., 2023). Therefore, the bioavailability, half-life, apparent volume of distribution, clearance, and other parameters calculated from low-dose pharmacokinetic studies may not be applicable to high-dose TK investigations.

A PBTK model is a “bottom-up” mathematical model that combines the anatomical and physiological parameters of different species with physicochemical properties of drug to predict in vivo ADME. As this model integrates many drug-specific data such as solubility, permeability, metabolism, transporter interaction, etc., with species physiology, it provides a good understanding of all active processes influencing the pharmacokinetic properties of a drug (Zhuang and Lu, 2016). There is a strict separation between the physiology of the organism on the one hand and the properties of the compounds on the other, that allows the exchange of either the organism physiology or the drug physicochemical properties to predict the disposition of the same compound in different organisms or, in turn, the disposition of different drugs in the same species (Lippert et al., 2012; Meyer et al., 2012). This enables reliable interspecies extrapolation, dose extrapolation, and in vivo-in vitro correlation, thereby enhancing the dose-response relationship in toxicity testing and risk assessment (Lipscomb et al., 2012). Thus, the main advantage of PBTK models over the classical compartmental approaches is the ability of PBTK models to extrapolate outside of the doses or species that was evaluated experimentally (Tsamandouras et al., 2015). Fig. 1 illustrates the structure, input parameters, and output data of a PBTK model.

The current review focuses on the applications of PBTK modeling in toxicokinetics and its alignment with regulatory guidelines for risk assessment and interspecies extrapolation.