EFSA Modelling tools for human health, animal health and ecological risk assessment

Abstract

derive BMDs on an external dose basis 2. Dose response data adjusted for internal dose from the results of forward dosimetry simulations such as plasma or target organ concentrations on an acute or chronic basis. The use of the MIXTOX module to derive risk metrics for multiple chemicals is shown using exposure and hazard metrics as well as the use of DEB modelling to model the impact of chemicals for species of ecological relevance at the individual and population level. Finally, generation of automated reports from TKPlate simulations and discussion on further developments and training to support NGRA conclude this presentation. Open-source tools have become increasingly popular in the pharmaceutical and chemical industries, offering cost-effective and collabora-tive approaches to drug development and chemical risk assessment. PK-Sim, an open-source physiologically-based pharmacokinetic (PBPK) modeling and simulation tool within the Open Systems Pharmacology Suite (OSPS, www.open-systems-pharmacology.org), has experienced rapid growth since its transition to an open-source platform. As a non-animal method (NAM), PK-Sim is increasingly being used in regulatory submissions to support next-generation risk assessment (NGRA) in human , animal, and ecological studies. By providing a quantitative link between exposure assessment, internal dose, and toxicological data, these models facilitate the derivation of safe chemical levels and reduce animal testing. PK-Sim offers several advantages to the pharmaceutical and chemical industry, including a robust qualification framework and increased physiological detail for accurate exposure estimates, particularly relevant for NGRA in vulnerable populations and safety-critical drug development applications. As an open-source tool, PK-Sim promotes cost-effectiveness, transparency, and open collaboration, with community driven quality control enabled through the advanced development platform GitHub. In line with the session proposal, PK-Sim and other open-source tools address the scientific, ethical, and legislative demands to build a mechanistic understanding of toxicity, reduction of animal testing (NAMs), and optimization of risk assessment capacity with NGRA. PK-Sim allows for the prediction of exposure levels, identification of potential toxicities, and assessment of dosing strategies in multiple animal species and populations across various routes of absorption. Integration with a high-throughput-ready framework in R enables the assessment of thousands of chemicals within a NAM/NGRA paradigm. Originally developed for pharmaceutical applications, PK-Sim provides significant benefits to the industry, including a detailed database for high-precision kinetics/exposure assessments for chemical, harmonized qualification and assessment approaches to streamline cross-sector legislation, and inclusion of sensitive populations. Continued development and refinement of open-source tools like PK-Sim will further augment their utility for toxicological risk assessment, drug development , and regulatory decision-making. This fosters the open and transparent advancement of systems toxicology and pharmacology as a crucial pillar of model-informed, NAM-based NGRA and drug development. inhalation and dermal routes of exposure. chemicals. Altogether, the "httk" R package provides free access to data and models, which are continually updated, that can be easily used, evaluated, and traced to enable robust TK assessments for NGRA. This abstract does not necessarily reflect U.S. EPA policy. New approach methodologies (NAMs) are increasingly considered in next generation risk assessment (NGRA) particularly for hazard identification and characterisation and refer to non-animal-based testing approaches including in vitro, in silico and in chemico methods. Amongst such NAMs, quantitative in vitro in vivo extrapolation (QIVIVE) models and physiologically-based (PBK) respectively allow conversion of invitro to in vivo concentrations and prediction of internal dose such as plasma or target organ concentrations. Ultimately, the use of these QIVIVE and PBK models aims to derive reference points on an internal dose basis in species of relevance without relying on in vivo toxicological studies. Over the last few years, the European food safety Authority (EFSA) together with academic and national authorities have been developing modelling tools for human health, animal health and ecological RA. These include generic kinetic, QIVIVE and PB-K models for humans, test species (rat, mice, dog, rabbit) and several farm animal species (cattle, pig, sheep, chicken) as well as dynamic energy budget (DEB) models for species of ecological relevance. These models have been published open-access, assessed and validated using relevant case studies following the principles laid out in the OECD guidance on the use of PBK models in RA. In addition, the models have been implemented in the TKPlate graphical interface to further support their implementation in food and feed safety. TKPlate contains a number of modules 1. Input to select the model itself, species, exposure inputs and chemical specific data, 2. Forward dosimetry to predict kinetic parameters or blood/organ concentrations, 3. Reverse dosimetry to recalculate exposure metrics from internal dose or biomonitoring data. 4.Bench-mark dose (BMD) modelling 5. Mixture risk characterisation (MIXTOX) using a component-based approach (CBA) and methods from EFSA's scientific committee guidance, 6. DEB modelling. 7. Automated report summarising results. Here, applications of TKplate for human health, animal health and ecological NGRA of chemicals are illustrated first for the forward and reverse dosimetry modules. Derivation of reference points are then highlighted within the BMD module using 1. Dose response data to