A pragmatic workflow for human relevance assessment of toxicological pathways and associated new approach methodologies.

Currently, safety assessments of chemical substances are predominantly based on animal data. Multiple considerations call for the use of alternative testing strategies that are based on new approach methodologies (NAMs). However, the human relevance of these testing strategies is usually uncertain. This necessitates a harmonized and accepted workflow for assessing their applicability for regulatory purposes. This report proposes such a workflow, applicable for assessing the human relevance of a toxicological pathway and the relevance of NAMs related to the different components of the pathway. The workflow starts with an established toxicological pathway, of which the adverse outcome is relevant for human health risk assessment and that has sufficient weight of evidence. Human relevance is assessed through three main questions, related to the different components (steps) of the pathway, the pathology of human syndromes that have a similar adverse outcome, and quantitative aspects. The latter comprise both interspecies differences and in vitro - in vivo differences. The combined evidence is scored as 'strong', 'moderate' or 'weak' support of human relevance, based on expert judgement. The workflow developed was tested in a case study, through application to an AOP describing craniofacial malformations after in utero exposure to triazoles. Based on evidence collected for two of the three main questions, the case study provided moderate to strong support for human relevance of both the various components of the AOP and its associated NAMs. Furthermore, it demonstrated that the workflow is a promising approach that allows for a more transparent scientific evaluation of human relevance of toxicological pathways and associated NAMs. Therefore, despite some areas for improvement, we consider the workflow an important step forward for application of AOPs and related NAMs in human health risk assessment.