New approach methodologies in human regulatory toxicology – Not if, but how and when!

The predominantly animal-centric approach of chemical safety assessment has increasingly come under pressure. Society is questioning overall performance, sustainability, continued relevance for human health risk assessment and ethics of this system, demanding a change of paradigm. At the same time, the scientific toolbox used for risk assessment is continuously enriched by the development of "New Approach Methodologies" (NAMs). While this term does not define the age or the state of readiness of the innovation, it covers a wide range of methods, including quantitative structure-activity relationship (QSAR) predictions, high-throughput screening (HTS) bioassays, omics applications, cell cultures, organoids, microphysiological systems (MPS), machine learning models and artificial intelligence (AI). In addition to promising faster and more efficient toxicity testing, NAMs have the potential to fundamentally transform today's regulatory work by allowing more human-relevant decision-making in terms of both hazard and exposure assessment. Yet, several obstacles hamper a broader application of NAMs in current regulatory risk assessment. Constraints in addressing repeated-dose toxicity, with particular reference to the chronic toxicity, and hesitance from relevant stakeholders, are major challenges for the implementation of NAMs in a broader context. Moreover, issues regarding predictivity, reproducibility and quantification need to be addressed and regulatory and legislative frameworks need to be adapted to NAMs. The conceptual perspective presented here has its focus on hazard assessment and is grounded on the main findings and conclusions from a symposium and workshop held in Berlin in November 2021. It intends to provide further insights into how NAMs can be gradually integrated into chemical risk assessment aimed at protection of human health, until eventually the current paradigm is replaced by an animal-free "Next Generation Risk Assessment" (NGRA).