Evaluation of a non-animal toolbox informed by adverse outcome pathways for human inhalation safety.

INTRODUCTION: This work evaluated a non-animal toolbox to be used within a next-generation risk assessment (NGRA) framework to assess chemical-induced lung effects using human upper and lower respiratory tract models, namely MucilAir™-HF and EpiAlveolar™ systems, respectively. METHODS: A 12-day substance repeated exposure scheme was established to explore potential lung effects through analysis of bioactivity readouts from the tissue integrity and functionality, cytokine/chemokine secretion, and transcriptomics. RESULTS: Eleven benchmark chemicals were tested, including inhaled materials and drugs that may cause lung toxicity following systemic exposure, covering 14 human exposure scenarios classified as low- or high-risk based on historical safety decisions. For calculation of bioactivity exposure ratios (BERs), obtained chemical-induced bioactivity data were used to derive in vitro points of departures (PoDs) using a nonlinear state space model. PoDs were then combined with human exposure estimates, i.e., predicted lung deposition for benchmark inhaled materials using multiple path particle dosimetry (MPPD) exposure computational modeling or literature maximum plasma concentration (C(max)) for systemically available benchmark drugs. DISCUSSION: In general, PoDs occurred at higher concentrations than the corresponding human exposure values for the majority of the low-risk chemical-exposure scenarios. For all the high-risk chemical-exposure scenarios, there was a clear overlap between the PoDs and lung deposited mass and C(max) for the benchmark inhaled materials and therapeutic drugs, respectively. Our findings suggest that combining computational and in vitro new approach methodologies (NAMs) informed by adverse outcome pathways (AOPs) associated with pulmonary toxicity can provide relevant biological coverage for chemical lung safety assessment.