New approach methodologies (NAM) to characterize neurotoxic chemicals for use in risk assessment: case study of four neurotoxic pesticides

Neurotoxic pesticides: Cyfluthrin (CYF), βCYF, endosulfan (ENDO) and chlorpyrifos-oxon (CPFO) were examined using new approach methodologies (NAM) to assess ToxCast <i>in vitro</i> neurotoxic assay activities (50% concentration–AC₅₀s), chemical-target specificity (Z-Scores) and <i>in vitro</i> to <i>in vivo</i> extrapolation with ToxCast assay AC₅₀s as inputs to calculate human Administered Equivalent Dose (AED_Human mg/kg) through Integrated Chemical Environment (ICE) models. Fold differences (FD) between AED_Human NAM predictions and adjusted (animal to human interspecies extrapolation) <i>in vivo</i> regulatory neurotoxicity points of departure (POD) were calculated to assess modeled predictions against measured <i>in vivo</i> data. Interspecies (animal to human) <i>in vivo</i> adjusted PODs were 0.147 mg/kg/d (CYF/βCYF), 0.180 mg/kg/d (ENDO) and 0.106 mg/kg/d (CPFO). ToxCast neurotoxicity assay results supported <i>in vivo</i> observations: CYF, βCYF and CPFO showed depressed neuroactivity, while ENDO showed increased neuroactivity. βCYF is known to have twice the toxicity of CYF <i>in vivo</i> but was less toxic in the <i>in vitro</i> assays. Percent assays with Z-Scores greater than 3 for CYF, βCYF, ENDO and CPFO were 100%, 100%, 67% and 71% respectively. Most AED_Human were less than 10-FD indicating good predictivity of the <i>in vivo</i> adjusted PODs with ICE models. The NAMs were useful for characterizing the main neurotoxic activities and the ICE models were highly predictive of <i>in vivo</i> PODs to support mechanisms in risk assessment.