Benchmark dose (BMD) modelling analysis of in vitro data of human respiratory epithelium (MucilAir) to establish a toxicological point of departure (POD)

Abstract

The utility of conventional in vivo animal studies to characterize toxicity associated with repeat inhalation exposures to irritants is limited by the complexity of extrapolation to humans, as well as by animal welfare concerns. An in vitro model of the human upper airway epithelium (MucilAir™) is a functional model of the human respiratory tract and accurately represents potential effects in vivo. This model was used in the evaluation of human toxicity for an irritant chemical. A number of endpoints were assessed including Trans-Epithelial Electrical Resistance, Lactate Dehydrogenase Release and Resazurin metabolism which serve as markers of the surface irritant effect. Ten exposure concentrations ranging from 2 and 200 mg/L were evaluated in human respiratory epithelial cells as six replicates from five donors. A Hill model was fitted to the data using the BMDS software developed by the US EPA. Our assessments of model fit strongly favoured the Hill model over alternative polynomial models. Different definitions of the benchmark response (BMR) with various values of the benchmark response factor were evaluated, including the relative change BMR of 10% in the mean response over the control group, the one standard deviation BMR and the point BMR. The estimated Hill models well captured the patterns of dose-response curves. The BMD estimates were comparable across donors and endpoints, and their ranges and the 95% lower limits were used as the POD to determine chronic inhalation risk assessment values. Assessment of this POD against tissue-level exposure values demonstrated the safety of the assessed uses of this compound. This in vitro model is a fit for purpose alternative method for quantitative risk assessment of inhalation exposure to irritant materials.