Prediction of acute inhalation toxicity using cytotoxicity data from human lung epithelial cell lines

Abstract Recent research on in vitro systems has focused on mimicking the in vivo situation of cells within the respiratory system. However, few studies have predicted inhalation toxicity using conventional and simple submerged two‐dimensional (2D) cell culture models. We investigated the conventional submerged 2‐D cell culture model as a method for the prediction of acute inhalation toxicity. Median lethal concentration (LC50) (rat, inhalation, 4 h) and half maximal inhibitory concentration (IC50) (lung or bronchial cell, 24 h) data for 59 substances were obtained from the literature and by experiments. Cytotoxicity assays were performed on 44 substances with reported LC50, but without IC50, data to obtain the IC50 values. A weak correlation was observed between the IC50 and LC50 of all substances. Semi‐volatile organic compounds (SVOCs) and non‐VOCs (NVOCs) (16 substances) with a water solubility of ≥1 g/L were strongly correlated between 24‐h IC50 and 4‐h LC50, and this had an excellent predictive ability to distinguish between Categories 1–3 and 4 (Globally Harmonized System classification for acute inhalation toxicity). Our results suggest that the submerged 2‐D cell culture model may be used to predict in vivo acute inhalation toxicity for substances with a water solubility of ≥1 g/L in SVOCs and NVOCs.