Development and Evaluation of a High Throughput Inhalation Model for Organic Chemicals

Currently it is difficult to prospectively estimate human toxicokinetics (particularly for novel chemicals) in a high-throughput manner. The R software package httk has been developed, in part, to address this deficiency, and the aim of this investigation was to develop a generalized inhalation model for httk. The structure of the inhalation model was developed from two previously published physiologically-based models from Jongeneelen et al. (2011) and Clewell et al. (2001) while calculated physicochemical data was obtained from EPA’s CompTox Chemicals Dashboard. In total, 142 exposure scenarios across 41 volatile organic chemicals were modeled and compared to published data. The slope of the regression line of best fit between log-transformed simulated and observed combined measured plasma and blood concentrations was 0.46 with an r(2)= 0.45 and a Root Mean Square Error (RMSE) of direct comparison between the log-transformed simulated and observed values of 1.11.. Approximately 5.1% (n = 108) of the data points analyzed were > 2 orders of magnitude different than expected. The volatile organic chemicals examined in this investigation represent small, generally lipophilic molecules. Ultimately this paper details a generalized inhalation component that integrates with the httk physiologically-based toxicokinetic model to provide high-throughput estimates of inhalation chemical exposures.