Quantitative adverse outcome pathway modeling for cigarette smoke-inducible airway mucus hypersecretion. Part 2: Bayesian network modeling for probabilistic risk estimation.

Abstract

The development of in vitro tests that reproduce real-world situations is crucial for toxicity- and disease-risk assessment without animal testing. Because signs and symptoms of health concerns can be complex, it is helpful to create a simplified representation of such manifestations using a conceptual framework such as an adverse outcome pathway (AOP). Combining an AOP with computational models could be a potential tool for the extrapolation of in vitro results to real-world scenarios. Here, we applied Bayesian network-based probabilistic quantitative models for disease-related risk estimation using an in vitro dataset on the AOP of mucus hypersecretion-a known representative symptom of chronic airway disease-obtained by repeated exposure of human bronchial epithelial cells to whole cigarette smoke. We also used a computational aerosol dosimetry model to account for differences between in vitro exposure concentrations and human exposure scenarios. The results revealed dose- and exposure repetition-dependent increases in adverse outcome probability, suggesting that the model reflects the risk continuum of cigarette smoking. Furthermore, under certain assumptions, dosimetry modeling indicated that our in vitro exposure concentrations were similar to actual smoking scenarios. As an exercise, we also calculated in vitro odds ratios for chronic bronchitis that were comparable to the range of real-world odds ratios for chronic bronchitis due to cigarette smoking. Our combinatory risk-assessment approach could be a valuable tool for estimating the chronic inhalation effects of inhalable products and chemicals.