FORECASTING PULMONARY INFLAMMATION FROM IN VITRO ASSAY RESULTS FOR NANOPARTICLES

DESCRIPTION (provided by applicant): The rapidly developing field of nanotechnology shows promise by allowing designers to specifically select unique combinations of material properties as needed increasing the effectiveness of applications in medicine, coatings, lubrication, semiconductors, composites, and many others. These materials with their unique combinations of properties on exposure to humans may result in unanticipated hazards, however, putting workers in nanotechnology-related industries at risk. Traditional animal testing is expensive and too slow to evaluate potential risks for the current pace of new nanomaterial development. Both technology developers and regulators need more rapid methods to evaluate new nanomaterial configurations for their risk potential. Much hope is placed in high-throughput in vitro screening assays, but the relevance of these results to the potential for human disease or even the observed toxic effects in animal exposures is unclear. Some research has proposed Quantitative Structure Activity Relationships (QSARs) to predict in vitro nanomaterial toxicity in a few specific assays, but the applicability of these models to a wider group of materials, alternative in vitro assays, or in vivo toxicity has not been explored. If the primary exposure pathway for workers in the near term is inhalation, which in vitro assays will provide the most reliable risk information for that scenario? Two recently available data sources will permit this study to investigate this question: the Environmental Protection Agency's (EPA) ToxCast data for nanomaterials and the Nanomaterial Pulmonary Toxicity Database (NTDB), a collection of published peer reviewed studies observing pulmonary inflammation in rodents upon exposures to nanomaterials. This study will pursue the following specific aims: (1.) identify combinations ofin vitro assay results that can reliably forecast the results of pulmonary inflammation results in rodents; (2.) evaluate whether existing proposed QSARs for nanomaterial toxicity apply to a wider array of in vitro toxicity assays; and (3.) evaluate whether existing proposed QSARs for nanomaterial toxicity apply to in vivo pulmonary inflammation results. This study will employ machine learning methods to cluster similar nanomaterials between the various in vitro and in vivo results, and to identify combinations of in vitro assays that rank order the toxicity of nanomaterials most similarly to pulmonary inflammation results in rodents considering also how changes in specific chemical and physical particle properties exacerbate or mitigate observed toxicity. This study addresses documented research needs in the National Occupational Research Agenda (NORA) cross- sector Nanotechnology program including specific goals in the Human Health and Informatics categories. Implementation complies with the Research to Practice (r2p) Initiative in its formulation, design, and implementation plan including industry an public outreach. The insight generated by this study will improve nanomaterial risk screening capabilities and focus attention and effort on those measurements and techniques proven to be most effective and reliable enabling better management and control of the risks faced by workers.