Compilation and curation of an acute inhalation toxicity dataset
Animal tests have traditionally been used to identify acute inhalation toxicants and define safety standards that protect human health. NAMs that include in vitro and computational approaches have been proposed as complementary resources. These approaches have the potential to identify toxicants and elucidate toxicity mechanisms in a manner that is more relevant to human biology than animal models. However, developing and evaluating these NAMs requires robust, well-curated, and chemically diverse reference data.
Using data from open-access sources, NICEATM compiled a database of in vivo rat acute inhalation data for approximately 1,200 unique substances. In addition to LC50 values (exposure concentration of a toxic substance estimated to be lethal to half of the test animals), metadata collected for each entry included exposure type, exposure route, species, sex, and number of animals tested, when available. The chemical set was found to represent a diversity of physicochemical properties and functional use categories represented in the EPA Chemical and Products Database. EPA hazard categories (e.g., nontoxic, toxic, highly toxic) were assigned based on LC50 values. While all hazard categories are represented in the database, about two-thirds of the chemicals were classified as Category IV (least toxic/nontoxic). The final curated database contains 2,565 entries for 1,209 unique substances. Of these, 1,025 unique chemicals (2,109 entries) have chemical structures standardized for quantitative structure–activity relationship (QSAR) modeling.
A variability analysis was conducted on the 231 chemicals with two or more LC50 values by calculating mean absolute deviation, a measure of variability based on the absolute distance between each data value and the mean of the dataset. Principal component analysis was then used to evaluate any associations with inhalation LC50 variability, but no associations were found between LC50 values and either physicochemical properties or functional use categories. Conditional probabilities were also calculated for these chemicals to determine the probability that a chemical would be assigned a specific EPA hazard category given that it was previously assigned the same or another category on the basis of an LC50 from another study. This analysis showed Category IV categorization to be most consistent and Category III (mildly toxic) to be least consistent. The database can be downloaded from NICE, and a paper describing the database is in preparation.