The demand for alternatives to animal testing has increased, but there has been no significant progress in developing alternatives for repeated-dose toxicity tests despite their importance in chemical risk assessment. A comprehensive analysis of existing toxicity studies is the first step toward understanding toxicity and developing alternatives. However, such an analysis has yet to be performed for industrial chemicals. Therefore, we collected and organized publicly available repeated-dose subacute toxicity studies in male rats and constructed a database consisting of more than 2000 toxicity studies with about 500 toxicity-related findings. We then analyzed the no observed effect and lowest observed effect levels, toxicity-related findings, and organ categories in the database. The analyses revealed commonly and uncommonly observed toxicity-related findings and organ categories, as well as toxicity-related findings and organ categories with low and high median lowest observed effect levels. In addition, we extracted the toxicity studies registered in the Japanese and European chemical regulatory systems and conducted the same analysis for these datasets as the entire database. The results suggest that commonly observed toxicity-related findings were similar, but some toxicity-related findings differed in the frequency of observations between the two datasets.
Keywords: Chemical Substances Control Low; REACH; Toxicity test data; regulatory toxicology; statistical analysis.
More than 2000 repeated-dose subacute toxicity studies of industrial chemicals in male rats were collected and organized.The distributions of study NOELs and the medians of the LOELs of about 400 toxicity-related findings were examined.Commonly and uncommonly observed, and low-LOEL toxicity-related findings differed from those of safety failures in drug development.REACH-contained substances had relatively larger LOELs than CSCL substances.Some toxicity-related findings differed in the frequency of observations between CSCL and REACH datasets.