Abstract
There is a strong demand for the development of alternative methods to animal experiments in the safety evaluation of chemical substances, but it remains very difficult to develop such a method for repeated-dose toxicity and carcinogenicity tests. A method called read-across, which utilizes existing toxicity information, has been thus attracting attention. In this method, the toxicity of a target substance is expected from the toxicity information of “source substances” that have similar chemical and/or biological properties to the target substance. Therefore, the selection of appropriate source substances is very important, but the similarity judgment tends to be subjective, and the problem of reproducibility has been pointed out. Therefore, we are aiming to develop an objective read-across method to replace rat repeated-dose toxicity and rat carcinogenicity tests. In our studies, read-across is divided into several steps: For example, 1) definition of data set with substances having in vivo toxicity data, 2) selection of test substances, 3) determination of inter-substance distances using molecular descriptors, 4) selection of source substances based on the distance, 5) additional selection of source substances based on in vitro test results, and 6) toxicity evaluation of test substances by read-across. In this session, I will introduce our research strategy and recent findings and would like to discuss the (regulatory) science of read-across for the safety evaluation of chemical substances.
