Applicability of physicochemical data, QSARs and read-across in Threshold of Toxicological Concern assessment
Meta data
The present document has been produced and adopted by the bodies identified above as authors. In accordance with Article 36 of Regulation (EC) No 178/2002, this task has been carried out exclusively by the authors in the context of a grant agreement between the European Food Safety Authority and the authors. The present document is published complying with the transparency principle to which the European Food Safety Authority is subject. It may not be considered as an output adopted by EFSA. EFSA reserves its rights, view and position as regards the issues addressed and the conclusions reached in the present document, without prejudice to the rights of the authors.
Abstract
This report investigated how the applicability of TTC schemes can be improved by incorporating physicochemical data (both experimental and predicted) and toxicity data generated by non-testing methods such as Quantitative Structure-Activity Relationships (QSARs) and read-across within structurally related chemical groups. The objectives of the study were: TTC-datasets analysis; investigation of the possible use of physicochemical data and predicted toxicity data generated by QSARs and read-across; development of a new electronic datasets which support the analysis performed. In order to undertake these analyses, the non-cancer toxicological endpoints dataset of Munro et al (1996) and the Carcinogen Potency Database (CPDB) dataset were compiled into two new electronic datasets (available on the EFSA website to download) and their chemical space was characterized by means of statistically based methods using structural and physicochemical property descriptors. The analysis performed allowed the identification of groups, named clusters, of similar structures which were characterized in terms of single descriptors. The TTC datasets were also compared with the US EPA DSSTox dataset, a large dataset taken as reference. The investigation of the possible use of physicochemical data and predicted toxicity data generated by QSARs was explored to assess the capability of QSAR methods to improve the TTC approach. The Cramer scheme was applied on the TTC datasets and critically evaluated. The results showed that the Cramer scheme well fits the regulatory needs being conservative. Other statistical methods were applied to classify the datasets into classes of hazard and the results were compared with the ones provided by Cramer classification. Overall, the results of the study confirmed that the Munro and CPDB databases are broadly representative of the world of chemicals. They confirmed the protectiveness of the Cramer scheme for both non-cancer and cancer endpoints and indicated the potential of chemoinformatics for exploring relationships between chemical structure and toxicity.