Applicability of in silico models for the prediction of the sensitization potential of industrial chemicals

Abstract

The Integrated Assessment Scheme (IAS) defines weight factors for each piece of toxicological information under REACH in an Integrated Testing Strategy. This IAS is illustrated on skin irritation for Classification and Labelling (C&L) for five substances and using mostly (Q)SAR models. The models are the BfR Rulebase, Derek for Windows and TOPKAT, read across and pH-rules. The weight factors derived in the IAS show that for peracetic-acid and pentabromodiphenylether the C&L decision could be easily made. For bisphenol A additional information on read across was used to finalise a decision on C&L, while for methylphenyl-diisocyanate additional expert judgement was needed. For allylheptanoate only the TOPKAT prediction was in the applicability domain, which was insufficient on its own. Therefore, other non guideline testing information was used and in vitro testing results. The above examples on skin irritation information show how the IAS can aid in the decision making process and how it adds to the ToxRTool and the ITS of Hoffmann et al. on the same endpoint and similar methods.