Applicability of the in vitro skin irritation methods (EpiSkin™, EpiDerm™ SIT) to organosilicon-based substances

It is now widely accepted that the reconstructed human epidermis models (OECD TG 439) can be used as a standalone replacement of the in vivo rabbit assay (OECD TG 404) to accurately predict skin irritancy. Many legislations have now introduced the legal requirement to use in vitro methods as the first step. The applicability of these methods to organosilicon-based substances was not evaluated during the validation of this guideline. Therefore, the aim of the current work was to assess the applicability of EpiSkin™ and EpiDerm™ SIT in vitro methods for organosilicons. Ten substances were evaluated, and results were compared with existing rabbit data. The data showed that both test methods failed to accurately predict the in vivo skin irritation potential, with predictive capacities below the minimum test guideline requirements. The two models delivered consistent results in only 60% of the cases. Several hypotheses were explored to explain this high rate of discordance without success. As EpiDerm™ SIT showed 100% sensitivity, a new stepwise testing strategy is proposed for organosilicons consisting of starting with EpiDerm™ SIT, following by EpiSkin™ in case of positive outcome. While keeping protective, this adapted strategy avoids unnecessary animal testing.