Integrated skin sensitisation assessment Summary of three defined approaches

Integrated testing strategies (ITS) for regulatory safety assessments serve as a replacement for stand-alone tests and typically combine information from different sources (in vitro, in silico) to arrive at a prediction for a specified endpoint. Skin sensitisation is one such endpoint that has been prioritized by the OECD and other public agencies. It is also one of the few adverse outcomes for which an OECD-approved adverse outcome pathway (AOP) exists and validated assays assessing the key events have been developed. These assays include:

1. the direct peptide reactivity assay (DPRA) which measures covalent binding of a sensitiser to skin proteins,
2. KeratinoSens™ assay for the measurement of keratinocyte activation, and
3. the human cell line activation test (h-CLAT) for measuring dendritic cell activation.

These assays have also been used in several defined approaches using in silico and in vitro methods for skin sensitisation potency prediction for humans.

SaferSkin™ implements three defined approaches for making predictions on the skin sensitisation potential of a compound. All of the methods use a range of data which includes the substance’s physico-chemical properties, in silico data, and in vitro data from validated and OECD-approved skin sensitisation assays.

Approaches currently implemented in SaferSkin™

• Bayesian network integrated testing strategy (by Jaworska et al [1])
• Multiple regression analysis (by Natsch et al [3])
• “2 out of 3” weight of evidence (by Bauch et al [2])

Data used in SaferSkin™

Experimental values

Molecular descriptors