The intra- and inter-laboratory reproducibility and predictivity of the KeratinoSens assay to predict skin sensitizers in vitro: results of a ring-study in five laboratories.

Due to regulatory constraints and ethical considerations, research on alternatives to animal testing to predict the skin sensitization potential of novel chemicals has gained a high priority. Accordingly, different in vitro, in silico and in chemico approaches have been described in the scientific literature to achieve this goal. To replace regulatory approved animal tests, these alternatives need to be transferable to other labs, their within and between laboratory reproducibility must be assured, and their predictivity should be high. The KeratinoSens assay is a cell-based reporter gene assay to screen substances with a full dose-response assessment. It is based on a stable transgenic keratinocyte cell line. The induction of a luciferase gene under the control of the antioxidant response element (ARE) derived from the human AKR1C2 gene is determined. Here we report on the results of a ring-study with five laboratories performing the KeratinoSens assay on a set of 28 test substances. The assay was found to be easily transferable to all laboratories. Overall both the qualitative (sensitizer/non-sensitizer categorization) and the quantitative (concentration for significant gene induction) results were reproducible between laboratories. A detailed analysis of the transferability, the within- and between laboratory reproducibility and the predictivity is presented.