Data integration of non-animal tests for the development of a test battery to predict the skin sensitizing potential and potency of chemicals

Recent changes in regulatory restrictions and social views against animal testing have accelerated development of reliable alternative tests for predicting skin sensitizing potential and potency of many chemicals. Lately, a test battery integrated with different in vitro tests has been suggested as a better approach than just one in vitro test for replacing animal tests. In this study, we created a dataset of 101 test chemicals with LLNA, human cell line activation test (h-CLAT), direct peptide reactivity assay (DPRA) and in silico prediction system. The results of these tests were converted into scores of 0-2 and the sum of individual scores provided the accuracy of 85% and 71% for the potential and potency prediction, compared with LLNA. Likewise, the straightforward tiered system of h-CLAT and DPRA provided the accuracy of 86% and 73%. Additionally, the tiered system showed a higher sensitivity (96%) compared with h-CLAT alone, indicating that sensitizers would be detected with higher reliability in the tiered system. Our data not only demonstrates that h-CLAT can be part of a test battery with other methods but also supports the practical utility of a tiered system when h-CLAT and DPRA are the first screening methods for skin sensitization.

► A large-scale dataset of 101 chemicals was created for h-CLAT, DPRA and DEREK. ► h-CLAT and DPRA complemented the limited sensitivity of either test. ► The tiered system of h-CLAT and DPRA displayed high accuracy of 86.1% compared with LLNA. ► h-CLAT can serve as part of the test battery with other methods for prediction of sensitizing potential.