Retrospective review on in vitro phototoxicity data generated in 3D skin models to support the development of new OECD Test Guideline

• TR Number: 926
• Keywords: Phototoxicity, photo-irritation, validation, regulatory acceptance, EpiDerm (EPI-200), SkinEthic, EPISKIN, Keraskin, basic cosmetic raw materials, essential oils, pharmaceuticals, UV-filters, fragrances, OECD TG

Phototoxicity (also photo-irritation), is a light-induced skin reaction that occurs when photoreactive chemicals are activated to produce cytotoxic effects. Phototoxicity testing of chemicals is mostly conducted in vitro, following the methods described in the OECD test guideline (OECD TG 432), using the Balb/c 3T3 mouse fibroblast cell line. The test is highly sensitive, but low specificity has been reported in some studies (1). The test has also limitations with testing poorly soluble compounds. The use of 3D reconstructed human epidermis (RhE) tissue models has been proposed as the second tier in an integrated testing strategy to assess potential phototoxic activity, especially for topical exposures (e.g., chemicals, pesticides, cosmetics) (2). We collected published data on phototoxicity testing using various RhE models and protocols. The dataset contains more than 80 materials and over 800 entries utilizing different RhE models. The analysis conducted on the datasets revealed, that despite some differences between the protocols (e.g., exposure times, dosing, solvents), the RhE models have potential not only to distinguish between phototoxic and non phototoxic materials but also the potential to predict phototoxic potency as demonstrated in limited studies conducted in parallel to clinical studies in human volunteers. This database provides a valuable resource towards achieving regulatory acceptance of the method and source of information or the developers of similar methods. The database does not contain information on the systemic phototoxicity protocols developed with EpiDerm and EPISKIN models, nor information on the formulation testing. These additional protocols and approaches are however described in the literature and might be considered as supporting evidence of broad applicability domain of the reconstructed human 3D skin models.