359 © 2008, Japanese Society for Alternatives to Animal Experiments Comparison of human skin irritation and photo-irritation patch test data with cellular in vitro assays and animal in vivo data Dagmar Jirova 1 , Manfred Liebsch 2 , David Basketter 3 , Erin Spiller 4 , Kristina Kejlova 1 , Hana Bendova 1 , Marie Marriott 5 and Helena Kandarova 4 1 National Institute of Public Health, 2 ZEBET, Federal Institute for Risk Assessment, 3 St John's Institute of Dermatology, St Thomas' Hospital, 4 MatTek Corporation, 5 Safety and Environmental Assurance Center, Unilever Colworth Laboratory Corresponding author: Dagmar Jirova National Institute of Public Health, Prague, Czech Republic Phone: +(420)-26-708-2439, Fax: +(420)-267-08-2386, jirova@szu.cz Abstract Replacement of the rabbit Draize skin irritation test or the animal photo-irritation test is in course in Europe under the REACH chemical strategy and the Cosmetics Directive. Various in vitro protocols, including 3D skin models, have been assessed. One key difficulty in determining the validity of alternative in vitro methods is that the in vivo animal data is both scarce and often of limited utility for prediction of effects in man. Consequently, we have examined in human 4h patch tests a number of chemicals of EU borderline classification. In addition, in a specific group of cosmetic ingredients we assessed the potential of photo- irritation using results obtained in 3D skin models and in human photopatch tests. Several chemicals reported to be irritant in the rabbit were found to be without effect in humans. 3D skin model assays and human patch tests provided concordant results particularly in case of non-irritating and non-phototoxic substances. In our view, skin model tests seem to be a useful tool for the prediction of human skin irritation or phototoxicity hazard, particularly for consideration of initial concentration for confirmatory human patch tests to prove substance and product safety. Keywords: skin irritation, photo-irritation, human patch test, human epidermal model, rabbit Draize test Introduction The in vivo skin irritation/corrosion test in rabbits was introduced by Draize in the 1940s to predict hazardous effects of substances and formulations coming into contact with human skin (Draize et al., 1944). Several decades later, it was recognized that the design of the test and visual grading of the effects is highly subjective and that the test performed has low predictivity and relevance towards humans (Nixon et al., 1975, Basketter et al., 2004). Reproducibility of the assay has also been questioned (Weil and Scala, 1971). These studies concluded that results of the animal test should not be relied upon exclusively as 1) the test over predicts irritating effects of chemicals in relation to human skin; 2) although accepted by regulators, the test has never been scientifically validated. The replacement of the in vivo rabbit skin test is therefore not only matter of scientific effort but recently also a legal requirement of the European Cosmetics Directive (EC, 2003). During the past two decades, a number of reconstructed human epidermal (RHE) models became available, that can be used for studies of hazardous effects of topically applied substances. In the area of skin irritation, two assays using RHE models (EPISKIN TM , EpiDerm TM ) were recently evaluated in the ECVAM Skin Irritation Validation study (Spielmann et al., 2007). For EPISKIN assay, satisfactory levels of specificity and sensitivity were obtained when compared to existing rabbit data. The EpiDerm assay revealed high specificity but relatively low sensitivity, and further optimization of the protocol was recommended to reach concordance with the regulatory accepted in vivo test. However, knowing the overpredictive nature of the in vivo rabbit assay, it was further recommended to search and analyze relevant human skin irritation data and to look at the reasons for disconcordant classifications. In consequence, we decided to perform a limited number of human patch tests with some substances from the ECVAM skin irritation validation study and several other commercially available chemicals and AATEX 14, Special Issue, 359-365 Proc. 6th World Congress on Alternatives & Animal Use in the Life Sciences August 21-25, 2007, Tokyo, Japan