Comparison of in vitro and in vivo human skin responses to consumer products and ingredients with a range of irritancy potential.

Human skin equivalent cultures were investigated as possible pre-clinical skin irritation screens to aid safety assessments for chemicals and product formulations, and to facilitate design of safe and efficient human studies. In vitro responses in human skin equivalent cultures were compared directly to in vivo human skin responses from historic or concurrent skin tests for representative chemicals and products, including surfactants, cosmetics, antiperspirants, and deodorants. The in vivo data consisted of visual scores (i.e., erythema and edema) from skin-patch tests and diary accounts of skin irritation from product-use studies. In the in vitro studies, cornified, air-interfaced human skin cultures (EpiDerm) were evaluated using methods designed to parallel human clinical protocols with topical dosing of neat or diluted test substances to the stratum corneum surface of the skin cultures. The in vitro endpoints have previously been shown to be relevant to human skin irritation in vivo, including the MTT metabolism assay of cell viability, enzyme release (lactate dehydrogenase and aspartate aminotransferase), and inflammatory cytokine expression (Interleukin-1alpha). For surfactants, dose-response curves of MTT cell-viability data clearly distinguished strongly-irritating from milder surfactants and rank-ordered irritancy potential in a manner similar to repeat-application (3x), patch-test results. For the antiperspirant and deodorant products, all the in vitro endpoints correlated well with consumer-reported irritation (r, 0.75-0.94), with Interleukin-1alpha (IL-1alpha) release, showing the greatest capacity to distinguish irritancy over a broad range. IL-1alpha release also showed the best prediction of human skin scores from 14-day cumulative irritancy tests of cosmetic products. These results confirm the potential value of cornified human skin cultures as in vitro pre-clinical screens for prediction of human skin irritation responses. A preliminary report of these results has been published.