Comparison of the human skin grafted onto nude mouse model with in vivo and in vitro models in the prediction of percutaneous penetration of three lipophilic pesticides

The evaluation of the degree of percutaneous penetration of agrochemicals is a key part of risk assessment for operators. The availability of suitable and predictive experimental models is crucial, in particular in the case of lipophilic compounds which persist in the stratum corneum (SC). Regulatory models (rat in vivo, human and rat in vitro) and the innovative human skin grafted onto nude mice (HuSki) model were compared for their ability to predict the human skin absorption. Radiolabelled malathion, lindane and cypermethrin (4 mu g/cm(2)) were topically applied to each model. The % of applied dose absorbed and that present in skin and SC were evaluated at 24 h. Additionally, the absorption profile of cypermethrin was evaluated in the in vivo rat and HuSki models for up to 11 days. We found that the human in vitro and HuSki models closely predicted the human skin absorption at 24 h, while rat models overestimated the human skin absorption. Furthermore, our experiments with cypermethrin indicated that evaluation of % percutaneous absorption over extended periods of time was feasible with the HuSki model. In our studies the HuSki model overcame the limitations of the regulatory models and is promising to realistically refine the dermal absorption assessment of topically applied chemicals. (c) 2006 Elsevier Inc. All rights reserved.