Development and evaluation of a robust, common protocols for reconstructed human skin models: In vitro skin irritation/corrosion testing

Alemán. Due to the increased need for information on the safety of chemicals (driven mainly by new the chemicals policy of the European Commission (REACH)) a lot of attention is paid to development and validation of in vitro methods replacing or reducing the use of test animals. Reconstructed human skin models offer a possibility to satisfy regulatory requirements for testing of local toxicity and at the same they contribute to the 3R approach. For in vitro skin irritation and corrosion testing several test assays have been successfully developed in the past applying diverse reconstructed human epidermal (RHE) models. However, since several RHE models disappeared from the market after validation (and with them the validated assays), a common test protocols, applicable to all well developed and standardised RHE models had to be developed. For skin corrosion and acute phototoxicity, the usefulness of a common protocol has been already proven (Liebsch et al., 1997; Liebsch et al., 1999). Recently an attempt has been made to develop a \"common skin irritation protocol\" for EpiDerm and EPISKIN RHE models (Cotovio et al., 2005; Kandárová et al., 2005). This protocol is currently evaluated in an ECVAM Skin Irritation Validation Study. To show that the \"common protocol concept\" works ZEBET has in 2004 performed two additional studies employing the SkinEthic RHE model. First, in a co-operation with BASF AG (Germany) and SafePharm (UK) we applied the common skin corrosion protocol of OECD TG 431. Than in a co-operation with SkinEthic Laboratories (France) and SCHERING AG Germany) we applied the \"common skin irritation\" protocol to SkinEthic RHE. After minimal adaptations specific for SkinEthic RHE model almost identical results compared to EpiDerm and EPISKIN were obtained in both assays. In the above mentioned studies the reliability of the common protocol and similarity between three reconstructed human epidermal models have been evaluated and confirmed. Minor, probably model-related differences in prediction have been observed. However, taken into account large variability of in vivo responses, the above mentioned RHE models and protocols can provide reliable prediction of irritation and corrosion potential of chemicals, that are concordant between laboratories and over time.