Potential pitfalls in skin permeation experiments: influence of experimental factors and subsequent data evaluation

There is a growing demand for skin penetration and permeation data considering toxicological and potential drug delivery aspects for an increasing number of substances. Although there are official guidelines available, results from different skin diffusion studies are often inconsistent and sometimes even controversial. The aim of our study is to address and to investigate the influence of experimental parameters as well as mathematical problems for subsequent evaluation of the permeation raw data. To create a reliable database diffusion experiments across human stratum corneum were performed under highly standardized conditions. The experimental data were evaluated using linear and non-linear regression analysis to determine the influence on the permeability coefficient and the lag-time. Additionally, the influence of two critical experimental parameters, temperature and unstirred water layers, on the permeability was investigated in silico. Based on our results we suggest that the influence of temperature on the permeability coefficient is small compared to the effect of other experimental parameters. Thickness of unstirred water layers has a tremendous effect on the permeation and may lead to underestimation of the permeability by more than 90%. Non-linear regression analysis seems to be superior compared to linear algorithms hence is advisable for evaluation of the experimental data. Our findings may help to optimize the experimental set-up and to reduce total costs for future skin diffusion testing. With regards to EU's REACH-Initiative this will also help to create more and reliable data on safety issues of industrial materials.