Prediction of human percutaneous absorption of chemical substances from laboratory animal data

This study developed mathematical models that can be used to quantitatively predict the extent of percutaneous absorption across human skin for risk assessment purposes. A rigorous evaluation of the peer-reviewed studies that appeared in the public literature between the 1960 and 2002 resulted in the creation of a unique database of animal and human data generated under the same experimental conditions that compared in vitro or in vivo human percutaneous absorption to six species of animals including guinea pig, monkey, mouse, pig, rabbit and rat. The database was used to develop a series of basic regression equations that can be used as predictive models for human percutaneous absorption. The predicted value for human percutaneous absorption is defined by the same experimental conditions used in the animal study.;Three measures of in vitro and one measure of in vivo percutaneous absorption were modeled. With respect to % absorption in receptor fluid, the results showed very good correlation in rabbit, monkey, pig, guinea pig and mouse. Regression results for % absorption in receptor fluid plus skin data showed very good correlation in mouse, pig and guinea pig. Regression analysis of absorption rate data produced very good correlation in pig, rabbit, mouse and guinea pig. In general, the rat showed the lowest correlation to human data for all three measures. Therefore the three in vitro measures were evaluated by rat strain. The regression of rat data by strain improved the correlation of rat data to human data. With respect to in vivo data, sufficient data were available only for the monkey and rat. A significant regression was obtained only for monkey data.;The use of basic regression equations derived from comparative human/animal data as described herein is a novel and practical approach for the quantitative risk assessment of chemical substances with respect to the dermal route of exposure and has not been previously reported. This approach should result in a more accurate estimation of human percutaneous absorption than has been previously reported in the literature and quantitatively improve human risk assessments for dermal exposures.