Percutaneous penetration of para-substituted phenols in vitro.

The percutaneous penetration of 11 para-substituted phenols has been measured across full-thickness hairless mouse skin in vitro. The phenols, which spanned more than a 1000-fold range in octanol/water partition coefficient (P), were applied (14C-radiolabeled) to the skin surface in a small volume of volatile organic solvent. Permeation kinetics were continuously monitored and were characterized by the maximum observed flux (Jmax). The linear correlation of log Jmax with log P was very poor. However, inclusion of molecular volume (MV) in a multiple regression analysis considerably improved the relationship between the measured transport parameter and the physicochemical descriptors. Furthermore, significant parabolic (log Jmax = -0.18 + 1.35.log P - 0.30.[log P]2) and bilinear (log Jmax = -0.17 + 1.08.log P - 1.95.[log(beta.10logP + 1)]) dependencies were obtained, suggesting a change in the rate-limiting transport step (for compounds of high log P) from diffusion across the stratum corneum (SC) to partitioning at the SC-viable epidermis interface. Addition of a term in MV (or molar refractivity) further improved the absolute correlations, but with marginal statistical significance. A wider range of molecular size is necessary to unequivocally define the role of permeant dimensions in percutaneous permeability for this group of compounds. The quadratic log Jmax correlation with log P was compared to the previously reported steady-state permeability coefficients (Kp) of a different set of phenol analogs through human epidermis. Despite the different methodologies, different compounds, and different skin membranes employed, the patterns of behavior in the two data sets were consistent, and suggest that the form of this correlation may be suitable description of phenol permeability under a range of experimental conditions.