A multiphase microscopic diffusion model for stratum corneum permeability. II. Estimation of physicochemical parameters, and application to a large permeability database.

The full parameterization for the stratum corneum biphasic microtransport model presented previously in this Journal [95:620-648 (2006)] is developed through a combination of fundamental transport theory and calibration with existing data. Of the five microscopic transport properties, four (D(cor), K(cor/w), D(lip), K(lip/w)) are developed from sources independent of the existing steady-state permeability database. The fifth parameter, k(trans) (the mass transfer coefficient for transbilayer hopping), is derived from a fit of the model to the permeability data according to a modified free surface area function of the form log(10) k(trans) = A-B x (MW)(1/3). Examination of the experimental data in terms of the two dimensionless groups, R and sigma, arising from the analysis leads to the conclusion that SC permeation for most compounds is dominated by the transcellular pathway regardless of their lipophilicity, a striking departure from recent skin permeability models. Overall fit of the developed model(s) to the permeability data is somewhat better than for the Potts-Guy equation and variants thereof; however, marked improvement is seen in the estimation of lag times and the related potential for predicting skin hydration effects and transient skin permeation profiles. Simple approximations to the full numerical solution are presented that allow the developed model(s) to be implemented on a spreadsheet.

A multiphase microscopic diffusion model for stratum corneum permeability. I. Formulation, solution, and illustrative results for representative compounds.

A two-dimensional microscopic transport model of the stratum corneum (SC) incorporating corneocytes of varying hydration and permeability embedded in an anisotropic lipid matrix is presented. Results are expressed in terms of a dimensionless permeability (P(SC/w)(comp), which is a function of two dimensionless parameters, R and sigma. R is a ratio of transbilayer to lateral molecular flows within a lipid bilayer and sigma is the ratio of (lateral) permeability in the lipid phase, D(lip)K(lip/w), to that in the corneocyte phase, D(cor)K(cor/w.) The shape of the dimensionless permeability surface is also governed by the arrangement of the SC lipids, where Model 1 represents the extreme in which lipid-phase transport can occur with no transbilayer transport, whereas Model 2 entails maximum transbilayer transport. Model calculations are exemplified by characterizing the skin permeability of four representative permeants: water, ethanol, nicotinamide, and testosterone. A comparison with experimental steady state permeability and partition data supports that the transport properties of the SC lipids are highly anisotropic, with lateral diffusivities several orders of magnitude higher than the equivalent diffusivity calculated from transbilayer hopping. Nevertheless, the calculations suggest that corneocyte-phase transport plays a major role for all four permeants. These results confirm our previous calculations on water permeability and present a marked contrast to the commonly stated doctrine that the SC transport pathway is primarily intercellular.

A two-phase analysis of solute partitioning into the stratum corneum.

An analysis is presented of partition coefficients K(SC/w) describing solute distribution into fully hydrated stratum corneum (SC) from dilute aqueous solution (w). A comprehensive database is compiled from the experimental literature covering more than eight decades in the octanol/water partition coefficient K(o/w). It is analyzed according to a two-phase model following that of Anderson, Raykar, and coworkers (1988, 1989), which accounts for uptake by intercellular lipid and corneocyte (keratin plus water) phases having inherently different lipophilicities, as characterized by an SC lipid/water partition coefficient K(lip/w) and a partition coefficient PC(pro/w) quantifying cornoeocyte-phase binding. Regression of 72 data points yields useful best-fit recalibrations of power laws (or linear free energy relationships) giving K(lip/w) and PC(pro/w) as functions of K(o/w). The specific conclusions of the analysis are as follows: (i) The two-phase model offers substantial improvements over previously proposed analytical representations of K(SC/w), yielding an rms error in log(10)K(SC/w) of 0.30 limited by the scatter in the data. (ii) The best-fit description of the lipid phase is given by the power law K(lip/w) = 0.43 (K(o/w))(0.81), suggesting about half the absolute value of K(lip/w) relative to previous estimates. (iii) The best-fit description of corneocyte-phase binding differs negligibly from the correlation found by Anderson, Raykar, and coworkers for the more limited set of compounds studied by them. Explicit consideration of the two-phase nature of the SC also furnishes a rational basis for predicting the effects of varying hydration state upon K(SC/w).

Mobility of water in human stratum corneum.

At low water activities, stratum corneum (SC) water sorption resembles that in other keratinized tissues (i.e., wool and horn), whereas at high water activities, it resembles that in polymeric hydrogels. We propose that the concentration-dependent water diffusivity observed in these other systems applies to the corneocyte phase of the SC. An increase in SC hydration leads to increased water diffusivity in the corneocytes, in accordance with the predictions of both effective diffusion and free volume theories. Thus, theoretical results on effective diffusivity in a composite medium with random fiber obstacles and a free volume theory for water diffusivity in hydrogels (calibrated using data from wool and horn) have been applied to human SC water sorption data to estimate and establish theoretical limits on water diffusivity in corneocytes as a function of water activity. These results are used in conjunction with steady-state water permeability data to estimate the water permeability of both corneocyte and lipid phases of the SC under hydrated and partially hydrated conditions. The results of the analysis, when combined with previous spectroscopic analyses, strongly suggest that the lipids provide most of the SC water barrier in either case; thus, the diffusion pathway for water is primarily transcellular.