Quantification of vehicle mixture effects on in vitro transdermal chemical flux using a random process diffusion model

Abstract The effect of vehicle mixtures on transdermal permeation has been studied using transient flux profiles from porcine skin flow through diffusion cells. Such data characteristically exhibit a large amount of variability between treatments (vehicle and penetrant combinations) as well as noise within treatments. A novel mathematical model has been used that describes longitudinal variation as a time varying diffusivity. Between treatment variability was described by a mixed effects model. A quantitative structure property relationship (QSPR) was developed to describe the effects of the penetrant and vehicle mixture properties on the mean diffusivity and partition coefficient in the membrane. The relationship included terms for the logP and molecular weight of the penetrant and the refractive index of the vehicle mixture with R 2  > 0.95 for K and > 0.9 for partition coefficient (as K  ⋅  D ). This analysis improved on previous work, finding a more parsimonious model with higher predictability, while still identifying the mixture refractive index as a key descriptor in predicting vehicle effects. The concordance with established and expected relationships lends confidence to this new methodology for evaluating transient, finite dose, transdermal flux data collected using traditional experimental methods.