J. Cosmet. Sci., 64, 181–192 (May/June 2013) 181 Address all correspondence to Kenneth Sloan at sloan@cop.ufl .edu. Fit of fl uxes of sunscreens and other compounds from propylene glycol:water (30:70) through human skin and silicone membrane to the Roberts–Sloan equation: The effect of polar vehicle (or water) solubility KENNETH B. SLOAN, HEMAMALINI DEVARAJAN-KETHA, JENNIFER SYNOVEC, and SUSRUTA MAJUMDAR, Department of Medicinal Chemistry, University of Florida, Gainesville, FL (K.B.S., H.D., J.S., S.M.), Department of Nephrology and Hypertension, Mayo Clinic, Rochester, MN (H.D), and Department of Molecular Pharmacology and Chemistry, Memorial Sloan Kettering Cancer Center, New York, NY (S.M.). Accepted for publication August 29, 2012. Synopsis It would be useful to develop a surrogate for animal skin, which could be use to predict fl ux through human skin. The fl uxes (and physicochemical properties) of sunscreens and other compounds from propylene glycol (PG):water (AQ), 30:70, through human skin have previously been reported. We measured the fl uxes of several of those sunscreens and other compounds from PG:AQ, 30:70, through silicone membrane and fi t both sets of data to the Roberts–Sloan (RS) equation to determine any similarities. For both sets of data, the fl uxes were directly dependent on their solubilities in a lipid solvent [octanol (OCT), in this case] and in a polar solvent (PG:AQ, 30:70, or AQ in this case) and inversely on their molecular weights. The fi t of the experimental (EXP) fl uxes through human skin in vivo to RS was excellent: r2 = 0.92 if the vehicle (VEH) PG:AQ, 30:70 was the polar solvent (RS1) or r2 = 0.97 if water was the polar solvent (RS2). The fi t of the EXP fl uxes through silicone membrane to RS was good: r2 = 0.80 if the VEH PG:AQ, 30:70, was the polar solvent (RS1) or r2 = 0.81 if water was the polar solvent (RS2). The correlations between their EXP fl uxes through human skin in vivo and their EXP fl uxes through silicone membrane were good (r2 = 0.85). In addition, the correlation between EXP fl uxes from PG:AQ, 30:70, through human skin in vivo and their fl uxes calculated from the coeffi cients of the fi t of solubilities, molecular weights and fl uxes from water through silicone mem- branes from a previous n = 22 database to RS was even better (r2 = 0.94). These results suggest that fl ux through human skin can be calculated from fl ux through a silicone membrane. INTRODUCTION Human skin is an effective barrier to the absorption of many of the chemicals with which it comes in contact. This effi ciency has been considered to be a challenge to many in the
JOURNAL OF COSMETIC SCIENCE 182 scientifi c community who have attempted to use topical delivery of drugs as a means to treat local as well as systemic disease states (1). To identify those physicochemical proper- ties of drugs that predict good topical delivery, fl ux (J) into (dermal delivery) and through skin (transdermal delivery) and various experimental (EXP) physicochemical properties of drugs have been collected (2–5). Qualitative analysis of those physicochemical proper- ties of drugs that affect fl ux values has shown that those drugs that exhibit higher lipid (SLIPID) and aqueous solubilities (SAQ) and lower molecular weights (MW) give the higher fl ux values (2–10), regardless of whether a lipid (2,6–8) or an aqueous vehicle (VEH) is used (3,4,9,10). This dependence of fl ux on solubilities in both lipid and aqueous phases is due to the presence of ceramides in what is usually characterized as the lipid matrix in which the corneocytes of the stratum corneum (SC) are embedded. The ceramides have polar func- tional groups attached to long-chain lipid alkyl groups and are arranged in bilayers. The polar functional groups have water molecules associated with them (11), so that to effec- tively pass through the bilayers that comprise the lipid matrix of the SC, drug molecules must actually pass through alternating layers of lipid and associated water layers (the biphasic solubility model) (6,12), and exhibit some solubility in each layer. Thus, any dependence on the water solubility of the permeant is not dependent on its solubility in an aqueous VEH (or any VEH) but is dependent on its solubility in an aqueous (or polar) phase within the lipid matrix of the SC—the assumed rate limiting barrier to fl ux. The qualitative relationship between lipid and aqueous solubilities has been quantifi ed using an extrapolation of Fick’s law: eq (1) (2–4): 1 / ( ) M Mn J D L C C eq (1) In equation (1), the driving force for diffusion of a drug through a membrane is the con- centration gradient between the fi rst few layers of the membrane (CM1) and the last few layers (CMn) (12,13). If it is assumed that CMn approaches zero and that CM1 has reached its maximum value (the solubility of the drug in those fi rst few layers of skin—SM1) (12–17), then the maximum fl ux, JM, is dependent on SM1. Since SM1 is diffi cult to mea- sure, the product of the solubility of the drug in the VEH, SVEH, and the partition coef- fi cient between the fi rst few layers of skin, M1, and the VEH, KM1:VEH or (KLIPID:VEH)y·c (12,15), gives an estimate of SM1. If the lipid, LIPID, is octanol (OCT), and the VEH is water, AQ, expansion of K as solubilities, combined with the solubility in the VEH term, SAQ, gives log SM1 = y log SOCT + (1 - y ) log SAQ + log c. Insertion of this identity into Fick’s law ultimately gives one form of the RS equation: eq (2) (3,4) where water is the VEH and the lipid is OCT. It can be shown that the same equation results if a lipid VEH is used (2). MAQ OCT AQ log J x y log S (1 y) log S z MW eq (2) The RS equation not only tells the investigator what properties to optimize to increase fl ux, but it also tells the investigator what properties to optimize to decreased fl ux (the only
Purchased for the exclusive use of nofirst nolast (unknown) From: SCC Media Library & Resource Center (library.scconline.org)





























































