j. Soc. Cosmet. Chem., 46, 221-229 (July/August 1995) Predicting absorption of fragrance chemicals through human skin JURIJ J. HOST•NEK, Euromerican Technology Resources, Inc,, Lafayette, CA 94549. Received March 25, 1995. Synopsis Potts and Guy fitted an equation to the experimental permeation coefficients (Kp) of a variety of molecular structures through human skin (1). This equation was based on molecular weight (MW) and the octanol- water partition coefficient (log P). In the endeavor to identify a mathematical model predictive of the mostly unexplored skin penetration of fragrance chemicals, the Potts-Guy equation was tested here by correlating the published, measured Kp for 20 such compounds with predicted Kp values. The data, covering a broad range of structures, are well described by that model (r 2 = 0.86, p 0.001). Also, calculated octanol-water partition coefficients, an alternative in case of missing measured values, appear well founded when analyzed statistically: comparison of 33 measured partition values with the calculated log Ps show near-perfect correlation for the intermediate lipophilicity range typical of fragrance chemicals (0.5 log P • 3.5 r 2 = 0.97, p • 0.001). The Potts-Guy relation thereby recommends itself as a simple and reliable method for the quantitative evalu- ation of human skin absorption potential of small-molecular-weight nonelectrolytes, and of fragrance chemicals, in particular where such data are largely missing. INTRODUCTION Exposure to fragrances is becoming unavoidable in the course of daily activities, as an ever-growing range of consumer goods is perfumed, either for functional or esthetic reasons. Most exposure is olfactory, and thus fleeting, and with the rare exception of individuals suffering from chemical hypersensitivities, without notable adverse effects. Through one category of products, however, exposure to fragrances is immediate, repeated, and significant, and dermatologists as well as toxicologists are becoming increasingly aware of the potential for toxicity. A large number of consumer products, namely the category of cosmetics and toiletries designed for contact with the skin, mostly as so-called stay-on products, contain a relatively high percentage of fragrance materials. Perfumes and colognes consist entirely of highly complex fragrance mixtures, This paper celebrates the 70th birthday of Bruno P. Vaterlaus, Dr. Sc. Techn. ETHZ, who for many years directed the course of research at Givaudan with great integrity and vision, through good, and sometimes very difficult times. 221

222 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS natural, nature-identical, or fully synthetic. As far as they contain natural oils and essences, they rate a low priority for purposes of human health concerns, as those currently in use have been proven safe based on an age-old history of human exposure without adverse effects of note. However, some 3000 synthetic chemicals are currently also used in fragrance compounding (2). They find increasing application in personal care products, for esthetic as well as economic reasons yet only a handful have been tested on human or animal skin for permeation, and thus systemic availability assess- ment. It therefore appeared appropriate to evaluate the relevance of currently developed mathematical models as an expeditious approach for the prediction of the human skin penetration potential by fragrance chemicals. METHODS PERMEABILITY COEFFICIENT Kp In order to avoid or minimize the ethical, economic, and biological difficulties associ- ated with in vivo experimentation, in vitro studies using human skin are now being used most frequently for estimating dermal absorption. From such studies, where a standard protocol has been observed, a statistically adequate number of data has now been generated that lends itself to mathematical modeling for QSAR purposes. Common parameters that result from such percutaneous absorption experiments are the perme- ability coefficient Kp, defined as Jss = KpC, where Js• and C are the steady-state flux and donor concentration, the latter maintained at an unchanged (or infinite) value. The rate at which a chemical penetrates the skin, in the present paper expressed in cm/hr, is described as the permeability coefficient Kp or as its logarithm, log Kp•a value determined over sufficient time to assure conditions of steady state. When the permeant is applied from aqueous solution, it represents the in vitro method by which most compound-specific penetration data currently available in the literature have been gen- erated. In such studies, a piece of excised human skin is attached to a diffusion cell consisting of a top chamber to receive an adequate volume of the penetrant in solution, an O-ring to secure the skin in place, and a temperature-controlled bottom chamber containing saline and equipped with a sampling port to withdraw fractions at specific time intervals for analysis (2a). Although forearm skin is an ideal choice, most commonly abdominal skin obtained at autopsy is used for reasons of availability. It is now generally accepted that properly conducted in vitro tests using human skin yield a reasonably good predic- tion of the absorption rate for chemicals in humans. OCTANOL/WATER PARTITION COEFFICIENT Poet Polarity was found empirically to be a critical, if not the most important, physicochem- ical parameter determining the ability of chemical compounds to penetrate the skin barrier. It describes the partitioning behavior of a solute between a lipid and a polar phase, typically n-octanol and water. Expressed as the partition coefficient Poet or its logarithm log Poet, it represents the solvent system most frequently used to describe the polarity of chemicals.