208 JOURNAL OF COSMETIC SCIENCE EXPOSURE CONSIDERATIONS RELATED TO HAIR DYE SAFETY William Dressier, Ph.D. Procter & Gamble (Clairol), Stamford, Connecticut The chemistry, formulation characteristics, and mode of use of hair dyes present a rather unique situation with respect to their potential for skin permeation, and pose a challenge in reliably estimating their systemic body burden under use conditions. For example, the oxidative (permanent) hair dye process involves the uptake of colorless, lower molecular weight precursors into the hair cuticle where they react to form colored polynuclear species that become permanently trapped in the hair. Thus, during the exposure period the concentrations of the more permeable dye precursors are reduced, while the increased molecular size of the reaction products serves to limit their penetration potential. In contrast, direct (semi-permanent) hair dyes are preformed colored materials usually of higher molecular weight and greater structural diversity. After coloring they can diffuse from the hair cuticle, resulting in more protracted skin exposure, at least at low levels, as they leach from the hair, from perspiration and showering during wear. Under use conditions, the hair itself serves as a large competing absorptive surface for the hair dyes. Hair may also modify the rate and/or extent of oxidative color reactions. While some of the applied dye may be adsorbed onto the stratum corneum (as evident from skin staining), or absorbed into the viable epidermis, loss of surface corneocytes due to the normal desquamation process serves to reduce the potential for systemic dye absorption. Carefully conducted studies in human volunteers using a series of •4C-ring-labeled oxidative and direct dyes have shown distinctive temporal patterns of systemic absorption characteristic of the chemistry involved. These historical in vivo data have served as useful benchmarks to evaluate the reliability of in vitro percutaneous penetration models using both human and animal skin. Such in vitro models facilitate more precise control or systematic variation of experimental conditions that may influence skin permeation potential. Such variables include dye concentrations and combinations, vehicle formulation components, and exposure time. Moreover, reliable in vitro models, especially those involving the relevant species (i.e., human), are useful to evaluate new dye molecules early in the development state to facilitate dose setting in toxicological studies. Accumulated data from a number of in vivo and in vitro systems has demonstrated that, overall, the systemic absorption of hair dyes is relatively low, with only fractional percentages of the applied amounts penetrating the skin. Because of possible substantial differences in product application rates between the in-use situation and in some in vitro studies, expression of results as a percentage of the applied dose can sometimes be misleading. Therefore, cumulative mass absorbed per unit area (ug/cm 2) is advocated as a preferred measure of percutaneous penetration. Expressed this way, most hair dye penetration data appear to fall within a relatively narrow "dynamic range" from about 0.1 ug/cm 2 to about 5-10 ug/cm 2 following a single application.

2002 ANNUAL SCIENTIFIC MEETING 209 From a toxicological standpoint, such low penetration values generally support relatively large safety margins for hair dyes, even for those dyes with no-observed-adverse effect levels (NOAELs) from toxicology studies as low as 10 mg/kg/day. Consideration of the temporal aspects of hair dye use on a discontinuous basis (e.g., once every 4-6 weeks) serves to further reduce estimated body burdens. As the denominator in the calculation of safety margins, where the numerator (NOAEL) reflects the results of extensive and expensive toxicological evaluations, reliable estimates of the extent of percutaneous penetration are of critical importance in the risk assessment process.