100 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS and spread evenly over the entire surface. Additional periodic anesthetization was carried out with sodium pentobarbital (30 mg/Kg, i.p.) over the duration of the entire exper- iment. One hour following application, the formulation on the skin was removed by swabbing thrice carefully with Kimwipes ©. At 0, 2, 4, and 8 h following removal of formulation, the animals were sacrificed by a lethal dose injection of sodium pentobar- bital. Full-thickness dorsal skin was then carefully excised, and the liver and urinary bladder were harvested. The skin section was then mounted on a wooden board and stripped as follows: two pieces of adhesive tape (Scotch Magic © Tape, St. Paul, MN), each 1.9 cm wide and about 6 cm long, were placed alongside each other on the skin surface to strip the skin. The tapes were of sufficient size to cover the area of skin that was in contact with the test formulation. The skin was repeatedly stripped with fresh twin tapes, usually about 15 times, until it appeared shiny and glossy. The remaining skin and bladder, along with the surface swabs and strips, were then assayed for glycolic acid or glycerol content using a scintillation counter after addition of 15 ml of Ecolite(+) scintillation cocktail (ICN Biomedicals, Inc., Irvine, CA). The amount adhering to the stratum corneum surface is operationally defined as the amount found in the first two strippings. The amount of glycolic acid found in the stratum corneum is then defined as the amounts found in tape strippings 3 through 15. The amount of glycolic acid in the living skin strata is defined as that determined by analysis of the remainder of the full-thickness skin. Liver samples were analyzed following combustion in a Model 306 tissue oxidizer (Packard Instruments). All experiments were carried out under non-occluded conditions, and a minimum of three animals per formulation per time point were used. In vitro diffusion experiments Male hairless mice (SKH-hr-1, 50-60 days old, Charles River Breeding Labs) were sacrificed by lethal dose injection of sodium pentobarbital, and full-thickness dorsal skin was excised. Subcutaneous fat was carefully removed using a dull scalpel. Appropriate- sized pieces of skin were then mounted on Franz diffusion cells with a surface area of 1.77 cm 2 and a receiver capacity of 7 ml (Crown Glass, Somerville, NJ). The epidermal side of the skin was exposed to ambient conditions while the dermal side was bathed by a 0.05 M isotonic HEPES, pH 7.4, buffer. The receiver solution was stirred continu- ously using a small Teflon-covered magnet. Care was exercised to remove any air bubbles between the dermis side of the skin and the receiver solution. The temperature of the receiver solution was maintained at 37øC. Following mounting of the skin, 25 }xl of the test formulation was applied to the epidermal surface of the hairless mouse skin and carefully spread evenly to achieve complete surface coverage. A minimum of three cells, using skin from at least three different animals, was used for each system tested. At 16 h, the diffusion setup was dismantled. Upon dismantling, the donor cap was rinsed in 10 ml of distilled water followed by a 20-ml methanol rinse. The methanol rinse was allowed to dry in a hood, at which time scintillation cocktail was added. The skin piece was then mounted on a board, and the epidermal side of the skin was swabbed carefully three times with a dry Kimwipe © to remove any undried formulation. The skin was then stripped repeatedly as follows: A piece of adhesive tape (Scotch Magic Tape, 810, 3M Commercial Office Supply Division, St. Paul, MN), 1.9 cm wide and about 6 cm long, was used to strip the skin. The tape was of sufficient size to cover the area of skin that was in contact with the test formulation. At least nine strippings were carried out, and

DEPOSITION OF GLYCOLIC ACID AND GLYCEROL 101 each strip was analyzed separately for radiolabeled drug. If at the end of nine strippings, the skin did not appear shiny and glossy, additional strippings were carried out until a glossy appearance ensuring complete removal of the stratum corneum was achieved. The remaining skin and the receiver solution were then assayed for radiolabeled drug. Assay of the donor rinses, surface swabs, strips, remaining skin, and receiver solution was carried out using a scintillation counter after addition of 15 ml of Ecolite (+) (ICN Biomedicals Inc., Irvine, CA) to each sample. Decarboxylation of glycolic acid by enzymes in the skin In order to determine the extent to which the glycolic acid marker may be decarbox- ylated by skin enzymes, glycolic acid formulations were incubated with freshly excised skin as follows: Hairless mouse skin was freshly excised and subcutaneous fat was removed. Skin pieces of approximately 2 cm 2 were homogenized (Tissumizer) in 1 ml isotonic pH 7.4 Hepes buffer each, and 200 p•l of 4% glycolic acid aqueous solution or Non-1 liposomal formulation was added to aliquots of the skin homogenate in scin- tillation vials. Following thorough mixing on a vortexer, the vials were capped and incubated at 37øC in an oven for varying periods of time. At 2, 4, and 20 h after incubation, the vials were uncapped and allowed to stand at room temperature in an air hood for 1 h. Assay for radiolabeled content was then carried out using a scintillation counter after addition of scintillation cocktail. The extent of loss of glycolic acid label was calculated by comparison with 200-p•l control formulations. Assays were conducted in triplicate for each formulation at a given time point. RESULTS AND DISCUSSION Table I shows the distribution of glycolic acid into various strata of hairless mouse skin as a function of time after topical in vivo application of various formulations for a period of 1 h. Pairwise comparisons were carried out using two-tailed Students's t-tests to obtain significance of differences between various formulations. In general, the accu- mulation of glycolic acid in the stratum corneum was in the order: aqueous solution = Non-1 = Non-2 W/O emulsion = O/W emulsion = 30% PG solution, both at 1 and 8 h. The amounts of glycolic acid in the living skin strata from Non-1 formulations were significantly higher than those from all other formulations at all time points examined except at 8 h. At 8 h Non-1 was similar to Non-2 and the W/O emlsion system. The rest of the formulations were similar to each other at all time points except for the 30% PG solution, which showed the poorest deposition behavior at all time points tested. The amounts in the urinary bladder at 8 h were significantly higher from Non-1 compared to the others, which were similar to each other. For Non-1 and Non-2 formulations, an inverse relationship between the accumulation of glycolic acid in the living skin strata and time was observed. It is also noted that for the liposomal formulations, the amount of formulation removed at the end of the 1-h application period is two- to threefold lower than for the other formulations and that the amount of glycolic acid found in the stratum corneum was twofold higher for the liposomal formulations. The amounts of vehicle remaining on the surface of the skin from liposomal formulations decreased as a function of time. These results suggest that