566 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS protection. There are two hairless animal species available, the hairless mouse and the Mexican hairless dog. The former are relatively inexpensive and obtainable in large numbers. Also there is a total absence of pigment cells in the skin of these animals, which eliminates that variable. Since erythema of mild degree is difficult to visualize in this species, it was felt that, instead of evaluating a preparation by determining the minimal erythema dose (MED) of skin treated with a product, a more realistic approach would be to compare the effects following exposures which produce severe damage in unprotected skin. Many sunscreen products are ineffective because they are eliminated by perspiration or removed while swimming (4). Therefore, immersion in water following application of the test products is necessary in a realistic test of efficacy and substantivity. We have also found that when hairless mice are painted with sunscreen preparations their natural grooming instinct results in the removal of some or all of the preparation. As an example, mice painted with 5% p-aminobenzoic acid, a proven effective sunscreen in man, 1 hr before exposure to the sunlamp had the same degree of injury as animals not treated with a sunscreen. We have arrived at dosages of tranquilizers which block the grooming response of these animals until after termination of the ex- posure to the suniamp. In this paper we present a method utilizing hairless mice which we feel is an effective and practical model for the evaluation of the efficacy of sunscreen preparations without, or following, immersion in water. MATERIALS AND METHODS Male hairless mice (JAX hr/hr, Jackson Laboratories, Bar Harbor, Maine), seven to ten weeks of age, were used in these experiments. They were housed in plastic cages, five per cage, on wood shavings and maintained on Purina © Laboratory Chow. The ul- traviolet exposure was based on the method of Owens et al. (5). The mice were placed in groups of five in clean cages 7 in. wide x 12 in. long with 0.5-in. wire screen lidso No shavings were used in the cages and a few fbod pellets were placed on the floor to minimize climbing on the screen lids. Four cages were placed under a fluorescent light fixture with two Westinghouse FS40-T12 lamps. The lamps were 30 cm above the floors of the cages. Three formulations containing sunscreens were tested. Product A was composed of 5% p-aminobenzoic acid in 95% ethanol. Product B contained 5% p-aminobenzioc acid, 55%, alcohol and an undisclosed content listed as skin moisturizers. Product C contained 3.3% octyl dimethyl p-aminobenzoate and 8.6% ethanol in a vehicle which formed a polymeric film on the skin and is easily removed with soap and water. Products A and B contained 0.365 M p-aminobenzoic acid and Product C contained 0.140 M octyl dimethyl p-aminobenzoate. Two test procedures were employed in this study: 1) UV exposure of mice immersed in water following sunscreen applicaiton and 2) UV exposure of mice not immersed in water following sunscreen application. In the nonimmersed study, all animals including controls were administered chlorpromazine hydrochloride, 9 mg/kg intraperitoneally. An additional control group was not treated with chlorpromazine hydrochloride in order to determine any possible phototoxic potential of the medication. A volume of

POLYMERIC FILM-FORMING SUNSCREEN 567 35 /•1 of the appropriate sunscreen was then applied over a 2 x 3.5-cm area on the backs of the mice (5/•i/cm2). One hour after application of the sunscreen, the animals were exposed to the UV source for • •n min. This was the equivalent of 17 x MED, determined by exposing groups of ten untreated animals/group and examining them for erythema at 24 hr. For the immersion study, all animals were first administered 1.5 ml/kg intramuscularly of a 1:10 dilution in 0.15 M saline of Innovar©-Vet (Pitman-Moore, Inc.). The appro- priate sunscreen was then applied as above. One hour after sunscreen application, all animals were immersed individually for 30 min in beakers containing enough water so that the animals could just stand on their hind legs. The water temperature was main- tained at 37øC by the use of a waterbath. Upon removal from the waterbath they were administered chlorpromazine hydrochloride, 9 mg/kg intraperitoneally, and exposed to the UV source for 150 min. All animals were examined daily for four days. Written descriptions of the appearance of the animals were later translated into a numerical scoring system as follows: no visible damage = 0 erythema, dry scaley skin, but normal or nearly normal by 96 hr = 1 scattered whitish patches (edema?) with few scattered pinpoint erosions = 2 large area of whitish, thickened skin with numerous, tiny ero- sions = 3 large erosions 3 to 5 mm long = 4 and erosions 5 mm long = 5. The scores for each animal in the group were totaled and the total divided by the number of animals in the group to arrive at the "average burn score." RESULTS No differences in the various groups could be detected immediately following ex- posure to the suniamps. All animals appeared erythematous. Differences could be dis- tinguished at 24 hr but were much more obvious at 48 hr. In the animals that had received only slight damage, the treated areas appeared visibly normal within 96 hr. Animals that received little or no protection still had gross evidence of severe damage at 96 hr. NONIMMERSION STUDY The average burn scores are presented in Table I. No clear differences could be distin- guished between the nondrug control group and the drug control group. These animals were generally erythematous with small pinpoint erosions and edema at 24 hr. At 48 hr there was marked edema with numerous small encrusted erosions. The sizes of the en- Table I Average Burn Scores of Hairless Mice Exposed to Ultraviolet Light Not Immersed Immersed Treatment N a Score N a Score Control 5 4.0 5 5.0 Drug Control 5 4.0 5 5.0 Product A 5 1.0 5 4.4 Product B 5 1.0 5 4.0 Product C 5 1.0 5 2.0 aNumber of animals.