j. Soc. Cosmet. Chem., 48, 289-295 (November/December 1997) In vitro assay of high-SPF sunscreens R. P. STOKES and B. L. DIFFEY, Regional Medical Physics Department, Dryburn Hospital, Durham DH1 5TW, UK. Accepted for publication December 1, 1997. Synopsis In vitro spectral transmission measurements using excised human epidermis as the substrate were used to determine the photoprotection provided by physical and organic chemical sunscreens encompassing a wide range of sun protection factor (SPF). The measured SPFs were in good agreement with the quoted SPFs of the products. This in vitro technique using human epidermis could prove reliable for evaluating the SPF of high-protection sunscreens for which in vivo assay is problematic due to the impractically long irradiation times to achieve erythema on sunscreen-protected skin. We also compared our calculated SPFs, which assumed a natural sunlight spectrum, with those that would have been obtained assuming a xenon-arc solar simulator spectrum. We found that for products with relatively low UV-A absorption, the use of a solar ß simulator for in vivo testing overestimates the SPF that would be expected in sunlight. INTRODUCTION The photoprotection provided by a sunscreen product is assessed in terms of its sun protection factor (SPF). Sunscreen SPFs are generally measured by in vivo assay and defined as the ratio of the ultraviolet (UV) dose required to cause minimal erythema in protected skin to that required for unprotected skin. Internationally agreed procedures (1,2) define protected skin as that to which a 2-mg/cm 2 layer of sunscreen has been applied. In vivo assay is problematic for high-protection sunscreens (SPF 25) because of the impractically long UV irradiation times and variability of results (3). Conse- quently, it would be particularly desirable to use a reliable in vitro assay for these products, whereby the transmission of UV radiation is measured first through a substrate and then through the substrate with applied sunscreen. The ratio of UV transmission without sunscreen to that with sunscreen gives a measure of photoprotection. A wide range of substrates has been used for in vitro assay, including wool, pig skin, hairless mouse epidermis, human epidermis, human stratum corneum, synthetic skin casts, and surgical tape (4). Unfortunately, most substrates are least reliable when products offering high protection are assayed. The substrate expected to give results closest to in vivo assay is human epidermis. In the study reported here, we have applied physical sunscreens, with varying concen- trations of the same active ingredient, and organic chemical sunscreens to excised human epidermis, and using a spectral transmission technique (5), measured the SPF at an 289

290 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS application thickness of 2 mg/cm 2. Our aims were to determine whether the technique was a reliable way of measuring high SPF (25) products and to evaluate the difference between SPFs measured in vivo using a xenon arc solar simulator and those expected in natural sunlight. MATERIALS AND METHODS ISOLATION OF EPIDERMIS Skin was taken from the underside of the female breast during the operation of breast reduction. It was obtained by a process known as de-epidermalization, the principle of which is to remove the epidermis and epidermal appendages while leaving the deepest layers of the dermis in situ. The skin (approximately 10 x 4 cm) was received within one day of surgical operation, and these strips were cut into squares of approximately 4 cm x 4 cm. The samples of skin were placed in a water bath at 60øC for 45 seconds (6). On removal from the water bath, the epidermis was gently separated from the dermis by careful peeling. Epidermal sheets were stored in physiological saline at 4øC until re- quired, which was normally within five days. Sheets of epidermis can be stored at 4øC for several weeks without loss of barrier function (6). SUNSCREEN PRODUCTS Five physical sunscreen products were used, each containing titanium dioxide as the sole active ingredient at concentrations of 4.4%, 6.9%, 7.8%, 8.6%, and 12.0%, respec- tively. The first four products were commercially available and had quoted SPFs of 8, 15, 25, and 35. The product with the highest concentration was not yet available commer- cially but was expected to have an in vivo SPF of 35 or higher. These sunscreens are identified as P8, P15, P25, P35, and P35+, the numbers denoting nominal SPF. Five organic chemical sunscreens were used and these are identified as C5, C15, C20, C30, and C50, the numbers again denoting quoted SPF. The active ingredients contained within these sunscreens were as follows: C5: Butyl methoxydibenzoylmethane and methylbenzylidene camphor. C15: Butyl methoxydibenzoylmethane, methylbenzylidene camphor, and octyl salicyl- ate. C25: Butyl methoxydibenzoylmethane, methylbenzylidene camphor, octyl methoxycin- namate, and titanium dioxide. C30: Butyl methoxydibenzoylmethane, methylbenzylidene camphor, octyl salicylate, and titanium dioxide. C50: Octocrylene, octyl methoxycinnamate, octyl salicylate, and oxybenzone. EXPERIMENTAL TECHNIQUE A piece of epidermis (2 x 2 cm) was placed over a circular aperture of diameter 1.5 cm cut into an aluminium holder. The holder was positioned so that the circular aperture was directly over the teflon input optics of an Opttonic model 742 spectroradiometer controlled by a Hewlett Packard HP85 microcomputer. Radiation from a 75 W xenon arc lamp (filtered by a Schott UG5 filter) was directed onto the epidermis via a light