418 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Studies on reversibility of effect were not included in these investiga- tions. DISCUSSION AND CONCLUSIONS It is immediately apparent that there is poor concordance in the results obtained by the different methods. For example, T9 fluctuates between most irritant, intermediate and least irritant for both human skin and animal skin. Indeed, Smeenk {2) has concluded that the in vitro tests he used with human skin and the patch tests carried out with human volunteers were not suitable for screening the irritancy of detergents on the skin. This conclusion presupposes that the arm immersion test is the method to give the most meaningful results but it is well-known that many factors (e.g. size of test population, representativeness of the test group in relation to the population "at risk", ambient climatic conditions, etc.) affect the out- come of arm immersion tests (7-9). In a detailed report on a wider range of chemicals {10) it has been sug- gested that the guinea-pig might serve as a useful alternative to the rabbit in the assessment of skin irritancy, but it is our opinion, based on a much more varied range of chemicals than are reported here, that the guinea-pig and the rabbit tests are complementary rather than alternative to each other. However, given the same test conditions, there were no apparent differences in the reactivity of rabbit and guinea-pig skin with these surfac- tants. There was noticeably good agreement between the results obtained with occluded patches on formalin treated rat skin and those on untreated rabbit skin. However, in our laboratory we have been less successful in discriminating between a series of concentrations of sodium lauryl sulphate than were the originators of the formalin treated rat skin method {11, 12). The skin of "hairless" mice behaved in an anomalous manner under the influence of the nonionic detergent T4. Using histochemical methods three dehydrogenase systems were stimulated (Table II). Monoamine oxidase activity was also increased by T4. The only other surfactant in this series with an effect on monoamine oxidase was T8 and this is less surprising because the surfactant is a derivative of dimethylamine and a latent reserve of monoamine oxidase is known to be present in mouse skin (13). The lack of correlation between oxygen uptake in vitro and exposure of "hairless" mouse skin to these surfactants has already been reported (6). A heterogeneous group of chemicals such as surfactants may react with skin in many different ways (14). Given different conditions of exposure the same chemical may behave in a different manner in contact with the skin

PREDICTIVE IRRITATION TESTS WITH SURFACTANTS ON SKIN 419 (e.g. T6, a soap famed for its "mildness" to skin, was at the low end of the irritancy scale in uncovered tests in humans and animals but in the medium irritancy grade in patch tests, and in the most irritant grade in some of the biochemical tests). In reality, the skin is seldom abused by a surfactant alone the presence of fillers, perfumes and other components may modify the irritancy potential substantially [e.g. T8 was found to be among the most consistently irritant materials in this series and yet other laboratories have shown (15), and we have confirmed (unpublished results), that formu- lations of T8 with some added anionic surfactant are less irritant than is T8 alone] however, an analysis of the chemical and other factors involved are clearly outside the scope of these investigations. From these data we conclude that preliminary predictive tests of the type described using animals are at least as useful as comparable tests in humans. If appropriate, human studies should be undertaken after pre- liminary animal experimentation, but these human studies must be on an adequate scale and statistically designed to be meaningful (16). In the long term, studies on the biochemical changes brought about by surfactant-skin interactions will aid understanding of the aetiology of detergent dermatitis. Currently such studies do not offer much hope for simplified predictive testing. (Received: 23rd February 1971) REFERENCES (1) Smeenk, G. De Invloed van Detergentia op die Huid. Doctorial thesis, Leiden University, Netherlands (1968). (2) Smeenk, G. The influence of detergents on the skin (a clinical and biochemical study). Arch. Klin. Expl. Dermatol., 9,8õ 180 (1969). (3) Smeenk, G. and Polano, M. K. Methods for comparative estimation of the irritancy of various detergents on human skin. Trans. St. John's Hosp. Derrnatol. Soc., õ1 90 (1965). (4) Polano, M. K. The interaction of detergents and the human skin. J. Soc. Cosmet. Chem., 19 3 (1968). (5) Brown, V. K. H. and Clarke, R. A. Sulphan blue as an aid to the laboratory assessment of primary skin irritants. J. Invest. Derrnatol., 45 173 (1965). (6) Brown, V. K. H. The influence of some detergents on the uptake of oxygen by "hairless" mouse skin. J. Soc. Cosmet. Chem. 20 413 (1969). (7) Justice, J. D., Travers, J. J. and Vinson, L. J. The correlation between animal tests and human tests in assessing product mildness. Proc. Sci. Sect. Toilet Goods Assoc. No. 35 12 (1961). (8) Brunner, M. J. Pitfalls and problems in predictive testing. J. Soc. Cosmet. Chem. i8 323 (1967). (9) Idson, B. Topical toxicity and testing. J. Pharm. Sci. 57 1 (1968). (10) Roudabush, R. L., Terhaar, C. J., Fassett, D. W. and Dziuba, S. P. Comparative acute effects of some chemicals on the skin of rabbits and guinea-pigs. Toxicol. Appl. Pharma- col. 7 559 (1965). (11) Finkelstein, P., Laden, K. and Miechowski, W. New methods for evaluating cosmetic irritancy. J. Invest. Dermatol. 40 11 (1963).