TESTING DRUGS FOR DERMAL TOXICITY 391 The intact dermis forms a natural barrier to absorption of compounds through the skin but skin absorption can be altered due to changes in formulation and also the state of the skin itself. If the skin is damaged then the dermal barrier is removed and considerable absorption can occur. The effects may be studied by placing the substance on the skin and observing the subsequent toxic effects which occur. This problem may be studied in a rather simpler way. We determine the amount of material we are likely to put on the skin. This is related to the chronic toxicity of the compound when given by the oral route in a normal feeding trial. If only a fraction of 1% of that substance in the formulation is to be applied to the skin, it does not seem necessary to carry out long-term skin toxicity tests. Another approach is to determine the amount of the substance absorbed and to relate this to the known toxicity of the compound. Thus we can assess the safety margin. If we wish to carry out toxicity tests following dermal application, problems do arise. We must be quite certain that oral contamination does not occur. There are substances we put on the skin which when taken orally would be highly toxic. Lastly I would draw attention to danger of carcinogenicity with sub-. stances applied to the skin. This is a very real problem we have to face. Here we need to apply the substance to the skin, usually the mouse or the rabbit, over a period of 12 months and then continue our observations of them for a future year. Again we look for local effects, but we must remember that a carcinogen may be absorbed through the skin and produce distant carcinogenic effects. We need not cover the skin to prevent oral contamina- tion. DISCUSSION MR. J. SOUTHWOOD: I would like to have your opinion on the value of' tests for carcinogenicity by the subcutaneous route. How would you view a substance which shows activity by this route and not by skin application ? TH• L•CTUR•R: We all know that when we inject substances into the rat subcutaneously we may get false positive results or at least we think so. Even injecting a solution of glucose subcutaneously for a period of time can cause rumour-formation in the rat. We therefore have to be very careful in controlling these things and interpreting them. I feel that if there is a suspicion of carcinogenicity we should avoid the use of the substance unless there is a very real need for it. In the use of a medicinal substance, which is essential for a medical treatment, we may have to take a certain amount of risk--no one would stop using X-rays because there is a danger of carcinogenicity. We just have to accept a degree of risk.

392 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS DR. I. LEVENSTEIN: You mentioned the fact that you were able to obtain good responses with intradermal injections, and you feel that you can then screen out fairly hazardous materials without using the eye. I just wonder if you are using the same physiological system to screen out these prepara- tions. Injecting into the skin, and applying to a mucous membrane are entirely different. Can you correlate the two responses ? TI•F. L•CTURF. R: I am in complete agreement with you that application to a mucous membrane is rather different from injecting intradermally into the skin, but I do feel that the skin test picks out very irritant substances, like phenols and nitric acid, which would cause severe damage to the eye. All I am saying is that we must be a little humane in eye studies. If we have a substance, which in the intracutaneous test looks remarkably irritant, we should approach the eye test with caution. Dilute the test solution first--do not just instil the concentrated solution into the eye and blind the rabbit. There is no virtue in that at all. DR. L. GO•.BF. RG: I wonder whether you would comment on the work that has been carried out on the importance of lubricating oil on the electric clippers which are used to prepare the skin, and their influence on the result ? There is also the question of the use of a positive control when carrying .out tests for primary irritation. I have always looked on hydroxycitronellal as a very reliable positive control material. I may be quite wrong there, but I have heard of cases recently where it was applied to human skin and in the whole group it produced negative results. I would be interested to hear what other people use as a positive control material. T}•F. LECTURF. R: I do not think that the minute amount of oil that is -really necessary can have any influence at all. Perhaps it depends on the type of clippers which one uses. With regard to controls, I am in complete agreement. What we do need are substances of known irritancy in man, which can be used as controls for •comparisons in our animal tests. There is a great lack of information on substances which are known to be minor irritants in man I have sorbic acid in mind which shows nothing in the animal but does in man. There are a number of these examples which I think Dr. Levenstein has drawn .attention to. What is the reliability of our animal tests in these border line substances ? DR. P. H. WlTJENS: We have tested dinitromonochlorobenzene by intradermal injection in the normal guineapig sensitization test, and found that this product has no sensitizing properties according to the requirements ß of the F.D.A. because the reaction obtained in the first tests is just the :same as in the later challenge tests. According to the test requirements it