TESTING DRUGS FOR DERMAL TOXICITY 393 has no sensitizing properties although we know that it is one of the strongest sensitizers when used on the skin instead of by intradermal injection. I cannot explain it, but we have to accept it. THE LECTURER: To avoid confusion, I am speaking of two types of intradermal tests one for local irritant properties--you are quite clear about this. With regard to sensitization tests I feel quite strongly that any sensitization studies of this nature in animals are quite useless we need to do them in man--this applies also to drugs. You cannot sensitize animals to penicillin and yet it is known that this is a very real problem in man. We can only do it in man himself. In this country I do not think that we place any reliance at all on the animal tests for sensitization. DR. B. LESSEL: I should like to comment on the predictive value of the rabbit skin test. Nine topical antibacterial creams, as marketed, were examined for primary irritation by repeated application to rabbit skin and by patch tests in human volunteers. Some provoked a severe skin reaction in rabbits and had little or no effect on human skin while other preparations had the converse effect. The skin reaction in rabbits that appeared most predictive of irritancy to human skin was oedema or necrosis, whereas erythema of rabbit skin appeared to be a non-specific reaction. }V[R. J. }V[cL. PHILP: Have you had any experience of the use of rabbits for photosensitizing tests, and if so, could you give some details ? Do you place greater reliance on photosensitization tests in the rabbit than on guineapig sensitization tests ? I disagree with your statement that animals are of little value in sensitiza- tion tests, and that studies can only be made on man. One cannot have certainty from either animal or human tests. The animal test can be most useful, and may be a valuable prelude to human tests. Your reference to penicillin as being a sensitizer to the human was hardly relevant as it would have required a test panel of some thousands of human beings to show this sensitivity prior to the use of the material.. Tests carried out on humans can give erroneous results--nickel, for example, is well-known as a cause of contact sensitization in the human but cannot be shown to be a sensitizer by human tests it is, however, confirmed as a sensitizer using the guineapig test. THE LECTURER: In general I agree with the sensitization test in guinea- pigs. I was not suggesting that the test be abolished, it may be of some value. What I am stressing is that the test is not reliable. It will detect potent sensitizers. These substances would affect nearly everyone on whom they are applied. We are more concerned with the abnormal reactor that is the one person, perhaps one in a thousand, who has a different skin or

394 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS constitution than everyone else. This is a very real problem which we also have in the drug industry. We cannot detect these abnormal reactors in ordinary clinical thais, and they only become apparent in widespread use. In the cosmetic industry you become aware of them from the number of complaints which are received. In the drug industry the need is felt for an early warning system so that the doctor who observes them can draw our notice to them, and then we can determine their incidence. In sensi- tivity testing in animals we cannot predict with any degree of certainty what will happen in humans.