754 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS microscopic examination of epidermal cornified cells removed using a suitable adhesive tape, is thought to provide information about changes in the stratum corneum. Although epidermal stripping is best carried out on small areas of skin this is not regarded as a serious disadvantage. Epidermal samples from human or animal skins may be obtained using blisters raised by suction or application of chemical vesicant. Neither technique is con- sidered suitable for routine use in safety evaluations. Changes in the levels of enzymes in skin treated topically with irritants are likely to be sensitive markers of tissue damage, but since these assays require samples of tissues obtained at autopsy or from biopsy, they are restricted for use only in animal studies. A recommended investigation for primary irritants would ideally com- mence with animal studies and then be followed by trials using human volunteers. In animal studies, it is also considered advisable to obtain samples of skin from areas of treatment and to examine these histologically and histochemically for minor damage which is not visible to the naked eye. In human studies where this is not normally practicable, tape stripping is likely to be helpful in detecting minor changes. In testing for skin irritation, it is appropriate to allow a margin of safety. The use of a sensitive species like the rabbit and the use of the occlusive patch test go some way towards achieving this end, but it is difficult to quantirate this safety margin since the relative sensitivity of the rabbit to man will vary for different compounds. A better guide to the safety margin would be obtained using a series of concentrations of the substance to be tested, without substantially altering the conditions of test. For products that are washed away, prolongation of the time of contact beyond that envisaged for the eventual human application will give an added safety factor. The use of higher concentrations or prolongation of contact of the test material with the skin could be considered as a further safety measure, not only in animal tests but with due caution in human subjects also. SPECIALIZED TESTS Respirat9ry studies on the skin The observation that the respiration of the skin is influenced by its state of health (122) and its physiological condition (130, 131,128)suggests that a study of oxygen uptake might reflect the metabolic state in the tissue

APPRAISAL OF METHODS FOR DETECTING PRIMARY SKIN IRRITANTS 755 and serve as an early indication of damage. This view is supported by the observation that, in skin treated topically with carcinogens, antiseptics, some cosmetic chemicals and solvents, all of which are known to damage the tissues, oxygen uptake is depressed (132-134). Changes in the skin which might be recorded as a reduction in cutaneous respiration include alterations in the pH of the surface layer due to changes in the composition of the sebaceous secretion (132-135) and in the meta- bolism of the epidermal cells. Amersbach and his colleagues (133) used the Warburg apparatus in their investigation of respiratory changes in skin samples obtained by biopsy from patients treated topically with irritants. Skin samples were incubated with a suitable substrate and the oxygen uptake measured manometrically. Alternatively, samples of normal skin were incubated in the substrate which contained the compound under investigation. Changes in the respiratory rate shown in such an experiment would thus reflect the progressive alterations in the metabolism of the skin sample. Although potentially useful, the excessive variations which exist in the respiration of the human skin depending on race, sex, genetic type and environmental background would seem to present considerable difficulties in the interpretation of the results of such an investigation. Even in a single individual, the respiration of the skin would be expected to vary according to the part of the body from which the biopsies are taken (136). Furthermore, as Amersbach and his colleagues (133) pointed out, in the evaluation of the safety of cosmetic substances where tests are normally carried out using occlusive conditions, the respiration even in undamaged skin is likely to be lower than normal. Occlusive patches are known to cause increased sweat- ing, hyperthermia and inflammation (4). Therefore, unless the cutaneous damage caused by the test substance is sufficient to cause appreciable changes in the rate of oxygen uptake, the technique would appear to be insufficiently sensitive. Recent evidence that skin respiration studies are of doubtful value in the investigation of skin irritants, was given by Brown (137). Although this author believed that changes in respiratory activity might be indicative of alterations in carbohydrate metabolism (138-142), he found that the changes in respiratory activity were less than differences shown in patch tests. Thus, on available evidence, the Warburg technique for detecting changes in skin respiration would seem to be an unreliable way of detecting damage. Using the Cartesian diver method, no more success was achieved in