182 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS rinsed off the skin during normal use soon after application. In these instances, loading a patch with the undiluted product and leaving it in contact with the skin at full concentration for many hours is highly un- realistic and may well prove intolerable for the volunteers. For a shampoo, a more informative method of patch testing is to apply the product to the patches at a dilution of 10-20•o with water. When a physiologically-active constituent occurs in the formula (e.g. an anti- microbial agent) an alternative procedure is to load the patch with a quantity based on an amount per unit area of skin equivalent to the residue left on the scalp after shampooing, as determined by assay. Where the assay of actual residues presents serious difficulty, open patch tests may be carried out by actually shampooing a small test area of skin daily for several days experience suggests that this procedure may even have better predictive value than a closed patch test based on estimates of residual quantities after rinsing. There should thus be no insuperable difficulty in showing an adequate safety margin for the topical administration of cosmetics and toiletries by reasonable exaggeration of the exposure conditions in tests for irritancy. Since, for the reasons already stated, the present authors contend that no allowance is usually necessary for interspecies variation, the choice between using human volunteers or laboratory animals does not appear to have great significance from the investigator's standpoint. If, however, the irritant potential of the test material really is unknown, initial screening is certainly best carried out with laboratory animals. Moreover, if effects on damaged skin are being examined, an animal screening test is obviously desirable before extending the study to human skin, as a reasonable safe- guard for the volunteers. Despite the similarity in responses to irritants generally shown, if interspecies differences are discernible when results for animal and human irritancy tests are compared, greater reliance should obviously be placed on the human data. To avoid uncertainty in extrapolation from animal responses to man whilst minimizing the risk of serious harm to volunteers, human studies may sometimes take the form of tolerance tests. The degree of exposure (with respect to amount of test material applied, its concentration or the duration of contact) is gradually increased until a threshold response occurs. The predictive value of such a test should be satisfactory provided that precautions are taken to avoid fatigue by careful choice of site of application or interval between applications. Human eye irritancy testing should generally follow this type of cautious approach (12). The drawback to more

EXAGGERATED EXPOSURE IN PREDICTIVE TESTING 183 widespread use of the human tolerance test, however, is its time-consuming nature and the difficulty of recruiting volunteers. SENSITIZING POTENTIAL Sensitization is an important type of possible adverse reaction to toiletties and cosmetics it usually involves the risk of causing an allergic contact dermatitis and needs to be considered separately from irritancy. Allergens may sensitize occasional individuals at concentrations which are without any significant effect on the majority of the population this feature of allergenicity causes great difficulty in predictive testing. Some dermato- logists nevertheless take the view that exposing a group of randomly-chosen subjects under test conditions to an exaggerated concentration of a sus- pected allergen increases the chance of recognizing its sensitizing potential. If so, what degree of exaggeration is appropriate? Marzulli and Maibach (15) recently reported a detailed investigation of sensitizing properties using exaggerated concentrations of test materials. For example, with sorbic acid tested at 20•o concentration, one of the 33 subjects they tested gave a positive reaction. The crucial question is what proportion of users, if any, would be sensitized at a typical use-concentra- tion of about 0.1-0.2•o. Clearly this is unanswerable without knowing the shape of the dose-response curve for an allergen of the same or a similar type, extending right down to normal use-level concentrations of the test substance. These authors did, in fact, conduct tests at several concentration levels with a number of compounds but the proportion of subjects with positive responses did not show a dose-response pattern justifying any broad conclusions a graded dose-response relationship may perhaps be inferred on theoretical grounds but even the extensive work carried out by Marzulli and Maibach (15) was evidently insufficient to demonstrate it clearly. Some toxicologists concerned with toiletties and cosmetics prefer to conduct tests for sensitizing potential at normal use-concentration of the test materials, using human volunteers. If such tests involve multiple ex- posures under occlusive patches, the resulting minor degree of skin damage should marginally increase the likelihood of a positive response. However, since it is well known that toiletties and cosmetics in general sensitize less than, say, 1 in 10 000 users, the predictive value of any use-concentration test carried out with only a few hundred volunteers must be exceedingly