J. Soc. Cosmet. Chem., 40, 109-117 (March/April 1989) The role of the resident microflora in the pathogenesis of dandruff D. SAINT-LEGER, A. M. KLIGMAN, and T. J. STOUDEMAYER, Laboratoires de Recherche de L'Ordal, Dgpartement de Biophysique, 93601 Aulnay-Sous-Bois, France (D.So -L. ), Department of Dermatology, University of Pennsylvania, Philadelphia, PA (A.M.K.), and Biosearch, Inc., 3408-50 B Street, Philadelphia, PA 19134 (T.J.S. ). Received November 29, 1988. Synopsis Current thinking implicates P. ovale as the cause of dandruff the condition invariably resolves when this yeast is suppressed. However, P. acnes and coagulase-negative cocci are also abundant on the scalp. A contributory role for those members of the resident microflora has not been ruled out. We used a new technique to collect and weigh scales. We have determined the percentage of nucleated cells in scales, a measure of inflammation. The density of P. acnes, P. ovale and cocci were followed during and after treatment with antibacterial solutions and two anti-fungal shampoos, Octopirox © and Magnesium Oma- dine © . The level of dandruff could not be correlated with changes in the numbers of aerobic and anaerobic bacteria. We concluded that resident bacteria probably play no role in the etiology of dandruff. Octopirox © shampoo was more efficacious than Magnesium Omadine © in reducing scaling and P. ovale. While P. ovale is necessary, Koch's postulates have not been fulfilled. The cause(s) of drandruff is (are) still unknown. INTRODUCTION The modern method of treating dandruff with antifungal shampoos grew out of demon- strations that the active ingredients inhibited the resident yeast, P. ovale, in vitro and in vivo (1-4). Scales, the sole manifestation of the disorder, practically disappear whenever P. ovale is eradicated. When treatment stops, the return of scaling parallels the restora- tion of P. ovale. A single discordant result from the University of Pennsylvania group was technically flawed in using a pharmaceutically questionable formulation of ampho- tericin (5). The data base strongly suggests that P. ovale is the determinant causal factor in dandruff, a position vividly championed by Shuster (6). Still, Koch's postulates have not been fulfilled. In one of these postulates, the pathogenicity (distinct from an adven- titious form) is strictly defined by the fact that the "microbe" is found in a lesion. It 109

110 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS seems very unlikely that dandruff is simply an infection by a pathogenic yeast since it is recovered from normal scalps. An alternative explanation is that increased scaling is the result of an underlying inflammatory process (7), throwing up clumps of thickened, but porous, horny layer in which P. ovale flourishes. Mediators produced by a greatly ex- panded population of P. ovale could then percolate downward, aggravating the inflam- matory process. Indeed, it is known that P. ovale can induce inflammation through activation of the complement (8), leading to the release of chemotactic factors. The latter would then attract neutrophils into the epidermis, disturbing keratinization. In this scenario, P. ovale would function in a secondary role. The response of scalp psoriasis to ketoconazole has similarily been ascribed to the pro-inflammatory activity of a large population of P. ovale (9). Neither we nor others have found a way to induce dandruff in non-dandruff subjects. We assume that the etiology is complex and is influenced by various constitutional factors, including heredity. We lack a good understanding of why a minority of persons exhibit clinical dandruff when all individuals produce some scales. The difference between dandruff and non- dandruff seems to be merely quantitative, greater and larger scales (10). We undertook this study with several questions in mind: 1) Since the dandruff scalp also richly harbors P. acnes and cocci, can these aggravate scaling? 2) In view of improved methods of assessment, how good are the correlations between (a) weight of collected scales, (b) clinical grades, (c) P. ovale counts, and (d) the index of inflammation, based on the percentage of nucleated cells in scales? (3) Finally, Octopirox © (ethanolamine salt of 1, hydroxy-4-methyl-6-(2,4,4 trimethyl pentyl)-2-(1H)-pyridinone) and Magnesium Omadine © (magnesium salt of 2-pyridinethiol-1-oxide) are fungistatic agents included in popular shampoos, at least in the European countries. Can these be discriminated in regard to efficacy with the more sensitive methodology now available (10)? MATERIALS AND METHODS GENERAL DESIGN Twelve white males, aged 24-43, with moderately severe dandruff served as paid volunteers. By clinical criteria none had seborrheic dermatitis. During a preparative period of three weeks, they shampooed their scalps thrice weekly with non-medicated 6.25% lauryl ether sulfate (LES) provided by us. Scale production was measured ac- coMing to our recently published method (10). Basically, scales were harvested by voluminous shampooing and the wash water filtered to collect all the scales. These were dried and weighed. Scales were always harvested two days after shampooing, when scale production nears a plateau. The severity of dandruff was assessed on a 0 to 10 clinical scale. Grades 2 and below designate non-dandruff. Grade 3 is marginal dandruff. Grades 4, 5, and 6 reflect respectively mild, moderate, and severe dandruff. Our sub- jects were chiefly grade 5. The plan of the study is shown in Figure 1. Two groups, A and B, comprising six subjects each were compared. In the three weeks pretreatment phase the scalps were shampooed thrice weekly with 6.25 LES. In phase I, lasting four weeks, Group A received 5 ml of an ethanol:water solution (1:1), 5 ml/scalp, thrice weekly. Group B received an ethanolic solution (1:1), 5 ml/scalp, containing 1% clin- damycin hydrochloride and 1% Octopirox ©, thrice weekly. The latter formulation was