j. Soc. Cosmet. Chem., 30, 279-295 (September/October 1979) Mechanism of antiperspirant action of aluminum salts ERHARD H6LZLE Department of Dermatology, University of Munich, Munich, IVest Germany and ALBERT M. KLIGMAN Department of Dermatology, Duhring Laboratories, University of Pennsylvania School of Medicine, Philadelphia, PA 19104. Synopsis Complete anhidrosis was induced by occlusive application of aluminmn chloride hexahydrate for 3 and 24 hr. Immediately after exposure, an amorphous aluminum-containing cast was demonstrated in the sweat ducts by the fluorescent morin stain. The cast was PAS-positive, diastase-resistant and extended throughout the acrosyringium down to the upper and mid-dermis. Histologically, the luminal cells of the acrosyringium were damaged and soon sloughed, fusing with the cast. There were no inflammatory changes. The physically occluded ducts did not sweat for two to three weeks until the normal process of cell renewal resulted in replacement of the acrosyringium. ANHIDROSIS induced by ALUMINUM SALTS has unique characteristics which distinguish it from all other procedures which interfere with sweat delivery. INTRODUCTION Of a great variety of metallic antiperspirant salts, aluminum chloride, introduced by Stillians in 1916 (1), is still the most effective. More than 60 years later it is startling that there is virtually no agreement regarding the way in which aluminum chloride, or a host of other metallic salts, brings about inhibition of eccrine sweating. This is not because of lack of interest. Fiedler's sturdy review (2) of the antiperspirant literature lists 411 references! Writings on the mode of action provide a vivid panorama of disparate observations and contrary conclusions. The field is a battleground of contentions and confusions. Despite intensive study by sapient scientists, the prevail- ing ideas are almost wholly speculative. Investigators have sweated mightily, but the problem is as slippery as ever the sweat apparatus refuses to yield its secret. The most ancient theory ascribes antiperspirant activity to astringency, that is, simple shrinkage of the pore. It is subjectively appealing to imagine that the low acidity of aluminum chloride (pH = 2.0) can shut off the flow of s•a eat by contracting the orifice. However support for this idea rests on pharmacologic folklore, not scientific observations. Attempts have been made to salvage the astringency theory by showing that acidic aluminum salts can precipitate proteins the latter could obstruct the flow of sweat. However, a number of chemicals which form strong flocculants with proteins, for example, tannic acid and sulfosalicylic acid, have minimal antiperspirant activity. 279

280 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Besides, aluminum sulfate is a poor antiperspirant even though its protein-precipitating ability is equivalent to aluminum chloride (3). The first experimental effort to explain the anhidrotic effect of aluminum salts was made by Sulzberger, Zack & Herrmann (4). They applied a cream containing 20% aluminum sulfate to the axilla of eight subjects at t2-hr intervals for three applications. Histologic examination showed an intense periductal inflammatory reaction in the dermis, primarily consisting of neutrophils with a variable mixture of lymphocytes. These cell later appeared in the lumen in large numbers. Concomitantly, degenerative alterations in the aprocrine glands were noted. These workers suggest that neutrophils are attracted to the ducts by chemotactic activity of the aluminum salt. These observations remain unverified and are all the more remarkable since aluminum sulfate is a feeble antiperspirant. Another version was soon proffered by Sulzberger and his co-workers (5). They postulated that the sweat duct possessed an electrophysiologic gradient which was most negatively charged toward the surface. Hence positively charged electrolytes would nullify the gradient and inhibit sweating, while electro-negative ones would be enhancing. This quixotic notion has so many inconsistencies that experimenters by and large have not stirred themselves to a refutation. Then came the classic experiments of Shelley & Horvath (6,7). They produced millaria in a variety of ways, viz., ultraviolet radiation, heat freezing, iontophoresis and application of aluminum chloride. Their view was that the histopathologic picture was the same regardless of the way in which miliaria had been provoked, namely, hyper- and parakeratotic plugs within the sweat duct orifices. We have recently reviewed in detail the history of the plug theory (8). Suffice it to say here that the idea is no longer tenable. The horny plugs are a response to injury and are part of the repair process. They follow miliaria but are not its cause. Plugs have not been observed in the stage of anhidrosis which precedes miliaria. Papa & Kligman (9) noted that many of the photographs purporting to show plugs merely demonstrated the normal architecture of the sweat pore. The terminal intra-corneal spirals of the duct wind through a "beaker" of horny material, which indeed resembles a plug but is perfectly normal. Still another theory holds that aluminum chloride affects the secretory portion of the sweat gland, preventing the formation of sweat rather than impeding its delivery to the surface. Shelley (t0) has disproved this possibility by showing that glycogen disappears from the secretory coil beneath the miliarial lesion, a reliable sign of active sweat secretion. Besides, miliaria cannot occur except by the extravasation of sweat into the peri-ductal tissue. Brunet al. (tt) postulated that aluminum salts produced epidermal thickening (acanthosis) and that this made it difficult for sweat to escape. However, Hunziker et al. (12) showed with a variety of organic acids, aldehydes and alcohols that there is no correlation between antiperspirancy and acanthogenicity. Without doubt, the strangest theory of all was proposed by Papa & Kligman (9). Because they could not find a plug, could not relieve the anhidrosis by stripping but could obtain methylene blue pore patterns, indicating at least partial ductal patency, they contrived the leaky-hose hypothesis. According to this triumph of deductive reasoning, an obstruction did not exist. Instead, sweat poured through a duct made