PSEUDOMONAS INOCULATION ON SKIN 27 superhydrated skin. Apparently toxins are conveyed into the skin through follicular shunts. As in experimental Candida albicans infections, the process is not dominantly follicular and is provoked by transepidermal diffusion of toxic materials, a process which we have dubbed a "biological type of contact dermatitis." Our failure to provoke toxic reactions in the absence of a prior injury to the horny layer barrier seems to parallel the clinical scenario. Whether it be a deep burn, recurrent otitis externa, interdigital athlete's foot, or a scratched cornea, skin damage exists before an overgrowth of Pseudomonas can harm the skin. Under these special circumstances, tissue injury may be quite severe if high densities of Pseudomonas develop. No significant differences existed among the 16 strains of Pseudomonas with respect to inflammatory reactions. Pseudomonas strains produce a wide variety of extracellular toxins including phospholipase, proteolytic enzymes with trypsin-like activity, entero- toxins, exotoxins, and Pseudomonas slime. The proteolytic enzymes are believed to be responsible for the hemorrhage and necrotic changes which occur in experimental animal infections and probably in corneal destruction (17). There are two or possibly three proteases which can be separated by physical means. They are capable of liquifying gelatin and of degrading elastin, fibrin, and collagen (17). Upon injection into animal skin Pseudomonas protease will also produce indurated necrotic abscesses (18). Despite this formidable array of proteolytic enzymes, Pseudo- toohas appears incapable of damaging intact, normal skin. This is somewhat surprising, since 48-hr patches of 0.5% trypsin and subtilisin will provoke an inflammatory response on normal skin. Apparently the surface concentration of Pseudomonas proteases are too low. When directly injected into skin, tissue necrosis is to be expected whether or not growth occurs. Doses less than 105 organisms did not produce cutaneous damage even on scarified skin. An important aspect of the relatively high inoculum required to damage skin is the high moisture requirement of Pseudomonas. It slowly dies on a dry surface. Even 6 hr after inoculating scarified, semi-occluded skin the number of P. aeruginosa has rapidly declined, and by 12 hr viability was lost altogether. Our findings raise serious questions regarding the "epidemics" of skin infections in groups of people exposed to swimming pools and whirlpool baths contaminated with P. aeruginosa. The lesions observed were often similar to our experimental papulo- pustular eruptions. However, it may be significant that the folliculitis so prominent in the heated pool episodes was completely absent in our studies. The evidence would appear to us to be insufficient to incriminate Pseudomonas as the cause of the swimming pool rashes. Although we achieved high surface densities with the very same organisms isolated in those epidemics, no lesions were induced on normal skin unless it was damaged by either scarification or superhydration. Since our data indicate that a relatively high inoculum (1 x 105) is required to induce a lesion on scarified skin and 1 x 106 is required in superhydrated skin over a period of days, it is difficult to imagine how a pool could become so densely populated. Of the three epidemic reports, the level of contamination was estimated in two studies. In one study, the level was 3 colonies per 100 ml while in the other study the coliform count was 5 per ml, levels far below what we found necessary to induce lesions (12, 13). Contamination of minor cuts and abrasions or insect bites, particularly if excoriated, could drive Pseudomonas deeper into the skin. If then the site could remain relatively moist, an

28 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS eruption could conceivably develop with a low inoculum. In view of the ubiquity of Pseudomonas and the frequency of cuts and abrasions, Pseudomonas infections are a decided rarity even in tropical areas. Finally the question arises of the risk entailed in using a cosmetic cream which has inadvertently become contaminated with Pseudomonas. There is no doubt that in the area of eye products, mascara, or cleaning fluids for contact lenses, for example, the consequences are so serious that no level of contamination can be tolerated, even though the attack rate is probably very low. Taking our results with multiple strains of Pseudomonas and applying a 3 order of magnitude safety factor, we would propose the following guidelines. For an agent designed to be applied to an abrasion, cut, or wound, the risk is considerably less than for an eye product but may still be appreciable. Cuts and abrasions are routinely bandaged, and even band-aids may be occlusive enough to promote bacterial growth. While we would agree that zero levels are ideal, a low-level contamination (1 x 102 or less) would be unlikely to produce injury, since 1 x 105 organisms had no effect on scarified skin. With products that are usually applied to normal skin, Pseudomonas contamination appears not to pose a serious problem. Even at very high concentrations (1 x 106 per sq cm) P. aeruginosa is not hazardous to intact skin. For this type of product we would envision no danger if the contamination level did not exceed 1 x 103 organisms. REFERENCES (10) (11) (12) (13) (14) (15.) (1) R. E. Buchanan & N. E. Gibbons, eds., "Bergey's Manual of Determinative Bacteriology," 8th ed., Williams & Wilkins: Baltimore, MD, 1975 pp 217-219, (2) S. Tanenbaum, Significance of pseudomonads in cosmetic products, Amer. Perfum. Cosmet., 86, 33, 1971. (3) H. B. David, T. A. McMeekin and C.J. Thomas, Spoilage association of chicken skin, AppL and Environ. Micro., 37 (3), 399, 1979. (4) L. Goldman and H. Fox, Greenish pigmentfition of nail plates from bacillus pyocyaneus infections, Arch. Derm. & Syph., 49, 136, 1944. (5) M. Saltzman, Otitis externa clinical aspects and bacteriologic studies, Clin. Med., 70, 559, 1963. (6) R. A. Amonette and E. W. Rosenberg, Infection of the toe webs by gram negative bacteria, Arch. Derm., 107, 71, 1973. (7) D. Taplin, D. C.J. Bennett and P.M. Mertz, Foot lesions associated with Pseudomonas cepacia, Lancet, 568, 1971. (8) R. W. Heftnet and G. F. Smith, Ecthyma gangrenosum in Pseudomonas septicemia, Amer, J. His. Child, 99, 524, 1960. (9) F. N. Marzulli, J. R. Evans and P. D. Yoder, Induced Pseudomonas keratitis as related to cosmetics, J. Soc. Cosmet. Chem., 23, 89, 1972. W. C. Noble and J. A. Savin, Steroid cream contaminated with Pseudomonas aeruginosa, Lancet, 347, 1966. W.J. McCausland and P. Cox, Pseudomonas infection traced to motel whirlpool, J. Envir. Health, 37, 455, 1975. J. Washburn,J. A.Jacobson, E. Marston and B. Thorsen, Pseudomonas aeruginosa rash associated with a whirlpool, J. Amer. Med. Assoc., 235, 2205, 1976. W. F. Sausker, J. L. Aeling, J. E. Fitzpatrick and F. Judson, Pseudomonas folliculitis acquired from a health spa whirlpool,J. Amer. Med. Assoc., 235, 2205, 1976. M. T. Hojyo-Tomoka, R. R. Marpies and A.M. Kligman, Pseudomonas infection in superhydrated skin, Arch. Derm., 107,723, 1973. G. Singh, Pseudomonas infections of skin: an experimental study, Inter. J. Derre., 13 (2), 90, 1974.