.-\cta Derm Venereol (Swckh) 1997; 77: l-3. Duct Disruption, a New Explanation of Miliaria

SAM SHUSTER

Department of Dennatology. Uni1•ersity of Ne1vmstle, Ne 1~·castle upon Tyne, UK

... From argument and a few personal observations, the hypothesis had noticed with natural sunlight in tbe tropics, miliaria bas crystallised that the millaria commonly occurring in unaccli­ developed at 6 days at the si te of UV exposure, but not in the matised Caucasians visitlog bot elimates is eaused by exposure unexposed, noro~al skin. To confirm the uosurprising guess to ultraviolet irradiation, by an effect on the cells of the upper that the lesions had formed in the stratum corneum. thatlayer ~pi dermis , whiob eventually allows a split to develop between was removed as completely as possible with cellophane tape. them and the new stratum corneum that grows up beneath, into on a separate occasion, in hal f of the area exposed to the same "bich sweat from the disrupted ducts can collect as microcysts. dose of UVR 7 days earlier. As before, miliaria occurred with This debiscence is the probable explanation of sunburn peeling thermally induced sweating in the UV-exposed skin. but not and photo-onycholysis. h is concluded that duct disruption, not where the stratum corneum had been removed. From these blockage or dysfuuction, is the immediate cause of the miliarias. observations, T can only conclude that exposure to UVR (Accepted May 31, 1996.) induces miliaria crystallina by an effect which Jeads to a cbange in the stratum corneum 5- 6 days later. Acta Derm Venereol (Stockh) 1997; 77: 1- 3. What might this change be? The Jong-accepted view (2) thal S. Shuster. Department of Dem1atology, University of the primary change in miliaria is ·'k.eratotic plugging of the i'-/ewcastle, NEJ 4LP Newcastle upon Tyne, UK. sweat duct orifices" (3) can be rejected on general grounds because miliaria does not occur with superficial sweat duct blockage (6-8). It can likewise be rejected as the explanation Miliaria, a common problem for Caucasians in hot climates, of UV-provoked 1niliaria: VVR could not evoke the manufac­ is generally attributed to sweat duct blockage (l). I have never ture of plugs from the dead keratinocytes of the surface, and fo und this explanation convincing, because l consider the a plug of keratinocytes newly formed in the epidermis would histologica.l evidence ofblockage (2-5) to be poor and because take longer lhan 5- 6 days to grow out to the surface (l l). miliaria cannot be produced by stimulation of sweat glands Furthermore, in the former case, the occurrence of miliaria wbich have ducts blocked by disease (6) or experiment (7, 8); wouJd be synchronous with exposure or, at any rate, more indeed, during such blockage, the coil continnes to secrete rapidly than 5- 6 days, as it would likewise if blockage were sweat and the duct is able to absorb allthat is secreted (7, 8). eaused by material in the lumen of the terminal duct produced I therefore concluded that the immediate cause of miliaria was by damage more deeply in the skin. An explanation other not duct blockage buL a failure of the duct to absorb all the than duct blockage is required. 5'\ eat secreted by the coil ( 6, 9). T hese l wo opposing hypo­ Although the lesions of UV-induced miliaria are discrete, I theses have had as little efrect on one another as on the found them to be easily spread by gentie pressure, and they dl.sorder they attempted to explain, and thjs continued mechan­ join up with one another in the stratum comeum. Tbey can l'mt" sterility. together with a few personal observations, com­ t hen be peeled off as a sheet, revealing a new layer of normal­ p!Ued a reconsideration of the problem which produced a new looking stratum corneum beueath. in which the Lniliaria do b}pothesis with promise of a better outcome. not occur. Thus, the sweat, which initially forms discrete vesicles, is collecting in a potential plane of separation in the i!Dle () ' ev irradiation stratum corneum induced by UVR damage, and the ease with l usually develop millaria crystallina in tropicaJ climates. The which the vesicles can be spread into this plane by pressure .e-.,oru. appear rapidly when I sweat heavily and are easily reveals this potential split The commonly observed. but stopped by taking a cold shower. But l have noticed that the equally commonly unremarked upon, un((ormity of the sheet lesions oever occur before the end of the first week after of ·'sunburn peeliug·' fits coolfortably wi th this interpretation. arrival, and that they have a clear relationship to sun exposure, The thickoess of one such sheet, measured by micrometer, was the millaria being confined to the areas of mild sunburn (T 50 ~tm , w h ich is more tbao the thickness of the normal stratum rarely burn severely), never occurring in unexposed skin. As I corneum, suggesting that the primary action of UVR is on the have made this observation many timeson royself and found, epidermis, the first living cells to be encouotered by UVR by questioniog, the identical phenomenon in others, l can only penetrating the skin. Histological examination of several sheets conclude tbat exposure to ultravioler irrad:iation ( UVR) is the of sunburo peeling and the outer walls of a group of fresh common cause of miliaria crystallina. The productioo of miliarial vesicles, produced 7 days after UV exposure of my lesions by UVR has Iong been known, and it is therefore flexor forearm (Fig. l). conlirmed this si te of effect, the outer surprising that the prime ro le of UVR has been Jargely ignored, wall of the miliarial microcysts (and the sun-burn peeling) or indeed refuted ( 1-5). consisting of the old stratum corneum Jying on 2-3 layers of The opportunity to examine the relationship of rniliaria to damaged epidermal cells which bad separated from the new UVR came in the course of an unrelated study (lO) in which stratum cornetm1 beneath. a 4 x 6 cm area of my forearm was exposed to 5 joule(sq.cm of 311 nm UVR. Wheo heavy thermogenic sweating was Mechanism induced 3 and 6 days later, by playing squash with marginally As discussed in rejecting terminal duct blockage in the stratum better opponents, no miliaria occurred at 3 days; but, just as corneum, tbe dela y in onset of lesions excludes the possibility

<\;) 1997 Scandinavian University Press. lSSN 0001 -5555 Acta Derm Venereol (Stockh ) 77 2 S. Shuster

Mlllartatsunbum peeling

Day 1 Day 5-7

Fig. 2. Diagramatic explanarion of UV-provoked miliaria crystallina by duct disruption; l ) UVR damage to cells of upper epidermal cells; 2) The UVR-damaged epidermal cells are extruded, sandwiehed between the old and new stratum corneum cells: 3) The potential for dehiscence of the UVR-damagcd cpidcrmal cells from the new stra[Um comeum becomes apparent5- 7 days later wllen, during heavy sweat­ ing. swea.t escapes from the disrupted ducts imo the plane of dehi­ scence, forming miliaria. The plane of dchiscencc may also reveal itself as sunburn peeliog.

myself, chose a different conclusion ( 2- 6 ), despite the contrary evidence which was equally available then as now. In particu­ Fig. l. Outer wall of miliarial microcyst, consisting of the old stratum lar, Lowenthal ( 5 l (although less concerned with crystalline corneum and a layer of 2-3 damaged epidermal cells that bas split than other miliarias, and despile a curious attribution of the away from the new stratum corneum (not shown). whicb forms the disorder to saline tonicity) showed that the keratin ·'plugs" base of the microcyst. which had Iong been associated with duct blockage were a late event ( 5 ); yet despit e this they were not generally dismissed that UVR has a direcL dehiscent effect on the epidermis. Thus. as secondary phenomenon. whatever the initial effect of UVR on the cells of the upper epidermis, it is not detected as miliaria until the affected New llypothesis epidermal cells are moved out by the proliferating cells beneath The new hypothesis has obvious practical and theoretical them, and a plane of separation develops in the stratum implications. As ever, the former require less supporlive verbal corneum. between the UV-damaged epidermal cells and the dressing. Tn brief. l have found thai gradually increasing UV cells of the new stratum comeum beneath them. Clearly, such exposure and use of sun screens prevents mi liaria. as weil as a plane of separation, which may later progress to a sheet of being more enjoyable than avoiding the sun. Could duct sunburn peeling, would destroy the continuity of the sweat disruption, lhe essence of the new hypothesis, accoun t for the ducts, the complex spiral structure of which is maintained up Iong recognised, bu t unexplained, occunence of miliaria crys­ to the surface of the stratum corneum (12); sweat would then tallina alter iontophoresis ( 13 )? Certainly, as I have now leave the disrupted ducts and collect as miliaria in the plane belatedly recognised, it occurs with a similar latency to that of dehiscence, as shown diagrammatically in Fig. 2. Thus it of UVR-induced lesions. However, since the peeling after appears that the immediate cause of miliaria crystallina is du et iontophoresis is variable and irregular, if the action of the disruption. and this duct disruption is primary, not secondary iontophoretic current is on the cells of the upper epidermis as previously believed; duct blockage and inadequate absorp­ and sweat ducts. leading to disruption of their continuity as tion, the elderly mechanistic predecessors of this hypothesis, new cells of the stratum corneum (and sweat ducts) grow out, can now be Jaid to rest. Although the direct evidence for this the effect must be more focaJ than produced by UV exposure. conclusion comes from self-observations, these were consist­ Duc1 disruption would likewise explain the miliaria which ently reproducible on a number of occasions, and therefore follows the application of ·'irritants·· and diseases which show. at least, why l develop miliaria crystallina in the tropics darnage the epidermis (2, 4 ). leading ultimately to a layered (maceration millaria is discussed later). But since, personal desquamation. The unifying hypothesis of ducr disruption into oddities not withstan ding, 1 am unlikely to dilfer in the relevant a plane of potential dehiscence also explains the miliaria respects from other members of my species, this personal crystallina which follows superficial hydration. Of course, it is evidence can reasonably be generalised. More importantly, the weil known that hydration of tbe skin surface occludes the same conc!usion can be reached on theoretical grounds afld a sweat duct orifices ( 14 ), bu t for reasons already discussed this reinterpretation of published evidence: tbus it is now apparent would not produce miliaria so Iong as the ducts remain intact. that my new hypothesis resolves the conflict dormant, or Thus in the roitiaria crystallina which follows surface occlusion avoided, in previous studies (2-5), and also explains the (therapeutically, or in bot, humid climates), the eausal duct clinical features of crystalline rniliaria. It is interesting, and disruption is Iikely to be due to dissolution by maceration not entirely explicable, why previous investigators, including damage to the stratum comeum ( 15) (rather than secondary

A m1 Dem 1 r·enert?ol !Stock/1! 77 Duct disruption. a m.>ll' explanarion of mifiaria 3 to changes in the epidermis beneath), allowing sweat from the and miliaria crystallina by variou~ kmdl> of injury. J lnvest disrupted ducts to collect in the space opened by maceration. O.:nnatol 1950: 14: 9- 21. Finally. although du er disruption a ppears universall y applicable 4. Dobson RL. Lobitz WC. The pathogenesis of miliaria. Arch to the various forms ofmiliaria crystallina. it should be equally DerrnaloJ 1957: 75: 653-666. 5. Lowenthal LJA. The pathogencsis of maliaria, the role of sodium applicable to other miJiarias, except that a zoned dehiscence chloride. Arch Demmol 1961 ; 84: 52 67. would not be a prerequisite. l-l owevcr, and moving out of the 6. Johnson C. Sbuster s. Eccrinc sweating in psoriasis. Br J Dermatol sweat ducts, the zone of dehiscence does provide a novel t-969; 8 L: 119 124. explanation of drug-induced photo-onycbolysis: since the nail 7. Shuster S. Graded sweat duct occlusion: u tecbnique for studying is now known to be formed continuously a Iong its Jeugth ( 16 ). sweat gland function. Clin Sci 19631 25: 89-95. and not jul>l in the matrix as previously believed, dehiscence 8. Johnl>On C. Shuster S. The us.: of partia l sweat duct occlusion in following UVR damage to epiderrnal cells of the nail plate the elucidation ~f sweat duct funcuon in health and disease. J Soc would separate' naiJ from its active site of growlh in the bed. Cosmel Chem 1973: 24: 15-29. Photo-onycholysis is the simple analogue of sunburn peeling. 9 Shuster S. Dermatology in Intemal Medicine. Oxford. 1978: 110 III. 10. Fnrr P. Shuster S. Sunbum nod neuropeptide depletion. Br J Dcrmatol 1995; 133: 1012. ACK NOWLEDGEMENTs Il. Porter D. Shuster S. A new method for measuring replaccment of epidermis and stratum corneum. J lovest Dermatol 1967: -19: Sam Shuster holds a Leverhulmc Emeritus Fellowsbip and is grateful 251 255. to Eli~bcth Sveland of the Department of Pathology and Peter Farr 12. Johnson C, Dawber R, Shuster S. i\ppearances of rbe eccrinc of the Department of Dermatology. University of Newcastle upon sweat duct by scanning electron microscopy. Br J Derrnatal 1970: Tync. 83: 655 660. 13. Shellcy WB. Howatb PN, Weidman FD. Pitsbury DM. Productian of sweat retention anhidrosis and vesicles by means of iontophor­ REFERENCES esis. J lovest Dermatol 1948: Il: 275- 281. 14 Surkany l, Shuster S. Stammers M. O<:clusioo of the sweat pore l. Champion RH. Burton JL, Ebling FJG, eds. Textbook ofdermato­ by hydration. Br J Dermatol 1965: 77: 101- 104. logy. 5th edn. Oxford: Blackwell, 1995: 1758 1759. 15. Hombrick GW, Blank H. The microanalomy of miliaria crys­ 2. O'Brien JP. Study of miliaria rubra. topical anhidrosis and tnllina. J lovest Dennatal 1956; 26: 327 336. anhidrotic asthania. Br J Dermatol 1947: 59: 125. 16. Johnson M. Sbuster S. Contint1ous formation of nail along the 3. Shcllcy WB, Horrath PN. Produclion ofswcat retention anhidrosis bed. Br J Dennatol 1993: 128: 277- 280.

At'/O Derm Venereol ( Stockh ) 77