DOI:10.1111/j.1600-0625.2008.0712.x www.blackwellpublishing.com/EXD Controversies in Experimental Dermatology

Section Editor: Ralf Paus, Lu¨ beck

What causes hidradenitis suppurativa?

H. Kurzen, I. Kurokawa, G. B. E. Jemec, L. Emtestam, K. Sellheyer, E. J. Giamarellos-Bourboulis, I. Nagy, F. G. Bechara, K. Sartorius, J. Lapins, D. Krahl, P. Altmeyer, J. Revuz and C. C. Zouboulis

Abstract: Hidradenitis suppurativa (HS) – a rather common, very traces of terminological bickering remain visible, it does the HS chronic and debilitating inflammatory skin appendage disorder experts in our virtual debate room credit that they engage in a with a notoriously underestimated burden of disease – has long constructive and comprehensive dissection of potential been a playground for the high priests of nomenclature: Ask a pathogenesis pathways that may culminate in the clinical picture bunch of eminent dermatologists and skin pathologists to publicly we know under the competing terms HS or acne inversa. These share their thoughts on what causes HS, and they will soon get experts sketch more often complementary than mutually exclusive entrenched in a heated debate on whether this historical term is pathogenesis scenarios, and the outlines of a conceivable a despicable misnomer. Fortunately, the recently founded consensus on the many open pathobiology questions begin to Hidradenitis Suppurativa Foundation (HSF; http://www. emerge in these CONTROVERSIES. Hopefully, this heralds a hs-foundation.org), to which EXP DERMATOL serves as home welcome new tradition: to get to the molecular heart of HS journal, has broken with this unproductive tradition and has pathogenesis, which can only be achieved by a renaissance of solid encouraged publication of the current CONTROVERSIES feature. basic HS research, as the key to developing more effective HS This is exclusively devoted to discussing the pathobiology of this therapy. chronic neutrophilic folliculitis of unknown origin. Although

Please cite this paper as: What causes hidradenitis suppurativa?. Experimental Dermatology 2008.

Prelude as corticosteroids, immunosuppressive drugs and ⁄ or anti- TNF agents, may improve the disease Apocrine or not, that is the question… YET In the context of HS, it has become foolhardy to speak of One of the most useful drug regimens in cases of active apocrine sweat glands ever since pathologists have demon- inflammatory lesions is a combination of two antibiotics: strated that the primary histological event was in the follic- rifadin and clindamycin (1). ular duct, like in acne. It was then simple, if not to say simplistic, for many dermatologists to forget any major dif- Paradoxon 1 ferences between acne and HS, and to rename the disease The topography of involvement may be explained by two – ‘acne inversa’, thus replacing one possible misnomer (HS) not mutually exclusive – hypotheses: (a) the distribution of with another – and leading investigators to a dead end, apocrine glands; and (b) shearing forces, which originate in and practitioners to use ineffective treatments. large skin folds, especially of overweight patients. Any hypothesis about HS must take two paradoxons into Obesity and overweight are frequent in HS and could be account: a risk factor (2). However, HS is not infrequent in inter- 1 Hidradenitis suppurativa lesions have a very specific mammary folds and on the buttocks, where such shearing topography which is a copy of the anatomical distribution forces are absent. Moreover, individuals with low or nor- of apocrine sweat glands: axillae and groin as the main mal body mass index also do develop HS. Thus, while areas; breast, perineum and buttocks as accessory regions. obesity and overweight are strongly associated with severity YET in HS, they are insufficient to explain the specific topogra- Apocrine glands are not the primary target of the patholog- phy of the disease. ical process. So, what about apocrine sweat glands? The rejection of 2 Hidradenitis suppurativa is not primarily an infectious apocrine glands as a main factor in the pathogenesis of HS disease, and yet drugs that are conducive to infection, such originates in the clear demonstration that they are spared

ª 2008 The Authors Journal compilation ª 2008 Blackwell Munksgaard, Experimental Dermatology 1 Kurzen et al. by the initial inflammatory and destructive process. The New ones, psoriasin (7), are under consideration. Some of primary event is a follicular hyperkeratosis with plugging these peptides are produced by mature keratinocytes and dilatation of the hair follicle ensuing in inflammation, spontaneously or after stimulation, some are produced by abscess and sinus tract formation. Apocrine involvement eccrine sweat glands, and some are prominently subject to appears as a secondary phenomenon resulting from the modulation by antibiotics (7–9). diffusion of the granulomatous inflammation in deep However, until now, nothing is known about anti-micro- structures of the skin. bial peptides that are specifically produced in apocrine The mechanism by which follicular plugging occurs in sweat glands. This is where we need to search. acne is not known; various candidates are hypersecretion of sebum, proliferation of P. Acnes favoured by an alteration Conclusion of innate immunity and ⁄ or of inflammatory reactions, and Hidradenitis suppurativa is as multifactorial as any chronic possibly several others. disease and probably heterogeneous. Here, I have only con- Here, the specific anatomical relationship of the apocrine sidered the role that apocrine glands may have, a role that gland with the follicular canal has to be taken into account: might well be central – no matter, how much this gland In contrast to eccrine glands, whose ducts open onto the has fallen out of fashion in HS research. The role of other skin surface, apocrine glands empty their content into the factors – e.g. hair follicle anatomical ⁄ structural abnormali- follicular canal, just above the sebaceous gland duct. In HS, ties (highly probable at least in a subset of HS patients), hyperseborrhoea is definitely absent but the other factors the role of obesity, the role of cigarette smoking – all may be at work: deserve careful exploration. Any pathogenesis scenario, An abnormal secretion – excess or absence – of a sub- however, that completely discards apocrine glands and their stance that is present in apocrine gland secretion under specific distribution as key elements in the development of physiological conditions may therefore, after all, be the HS may soon turn out to have to be discarded itself! triggering factor of HS! Its morphologically recognizable effect could be in the acro infundibulum of the follicle, Jean Revuz with the responsible gland disguising itself as an innocent Department of Dermatology, Henri Mondor Hospital, bystander upon histology – a perfectly masked ‘criminal’. 94000 Creteil, France; E-mail: [email protected] Paradoxon 2 Hidradenitis suppurativa is a disease in which numerous References bacteria are present and active (3), and in which various 1 Mendonc¸a C O, Griffiths C E M. Br J Dermatol 2006: 154: 977–978. 2 Naldi L. Epidemiology. In: Jemec G, Revuz J, Leyden J, eds. Hidrade- antibiotic regimens have definitely improved the condition nitis suppurativa, Chap. 8. Springer Verlag, 2006: 58–64. in patients with severe inflammatory involvement (2). Sur- 3 Faye O, Poli F, Gabison G, Pouget F, Wolkenstein P, Revuz J. Associa- prisingly, numerous pro-infectious drugs have also been tion rifampicine ⁄ clindamycine dans l’hidrade´ nite suppure´ e. Ann Der- used with good results: corticosteroids, immunosuppressive matol Venereol 2005: 132: 68–69. drugs, anti-tumour necrosis factor (4,5). The coexistence of 4 Oprica C, Nord C E. Bacteriology of hidradenitis suppurative. In: Jemec G, Revuz J, Leyden J, eds. Hidradenitis suppurativa, Chap. 11. these two seemingly contradictory phenomena calls for a Berlin: Springer Verlag, 2006: 86–94. closer look at the properties of anti bacterial molecules 5 Nybaek H, Jemec G B E. Immunosuppressive therapy. In: Jemec G, which play a key role in innate and acquired immunity: the Revuz J, Leyden J, eds. Hidradenitis Suppurativa, Chap. 18. Berlin: so-called anti-microbial peptides (6). Springer Verlag, 2006: 136–140. They are known to exert both pro- and anti-inflamma- 6 Jacob S E, Kerdel F A. Biologic for hidradenitis suppurativa (Verneuil’s disease in the Era of biologics). In: Jemec G, Revuz J, Leyden J, eds. tory functions; they alarm and activate the adaptive Hidradenitis suppurativa, Chap. 20. Berlin: Springer Verlag, 2006: immune system and the keratinocytes that produce them. 145–149. They induce keratinocytes migration, proliferation – a role 7 Schauber J, Gallo R L. Expanding the roles of antimicrobial peptides in follicular occlusion? They are part of a complex network in skin: alarming and arming keratinocytes. J Invest Dermatol 2007: of cytokine and chemokine production. The absence or 127: 510–512. 8 Gla¨ ser R et al. Antimicrobial psoriasin (S100A7) protects human skin abnormality of one of these anti-microbial peptides would from Escherichia coli infection. Nat Immunol 2005: 6: 1–57. be a good candidate for explaining the infectious and 9 Tokura Y et al. Epidermal chemokines and modulation by antihista- inflammatory features of HS. Cathelicidins and defensins mines, antibiotics and antifungals. Exp Dermatol 2007: November 22 are the main representatives of this family identified today. [Epub ahead of print].

ª 2008 The Authors 2 Journal compilation ª 2008 Blackwell Munksgaard, Experimental Dermatology Controversies in Experimental Dermatology

Viewpoint 1 The scientific mind should view things without prejudice. may cause contractures, and far-reaching sinus tracts, the Hidradenitis suppurativa (HS) is a common dermatosis, latter often growing in a pseudo-invasive manner into the which significantly affects the lives of patients and which is surrounding tissue. notoriously difficult to treat. It is a well-defined clinical entity, consisting of recurrent crops of inflammatory lesions Aetiology in inverse areas (1). It looks infectious, but it is not. The wide range of clinical HS presentations, and of occa- The lesions are initially transient, but gradually become sionally successful HS treatments, invites the speculation intransigent and associated with significant scarring. Histo- that HS aetiology does not arise from a single external logically, it is a disease of the hair follicle; associated with factor, but rather represents a reaction pattern within the lympho-histiocytic inflammation, granulomatous reactions, afflicted patient. This concept is supported by genetic sinus tracts and scarring (2–5). It looks like acne – but it observations (22,23). is not. The effectiveness of antimicrobial drugs such as clinda- Treatment is often disappointing, and has significant mycin suggest that bacteria may occasionally be responsible negative impact on the patient’s quality of life (6–8). for starting the process (9,11). Similarly, the use of anti- Anti-inflammatory, immunosuppressive, antiandrogenic, inflammatory treatment such as corticosteroids or TNF- antibiotic and surgical treatments have been described as blockers, suggest a role for the immune system (24), while effective, although few randomized, controlled trials exist tobacco or physical trauma (described in convincing indi- that convincingly support these claims (1,9–11). vidual cases or epidemiological studies as trigger or aggra- So, what then is HS? vation factors) suggest that these factors may be involved (25). Most likely, however, they are only temporarily Disease hallmarks involved (or of secondary importance) as there is no evi- One of the most obvious hallmarks of the disease is the dence to suggest that their removal invariably cures HS. restriction to the skin areas affected. The disease is essen- Hypothetically, identical HS lesions can therefore be tially limited to inverse areas, although aberrant lesions produced by bacterial, immunological, chemical or physical may occur. This lead to the original proposition that factors affecting the predisposed hair follicle. apocrine glands are involved in the disease. It is good to suggest theories – but these can be rejected later. Pathophysiology The inverse areas are indeed rich in apocrine glands – There is consensus that the pathophysiology of the disease although the overlap between HS lesions and apocrine-rich involves hair follicle rupture and subsequent inflammation, regions of the integument is not good within an affected sinus tract formation and scarring. Most likely, folliculitis skin territory (3). The inverse areas are, however, also char- is one of the most common pathological events in the acterized by skin–skin contact and subsequent shear forces human skin – and yet this leads to HS only in a minority affecting the skin; increased moisture and bacterial flora, all of individuals. Why? of which may play a role. One possibility is the special biomechanical conditions in Within the affected regions, hair follicles are invariably the affected follicles and concave skin regions. It has been involved (2). The deep part of the follicle appears to be suggested that the biomechanics of the predominant kera- involved, rather than its superficial compartments, as seen tin found in the lesions is suboptimal, leading to repeated with acne affecting convex skin surfaces (12). Clinically, ruptures and development of lesions. The process may be closed comedones are not seen. further supported by the local biomechanical forces affect- Although a range of bacteria may be retrieved from HS ing the concave surface of, e.g. the axillae. It may be specu- lesions, these frequently appear to be sterile (13–15). Of lated that micro-tears of the hair follicle of predisposed the bacteria identified, only a small minority are recognized individuals could be the primary event. Clinical observa- pathogens, and no consistent IgG response to staphylococ- tions support such a mechanism. If unspecific factors cal antigens is found (13). Hormones are unlikely players induce inflammation, the subsequent control of the inflam- in the pathogenesis of HS: increased sebum excretion and matory process may offer an alternative explanation. Either other signs of cutaneous virilization are not seen, although an excess of pro-inflammatory signals, or a lack of inflam- single cases have been published to suggest this (16–19). mation containing signals may lead to the development of Similarly, the use of depilatories and other cosmetics are HS. Indeed, studies have pointed to abnormalities in the not associated with the disease (20,21). Scarring is promi- immunity of HS patients (26,27). nent, but often overlooked in this disease. The fully devel- Finally, the ‘containment’ or wound healing mechanisms oped disease is characterized by significant scarring and of the host may play a role. The wound healing mecha-

ª 2008 The Authors Journal compilation ª 2008 Blackwell Munksgaard, Experimental Dermatology 3 Kurzen et al. nisms responsible for the scarring and sinus tract formation 3 Jemec G B, Hansen U. J Am Acad Dermatol 1996: 34: 994–999. may depend on which cells from the damaged hair follicles 4 Jemec G B, Thomsen B M, Hansen U. APMIS 1997: 105: 378–383. 5 Boer J, Weltevreden E F. Br J Dermatol 1996: 135: 721–725. are recruited. A group of stem cell-like cells have been 6 Jemec G B et al. Clin Exp Dermatol 1996: 21: 419–423. identified in sinus tracts of HS lesions, and may provide 7 von der Werth J M, Jemec G B. Br J Dermatol 2001: 144: 809–813. the dynamic impetus for development of characteristic 8 Wolkenstein P et al. J Am Acad Dermatol 2007: 56: 621–623. lesions (28). A better characterization of the specific wound 9 Clemmensen O J. Int J Dermatol 1983: 22: 325–328. healing-related characteristics of the hair follicles in indi- 10 Mortimer P S et al. Br J Dermatol 1986: 115: 263–268. 11 Jemec G B, Wendelboe P. J Am Acad Dermatol 1998: 39: 971– viduals affected by HS may not only provide better insight 974. into HS pathogenesis, but may also promote our under- 12 Jemec G B, Gniadecka M. Arch Dermatol 1997: 133: 967–970. standing of the clinically challenging problem of fistula 13 Jemec G B et al. Dermatology 1996: 193: 203–206. formation in general, not only in HS patients. 14 Sartorius K et al. Dermatology 2006: 213: 305–312. The scientific mind must be without prejudice, fertile 15 Lapins J et al. Br J Dermatol 1999: 140: 90–95. 16 Jemec G B, Gniadecka M. Dermatology 1997: 194: 325–328. and prepared to reject the ideas it has fostered. It will be 17 Jemec G B. Br J Dermatol 1988: 119: 345–350. interesting to see which of the ideas outlined here will 18 Barth J H et al. Br J Dermatol 1996: 134: 1057–1059. survive. 19 Harrison B J et al. Br J Surg 1988: 75: 972–975. 20 Morgan W P, Leicester G. Arch Dermatol 1982: 118: 101–102. Gregor B.E. Jemec 21 von der Werth J M, Williams H C. J Eur Acad Dermatol Venereol Department of Dermatology, Roskilde Hospital, University 2000: 14: 389–392. of Copenhagen, DK-4000 Roskilde, Denmark; 22 Fitzsimmons J S, Guilbert P R. J Med Genet 1985: 22: 367–373. 23 Gao M et al. J Invest Dermatol 2006: 126: 1302–1306. E-mail: [email protected] 24 Alexis A F, Strober B E. J Cutan Med Surg 2005: 9: 296–302. 25 Konig A et al. Dermatology 1999: 198: 261–264. References 26 Ginder P A et al. J Clin Immunol 1982: 2: 237–241. 1 Jemec G B. J Cutan Med Surg 2003: 7: 47–56. 27 Giamarellos-Bourboulis E J et al. Br J Dermatol 2007: 156: 51–56. 2 Yu C C, Cook M G. Br J Dermatol 1990: 122: 763–769. 28 Gniadecki R, Jemec G B. Exp Dermatol 2004: 13: 361–363.

Viewpoint 2 reflecting the dynamic process of inflammation, prolifera- tion and stratification taking place in HS (7). Hyperkerato- I Fell Into A Burning Ring Of Fire sis of the infundibular follicular epithelium occurs in I Went Down, Down, Down response to androgenic stimulation. And The Flames Went Higher As the fire of HS becomes a bit bigger, the plug becomes And It Burns, Burns, Burns larger and then is invaded by bacteria from the skin com- The Ring Of Fire mensal bacterial population, usually staphylococci. Hyper- June Carter Cash, 1962 keratosis of the follicular infundibulum forms comedo-like After a cigarette is lit with a match, the match is perhaps impactions which occlude the pilosebaceous apparatus. The thrown into the corner of the room. The glowing match is keratincyte–keratin complex is not easily broken down or lying in the corner among dust, dirt, hair, microbes and resolved. This is followed by rupture of the follicular canal, grease. Slowly a fire starts. This tiny fire resembles the and the extrusion of foreign material – e.g. corneocytes, starting point of hidradenitis suppurativa (HS). The central bacteria, sebaceous matter and hairs – into the connective question, then, is: What started the fire…? tissue induces the formation of an inflammatory infiltrate, Recently, it has been shown that that the primary event, which consists initially of granulocytes, followed by mono- the smallest fire of HS, is follicular retention by keratinized nuclear cells and the formation of a foreign body-like gran- stratified squamous epithelium. HS would therefore appear uloma. Epithelial strands develop and produce keratin. to be a disorder of follicular rather than apocrine occlusion Squamous epithelium-lined sinuses, fistulas and secondary (1–6). Regardless of disease duration, follicular occlusion is comedones are typical features (Fig. 1). an early and important feature in the pathogenesis of the The leakage of the follicular content of bacteria and disease. Three types of phenotypes on the basis of the keratin into the perifollicular space causes more inflam- expression of cytokeratins of the epithelial cells of the HS mation and an abscess develops. After emptying or lesions have been suggested (7). Type I epithelium is corni- absorption of the abscess contents the follicular epithelial fying, type II is non-cornifying, type III is non-cornifying cells line the lumen of the abscess and form sinusoids. and strongly inflamed. The three phenotypes are charac- These cells that line the sinuses and fistulae form masses terized as pathologically stratified squamous epithelia of keratin harbouring a bacterial flora that lives in balance

ª 2008 The Authors 4 Journal compilation ª 2008 Blackwell Munksgaard, Experimental Dermatology Controversies in Experimental Dermatology

ing destruction of the epithelial strands in the abscess. Sur- gery is recommended early in the disease process and it has to be as radical as possible to minimize recurrence risk (8– 22). The main features of the recently described carbon dioxide laser technique are in the blood-less surgical field, stepwise vaporization, radical removal of the inflamed tis- sues of HS, including its content of follicular epithelial cells all of which obviate local recurrences, with minimal destruction of healthy tissues (23–26). Friction and small trauma might start the little fire, as well as the use of deodorants and depilatory products and shaving, which leads to minor primary lesions, boils, of HS that normally heal spontaneously in a week. Small balls of keratin are found in the openings of the boils. In the largest fire, i.e. when fistulas are formed, swelling of the keratinocytes occurs. Overweight is unlikely to be causal but may be an exac- erbating factor (27–31). Its importance in HS is probably due to shearing forces of skin and ⁄ or androgen effects. Weight loss may help control the disease. Keratin hydration is increased in sweat gland-rich regions of the body, and this has been shown to favour occlusion. Genetics and endocrine factors may also contribute. Patients frequently report cases among their relatives (32– 35). There appears to be an autosomal dominant inheri- tance with single gene transmission. One study supports the concept of a familial form of HS with autosomal domi- nant inheritance. A genome-wide scan was performed in a Figure 1. Middle, right – a reddish brown ‘beefy’ ball, which is a Hidradenitis suppurativa sinus-surrounded inflammatory active four-generation Chinese family to map the chromosome connective tissue. The fat tissue close to the sinus has turned white due location of the responsible gene (36). the inflammation. The apparent strong influence of sex hormones on HS is emphasized by an association with acne vulgaris comedones and hirsutism, development postpuberty, general decline in with the local immune system in this foreign body-like disease activity seen at the climacteric and improvement milieu. When the bacterial overgrowth in the epithelial often seen during pregnancy (1,30,37–45). However, not all keratin debris favours a vigorous chemotactic response authors agree. Most HS patients have normal androgen with an inflammatory cellular infiltrate consisting of neu- profiles and apocrine glands, and unlike sebaceous glands, trophils, lymphocytes and histiocytes, abscess formation are not androgen sensitive. Nevertheless, there have been develops, the even bigger fire of HS, leading to the anecdotes of symptomatic improvement after the use of destruction of the pilosebaceous unit and eventually the antiandrogens. other adnexal structures in the vicinity. Epithelial strands Finally, coming back to fire and cigarettes, smoking in are generated possibly from ruptured follicular epithelium patients with HS has been reported to be more frequent to form sinus tracts. If keratin-forming epithelial cells are compared with that in controls (46–48). Smoking induces left in the wound after surgery, the process can be set off altered chemotaxis of polymorhic neutrophils which may again from this living cell remnant. This chronic inflam- play a role, similar to that in palmoplantar pustulosis. It matory process causes severe deformation and fibrotic scar seems reasonable that smoking cessation should therefore tissue (Fig. 1) with subsequent functional defects, repre- be strongly encouraged in patients with HS. senting the largest fire of HS. Treatment of HS should aim to remove the causes of Lennart Emtestam, Karin Sartorius, Jan Lapins hyperkeratosis of the infundibular follicular epithelium. Section of Dermatology, Department of Medicine, Thus, the follicular mass of the apoptotic keratinocyte–ker- Karolinska Institutet, Karolinska University Hospital atin complex can be eliminated, and bacterial overgrowth Huddinge, SE 141 86 Stockholm, Sweden; reduced, thereby inhibiting the formation of and promot- E-mail: [email protected]

ª 2008 The Authors Journal compilation ª 2008 Blackwell Munksgaard, Experimental Dermatology 5 Kurzen et al.

25 Fairhurst M V et al. Mayo Clin Proc 1992: 67: 49–58. References 26 Dalrymple J C, Monaghan J M. Br J Surg 1987: 74: 420. 1 Attanoos R L et al. Br J Dermatol 1995: 133: 254–258. 27 Edlich R F et al. J Emerg Med 1986: 4: 369–378. 2 Brunsting H A. Arch Dermatol Syphiol 1939: 39: 108–120. 28 Barth J H et al. Br J Dermatol 1988: 118: 613–619. 3 Chicarilli Z N. Ann Plast Surg 1987: 18: 230–237. 29 Harrison B J et al. Br J Surg 1988: 75: 972–975. 4 Jansen T, Plewig G. Int J Dermatol 1998: 37: 96–100. 30 Barth J H et al. Br J Dermatol 1996: 134: 1057–1059. 5 Jemec G B et al. APMIS 1997: 105: 378–383. 31 Yosipovitch G et al. J Am Acad Dermatol 2007: 56: 901–916; quiz 6 Yu C C, Cook M G. Br J Dermatol 1990: 122: 763–769. 17–20. 7 Kurzen H et al. Br J Dermatol 1999: 141: 231–239. 32 Fitzsimmons J S et al. J Med Genet 1984: 21: 281–285. 8 Thornton J P, Abcarian H. Dis Colon Rectum 1978: 21: 573–577. 33 Fitzsimmons J S, Guilbert P R. J Med Genet 1985: 22: 367–373. 9 Banerjee A K. Br J Surg 1992: 79: 863–866. 34 Fitzsimmons J S et al. Br J Dermatol 1985: 113: 1–8. 10 Barron J. Dis Colon Rectum 1970: 13: 441–443. 35 Von Der Werth J M et al. Br J Dermatol 2000: 142: 947–953. 11 Broadwater J R et al. Am J Surg 1982: 144: 668–670. 36 Gao M et al. J Invest Dermatol 2006: 126: 1302–1306. 12 Brown S C et al. Br J Surg 1986: 73: 978–980. 37 Barth J H, Kealey T. Br J Dermatol 1991: 125: 304–308. 13 Chalfant W Pd, Nance F C. Am Surg 1970: 36: 331–334. 38 Camisa C et al. J Reprod Med 1989: 34: 543–546. 14 Cocke W M Jr. Plast Reconstr Surg 1967: 39: 178–181. 39 Farrell A M et al. Br J Dermatol 1999: 141: 1138–1139. 15 Duncan W C. J Dermatol Surg 1976: 2: 153–157. 40 Harrison B J et al. Br J Surg 1985: 72: 1002–1004. 16 Hartwell S W Jr. Surg Clin North Am 1975: 55: 1107–1109. 41 Lewis F et al. Br J Dermatol 1993: 129: 447–448. 17 Masson J K. Surg Clin North Am 1969: 49: 1043–1052. 42 Mortimer P S et al. Br J Dermatol 1986: 115: 263–268. 18 Rogers I W, Ryan R F. J La State Med Soc 1983: 135: 21–24. 43 Sawers R S et al. Br J Dermatol 1986: 115: 269–274. 19 Shaughnessy D M et al. JAMA 1972: 222: 320–321. 44 Stellon A J, Wakeling M. BMJ 1989: 298: 28–29. 20 Storino W D, Engel G H. Cutis 1978: 21: 338–341. 45 Woods G et al. Ohio State Med J 1972: 68: 864–866. 21 Tasche C et al. Plast Reconstr Surg 1975: 55: 559–562. 46 Bassukas I D, Hundeiker M. J Am Acad Dermatol 1997: 36: 1029. 22 Temelkov T, Troshev K. Acta Chir Plast 1984: 26: 246–252. 47 Freiman A et al. J Cutan Med Surg 2004: 8: 415–423. 23 Lapins J et al. Br J Dermatol 1994: 131: 551–556. 48 Konig A et al. Dermatology 1999: 198: 261–264. 24 Lapins J et al. J Am Acad Dermatol 2002: 47: 280–285.

Viewpoint 3 pational challenges. Together with the sequelae of the disease, including dermal contraction, keloid formation, According to the ‘Dessau definition’ of hidradenitis suppu- restricted limb mobility, lymphoedema and fistula forma- rativa ⁄ acne inversa (HS), this is a chronic, inflammatory, tion, this dominates the long-term burden of disease and recurrent, debilitating skin disease that usually presents greatly diminishes a patient’s quality of life. after puberty with painful, deep-seated, inflamed lesions in Aetiology and pathogenesis of HS are still unknown. the apocrine gland-bearing areas of the body, most com- Current theories implicate hyperkeratosis of the follicular monly the axillary, inguinal and anogenital regions (First epithelium as the morphological hallmark of HS patho- International Conference on Hidradenitis suppurativa, genesis, leading to occlusion of the apocrine glands with March 30–April 1, 2006, Dessau, Germany1). subsequent follicular rupture, inflammation and possible Its prevalence ranges from one per 600 to four per 100 secondary infection (3,4). From the genetic point of view, in the general population (1,2). HS affects more women the group of experts who participated at the First Interna- than men, with a female-to-male predominance as high as tional Symposium has accepted that HS must be a poly- 4:1. The majority of the patients are obese and smoke. The genic disease, with sporadic cases having defects in a disease occurs almost always after puberty and before the number of critical genes involved in its pathogenesis and age of 40 years. The apparent ‘rarity’ of HS in daily clinical familial cases with a probably highly penetrant defect(s) in practice may be explained by the prevalence of mild forms one of these genes.1 for which no consultation is sought, the weariness of HS Hidradenitis suppurativa has given rise to a number of patients discouraged by previous treatment results to seek lively clinical debates, such as the discussion on the exact further medical help, and by the generally poor level of skin appendage involved (eccrine sweat glands, apocrine education in the medical community on the proper recog- glands or terminal hair follicles) and the discussion on the nition and management of HS (this is manifest from the pathogenetic background, namely a genetic, hormonal, bac- often very late time at which the correct diagnosis is made, terial, genuine inflammatory, environmental or physical in the long run). The latter is particularly regrettable, as one. The significance of gender, body mass and smoking on the psychological impact on affected patients can be very disease prevalence is still under investigation. The major substantial, encompassing major social, personal and occu- therapeutic challenge in HS is the optimal choice between antibiotics, retinoids, corticosteroids, incision and drainage, 1Abstracts from the First International Hidradenitis Suppurativa local wound care, limited versus radical local excision, radi- Research Symposium, Exp Dermatol 2006: (15) 6: 478–482. ation, laser therapy and modern drugs, such as biologics, all

ª 2008 The Authors 6 Journal compilation ª 2008 Blackwell Munksgaard, Experimental Dermatology Controversies in Experimental Dermatology of which have been proposed – and whose relative impor- reduced risk to first-degree relatives, which falls short of tance and usefulness have been controversially debated. the expected 50% found in the study.

Which is the exact skin appendage involved? Hidradenitis suppurativa and autoimmunity One of the major debates in HS is the determination of the Clinical improvement with the application of therapies tar- exact skin appendage involved. Although initially accused, geted against tumour necrosis factor (TNF)-a may be com- the apocrine gland is probably ‘innocent’: neither exist sig- patible with the above theory of pathogenesis, as TNF-a is nificant differences in the size and density of the apocrine a major proinflammatory cytokine. In these studies, mono- glands in HS patients compared with normal controls (5), clonal anti-TNF-a etanercept receptors (12,13) or soluble nor significant morphological abnormalities of the apocrine TNF-a infliximab antibodies (14) were administered in a glands in diseased skin areas (6,7). No primary apocrine small number of patients. Positive responses with anti- involvement is detected; and the one-third of HS cases that TNF-a therapies have also addressed the question of present inflammatory changes involving the apocrine whether any probable autoimmune predilection might con- glands represents cases of very extensive inflammation tribute to the pathogenesis of HS (12). Based on the above that also engulf other structures such as the eccrine glands probability for the existence of some derangement of the and hair follicles. activity of the host immune function, a current study has A consistent finding in histological studies of HS is a fol- shown a reduction in the percentage of natural killer cells licular occlusion regardless of disease duration (8). The over time and a lower monocyte response to triggering by majority of specimens (50–85%) contain poral occlusion, bacterial components in patients with HS (15). sinus tracts or cysts (7). The histopathology of HS supports its classification to the follicular diseases. Newly formed The role of hormones in hidradenitis suppurativa nodules show an occluding spongiform infundibulofollicu- pathogenesis litis with secondary involvement of apocrine glands (9), i.e. The occurrence of HS in a narrow age spectrum and in a disease of the follicular epithelium in the terminal hair obese individuals as well as the female predominance have follicle. led to the theory that a hormonal component may be Different from acne vulgaris, HS is localized in non- involved in the pathogenesis of HS (16). Flare-ups have facial regions, where there are terminal, pigmented, coarse been linked with menses (17); shorter menstrual cycles and hairs, as in the axillae, groins, anal fold, mons pubis and longer duration of menstrual flow are associated with the scalp. These regions affected by acne inversa tend to be rich disease (18). Onset after menopause is rare (19). Sex hor- in apocrine glands, which are part of the apocrine-piloseba- mones may influence HS; association with acne vulgaris, ceous unit and can be engulfed in the inflammatory comedones and hirsutism, development post puberty, gen- process. The sebum excretion rate is not increased in HS. eral decline in disease activity seen at the climacteric, and The earliest inflammatory event in acne inversa is a improvement seen during pregnancy have been reported segmental rupture of the follicular epithelium, followed by (20–22). On the other hand, although most HS patients spilling of foreign body material, such as corneocytes, have normal androgen profiles (23), there have been bacteria, sebum products and hairs, into the dermis. The reports of symptomatic improvement with the use of anti- dumping of foreign products initiates an inflammatory androgen therapy (23–25). Indeed, HS has been regarded response provoking foreign body granuloma, and epithelial as an androgen-dependent disorder (26). HS is rarely a strands try to encapsulate the necrotic tissue. Once rupture presenting feature of premature adrenarche, leading sup- of the follicular epithelium has occurred, the disease spreads port to the view that it is androgen-dependent (27). rapidly. The draining sinus is a late complication of HS. It has been suggested that enhanced peripheral conver- sion of androgens by apocrine glands plays a critical part Hidradenitis suppurativa and genetic in its pathogenesis (26). However, equivalent activity of predisposition three peripheral androgen-converting enzymes in axillary Genetic factors may contribute to HS susceptibility. apocrine glands of subjects with HS was similar with those Patients frequently report HS cases among their relatives. of controls (28). A relationship between HS and In one study, 18 of 70 patients with HS (26%) had a posi- hyperandrogenism is largely based on the finding of an tive family history, whereas in 96 control subjects matched increased free androgen index (testosterone ⁄ sex hormone for age and sex, who did not have HS, only two of their binding globulin, SHBG) due to a low SHBG, but the sub- relatives suffered from HS (10). There appears to be auto- jects were not controlled for body mass index (BMI) (20). somal dominant inheritance with single gene transmission This finding is compromised, as many subjects are signifi- (10,11). A variable degree of gene penetrance and possibly cantly overweight (21) and SHBG is negatively correlated hormonal influence on gene expression may explain the with BMI (29). Another study could only demonstrate

ª 2008 The Authors Journal compilation ª 2008 Blackwell Munksgaard, Experimental Dermatology 7 Kurzen et al. hyperandrogenism in a subgroup of women who did not References experience a premenstrual flare in their disease (21) but no 1 von der Werth J M, Jemec G B E. Br J Dermatol 2001: 144: 809– supporting evidence for hyperandrogenism or suppression 813. of SHBG has been found in women with HS compared 2 Jansen T et al. J Eur Acad Dermatol Venereol 2001: 1: 532–540. 3 Slade D E M et al. Br J Plast Surg 2003: 56: 451–461. with age-, weight- and hirsutism-matched controls (19). 4 Wiseman M C. Dermatol Ther 2004: 17: 50–54. Hidradenitis suppurativa has also been reported to be 5 Morgan W P, Hughes L E. Br J Surg 1979: 66: 853–856. associated with classical endocrine disorders, such as 6 Yu C C W, Cook M G. Br J Dermatol 1990: 122: 763–769. Cushing’s syndrome (30) and acromegaly (31). This may 7 Jemec G B E et al. APMIS 1997: 105: 378–383. be interpreted as further indication that hormonal factors 8 Attanoos R L et al. Br J Dermatol 1995: 133: 254–258. 9 Boer J, Weltevreden E F. Br J Dermatol 1996: 135: 721–725. may play a significant role in HS pathobiology. 10 Fitzsimmons J S, Guilbert P R. Br J Dermatol 1985: 113: 1–8. 11 von der Werth J M et al. Br J Dermatol 2000: 142: 947–953. Outlook 12 Sullivan T P et al. Br J Dermatol 2003: 149: 1046–1049. In the era of molecular genetics, new diagnostic possibilities 13 Rosi Y L et al. J Dermatol Ther 2005: 16: 58–61. are emerging: gene expression profiling can be applied to 14 Cusack C, Buckley C. Br J Dermatol 2006: 154: 726–729. Giamarellos-Bourboulis E J et al. Br J Dermatol 2007: : 51– compare the gene expression pattern in the apocrine gland- 15 156 56. bearing areas of the body of HS patients versus healthy 16 Shah N. Am Fam Phys 2005: 72: 1547–1552. individuals. Bioinformatics will help to identify the most 17 Von der Werth J M, Williams H C. J Eur Acad Dermatol Venereol relevant among the differentially expressed genes. This may 2000: 14: 389–392. provide valuable insights into the basic pathologic processes 18 Jemec G B. Br J Dermatol 1988: 119: 345–350. 19 Barth J H et al. Br J Dermatol 1996: 134: 1057–1059. occurring in HS and may identify genes and biochemical 20 Mortimer P S et al. Br Med J 1986: 292: 245–248. pathways that could act as new targets for classical drugs in 21 Harrison B J et al. Br J Surg 1988: 75: 972–975. the treatment of HS. 22 Harrison B J. Br J Surg 1985: 22: 1002–1004. 23 Sawers R S et al. Br J Dermatol 1986: 115: 269–274. Christos C. Zouboulis 24 Goldsmith P C, Dowd P M. J R Soc Med 1993: 86: 729–730. Departments of Dermatology, Venereology, Allergology and 25 Camisa C et al. J Reprod Med 1989: 34: 543–546. Immunology, Dessau Medical Center, D-06847 Dessau, 26 Ebling F J G. Br J Dermatol 1986: 115: 259–262. and Laboratory for Biogerontology, Dermato-Pharmacology 27 Lewis F et al. Br J Dermatol 1993: 129: 447–448. 28 Barth J H, Kealey T. Br J Dermatol 1991: 125: 304–308. and Dermato-Endocrinology, Institute of Clinical Pharma- 29 Franks S et al. J Steroid Biochem Mol Biol 1991: 39: 835–838. cology and Toxicology, Charite´ Universitaetsmedizin Berlin, 30 Curtis A C. Arch Dermatol Syphilol 1950: 62: 329–330. Berlin, Germany; 31 Chalmers R J G et al. Br Med J 1983: 287: 1346–1347. E-mail: [email protected]

Viewpoint 4 This is best performed by considering selected key ques- tions in HS research. The most pressing ones arise immedi- Greek methodology has created the term ‘labyrinth’ for the ately, if one more carefully contemplates on: (a) the clinical convoluted indoor maze (Fig. 1a) that, according to legend, signs and symptoms presented by HS patients; and (b) the was constructed for King Minos of Crete (after the design proposal by Plewig et al. (4) that the term ‘HS’ should be by Daedalus) to hold the Minotaur (1), a creature half- changed to ‘acne inversa’. man, half-bull. Ranging from its clinical and histological presentation to What are the differences in clinical manifestation the many open questions posed by its pathobiology, hid- between HS and acne vulgaris? radenitis suppurativa (HS), in many respects, represents The characteristic lesions in HS are deep-seated subcutane- such a labyrinth. Clinically (Fig. 1b), and on histological ous nodules or abscess (2,5). In acne vulgaris, the primary analysis (Fig. 1c), the undermined sinus tracts in HS indeed lesions are either open or closed comedones (2,6). In HS, resemble a labyrinth (2). Despite intensive studies of HS closed comedones are never present, and open comedones (3), the aetiology of HS still remains unclear, and we are can be found as secondary lesions, but they are absent in still trapped in a maze of possibilities and conflicting early lesions of HS (2). They may appear in long-standing hypotheses. HS. Thus, the clinical appearance of HS and acne vulgaris Let us walk, then, through this labyrinth, searching for is quite distinct, and it adds little to our understanding of the Minotaur around which it may have been built. HS pathogenesis to adopt the ‘acne inversa’ terminology.

ª 2008 The Authors 8 Journal compilation ª 2008 Blackwell Munksgaard, Experimental Dermatology Controversies in Experimental Dermatology

(a)

(b)

Figure 2. Tufted hairs from one hair pouch in folliculitis decalvans.

Opening of fistels Undermined sinus tracts (c) Retention hyperkeratosis Exudate and pus discharging

Formation of Commuicating of sinus tract sinus tract

Rupture of the epithelium

Subcutaneous abscess Leaf-like structure Foreign body granuloma

Figure 1. Classical labyrinth (a) and clinical (b) and histological labyrinth (c) in hidradenitis suppurativa.

Are terminal hair follicles really the main theatre of HS action? Plewig and Kligman stated that HS occurs in the terminal follicle bearing pilosebaceous unit (2,4). If this were so, we Figure 3. Undermined fistulae in axilla in HS. Terminal hair follicles are not clinically affected in HS lesions. should find multiple thick hair shafts in HS. However, this is not really the case, and at least clinically (as opposed to histologically) terminal hair follicles actually are not the reverberated uncritically, in terminal (rather than in vellus) primary target in HS. In fact, despite some overlap, the hair follicles. areas of predilection for HS and the skin regions richest in But what about the apparent histopathological evidence terminal hair follicles seem to be rather distinct territories. for a predilection of HS for terminal hair follicles? Plewig In men, the buttock, a commonly affected site of HS and Kligman describe three types of hair follicles: vellus lesions, bears hardly any terminal hair follicles at all. In hair follicle, sebaceous follicle and terminal hair follicle (2). addition, if terminal hair follicles were the key targets of Well, in my personal experience, it is very difficult (if at all HS, multiple tufted terminal hair shafts from the follicle possible) to histologically detect terminal hair follicles in should be observed in HS, as is characteristically seen in HS. Even in standard textbooks (2,4,7), the classical charac- folliculitis decalvans (Fig. 2), dissecting folliculitis and peri- teristics of terminal hair follicles (thick, medullated hair folliculitis abscedens and suffodiens (Hoffman). Instead, shafts and large sebaceous gland), as described so beauti- multiple thick long terminal hairs are features one fails to fully in Plewig’s book (2), are essentially absent in HS. notice in the characteristic fistulae of HS (Fig. 3). Instead, Layton et al. (7) have observed tiny vellus hairs in Thus, from a clinical perspective, it is quite dubious follicular canals in HS. The lower part of hair follicles and whether HS really occurs, as frequently claimed and sebaceous gland are uninvolved in HS. Taken together,

ª 2008 The Authors Journal compilation ª 2008 Blackwell Munksgaard, Experimental Dermatology 9 Kurzen et al.

Coagulase-negative Staphylococcus (CNS), in normal flora on the skin, is the most common bacterium isolated from HS lesions (13). CNS is also the most common bacterium isolated from epidermal cysts (14). Therefore, CNS may tend to adhere to the adjacent, closed serrated epidermis in the intertriginous area, resulting in infundibular-like epi- thelium (cyst formation) in both HS and epidermal cyst. An alternative explanation may be that HS patients exhibit an altered immunologic response to CNS antigens, compared with normal individuals. Another possibility is that anaerobic bacteria (15) play a key role in retention hyperkeratosis, resulting in both sinus and cyst formation. Often, the protruding epithelium that forms from the Figure 4. CK 14 is expressed in the protruding epithelium in a leaf- like structure. wall of HS cysts shows a leaf-like structure, as shown in Fig. 4. Why do such leaf-like structures arise? Formation of the epithelial and mesenchymal interactions might be therefore, it is very dubious that HS predominantly affects related to undifferentiated keratin expression (CK14). In terminal hair follicles. addition, anastomosing epithelium that links sinus tracts Thus, even from the perspective of histopathology, HS is into a complex maze of tracts is frequently found in HS not a disorder of the terminal hair follicle. The fundamen- (Fig. 1c). Kurzen et al. (10) studied desmoplakin and des- tal change in HS is the retention hyperkeratosis of the mocollin in epithelia in HS, and demonstrated their differ- infundibulum (2,4,7,8). More likely, HS predominantly ences between three types of epithelium (10). Altered occurs in the infundibulum of vellus hair follicles, whose epithelial adhesion properties in HS might also result from tiny shafts can be observed in the follicular canal of abnormalities in e-cadherin expression. affected pilosebaceous units. The keratin expression pattern in HS has been studied Is the clinical phenotype of HS determined by an immunohistologically (9,10). Among cytokeratins, CK17 is abnormal immune response against bacteria? neither expressed in the cyst wall in most cases of HS (9) It is clear that bacterial infection is not the primary cause nor is found in the pilonidal sinus of HS (11). CK 14 is of HS. However, bacteria, their products, and immune more pronounced in the protruding epithelium in HS (9) responses raised against bacterial antigens may well greatly (Fig. 4) In terminal hair follicles, CK 16 is detectable in the influence the clinical phenotype of HS lesions as well as the lower portion of hair follicle below the opening of seba- course of HS in any given patient. From normal skin ceous duct (9–12). The fact that CK16 was only found in microflora, CNS, Staphylococcus aureus, Stereptococcus, two of 16 HS cases (9); therefore, further argues against Peptostreptococcus, etc. are isolated in HS, e.g. in perirectal terminal hair follicles as key targets in HS pathogenesis. and vulvovaginal lesions (16,17). However, it is unclear However, as these immunohistological results illustrate, HS whether these clinical isolates are involved in the pathogen- is definitely associated with abnormalities in keratinization. esis of HS or not. CNS obstructs the intra-epidermal sweat The next questions, then, are as follows. duct, resulting in miliaria formation (18). Thus, it is con- ceivable that sinus formation in HS involves a similar What causes retention hyperkeratosis, and how mechanism. does this lead to the characteristic, labyrinthic In terms of natural immunity, patients with HS may sinus tracts of HS? respond abnormally to these bacteria (e.g. Propionibacteri- Based on the above keratin expression patterns and the um acnes), whose products stimulate Toll-like receptors unconvincing evidence that terminal hair follicles are the (19), producing pro-inflammatory cytokines such as IL-1a, key organ where HS manifests itself, I propose the follow- and giving rise to abnormal keratinization (20). The hair ing hypothesis. follicle was recently recognized as one of the immune The sinus epithelium in HS is not a late, but a very early organs of skin (21). In particular, the infundibulum, which event and does not primarily arise from the hair follicle belongs to the distal arm of the human hair follicle epithelium, but by invagination from the epidermis immune system, appears to represent a special organ (Fig. 1c). This invagination of the epidermis may result in involved in immune response (22). These areas are the the formation of infundibulum-like epithelium that pro- preferred sites for perifollicular inflammatory cells in der- trudes into the deep dermis as epidermal cysts. Resident matoses such as lichen planopilaris, systemic lupus erythe- microflora may cause the adherence of the epithelium. matosus, scleroderma and folliculitis decalvans (23).

ª 2008 The Authors 10 Journal compilation ª 2008 Blackwell Munksgaard, Experimental Dermatology Controversies in Experimental Dermatology

Therefore, HS may just represent a certain clinical pheno- Is ‘altered epithelial adhesion’ and ⁄ or ‘altered immune type that reflects a disturbed hair follicle immune system. response against bacteria’ the answer…?

Important studies that remain to be performed in Ichiro Kurokawa HS research Department of Dermatology, Mie University Graduate In consequence, infundibular immunology, therefore, needs School of Medicine, 2-174, Edobashi, Tsu, to be systematically dissected in future. For example, Toll- Mie 514-8507, Japan like receptors may be related to the initiation of inflamma- E-mail: [email protected] tion in the infundibulum. In this context, the impact of pro-inflammatory cytokines on abnormal keratinization in References HS should also be carefully investigated. 1 Penelope Reed Dob, The Idea of the Labyrinth: From Classified Antiquity Through the Middle Ages. New York: Cornell University Other important areas for future HS research include Press, 1992. for example the precise mechanism by which new epithe- 2 Plewig G et al. Acne and Rosacea. Berlin: Springer-Verlag, 2006: lium is formed in the process of sinus drainage (in HS 309–341. and other diseases where this occurs). Scar formation 3 Jemec G B. J Cutan Med Surg 2003: 7: 47–56. involves a cell-mediated response that deserves study to 4 Plewig G et al. Acne and Related Disorders. London: Martin Dunitz, 1988: 345–347. clarify the pathogenesis of HS, like in acne vulgaris (24). 5 Poli F et al. Hidradenitis Suppurativa. Berlin: Springer-Verlag, 2006: With regard to possibly increased carcinogenesis in HS, 11–24. the CK expression in well-differentiated SCC arisen from 6 Cunliffe W J et al. Acne – Diagnosis and Management. London: HS epithelium is similar to that in normal infundibulum Martin Dunitz, 2001: 49–67. (25). By contrast, poorly differentiated SCC in HS con- 7 Layton A M. Hidradenitis Suppurativa. Berlin: Springer-Verlag, 2006, 25–33. tained simple epithelial keratins (CK7, 8, 18, 19). Thus, it 8 Jansen T et al. Int J Dermatol 1998: 37: 96–100. deserves to be carefully studied whether CK expression 9 Kurokawa I. J Int Med Res 2002: 30: 131–136. patterns reflect the clinical prognosis. Finally, the terminal 10 Kurzen H et al. Br J Dermatol 1999: 141: 231–239. stage of keratinization is related to filament-aggregating 11 Kurokawa I et al. Br J Dermatol 2002: 146: 409–413. protein (filaggrin), a major component of keratohyaline 12 Kurokawa I et al. Br J Dermatol 2003: 149: 99–104. 13 Lapins J et al. Br J Dermatol 1999: 140: 90–95. granules (26). Filaggrin expression has been linked to the 14 Nishijima S et al. Jpn J Dermatol (in Japanese) 2003: 113: 165–168. pathogenesis of acne vulgaris (27), epidermal cyst (28) 15 Brook H S et al. J Med Microbiol 1999: 48: 103–105. and nevus comedonicus (29). Therefore, an immunohisto- 16 Leach R D et al. Br J Med 1979: 2: 5–7. chemical study of filaggrin expression should be under- 17 Brook I. Arch Dermatol 1989: 25: 1658–1661. taken. As comprehensive ultrastructural studies of HS 18 Mowad C M et al. J Am Acad Dermatol 1995: 33: 729–733. 19 Kim J et al. J Immunol 2002: 169: 1535–1541. remain to be performed, this should be complemented by 20 Guy R et al. J Invest Dermatol 1996: 136: 166. electron microscopy. 21 Paus R et al. Trends Immunol 2005: 26: 32–40. In the long run, at least the ‘floor plan’ of the HS laby- 22 Christoph T et al. Br J Dermatol 2000: 142: 862–873. rinth is becoming better defined, defined scenarios can be 23 Paus R et al. J Invest Dermatol Symp Proc 1999: 4: 226–234. developed on how it is being constructed and evidence is 24 Holland D B et al. Br J Dermatol 2004: 150: 72–81. 25 Kurokawa I et al. J Cutan Pathol 2007: 21: 675–678. emerging that its epithelial walls show distinct abnormali- 26 Dale B et al. J Invest Dermatol 1979: 72: 257–261. ties of keratinization. 27 Kurokawa I et al. J Invest Dermatol 1988: 91: 566–571. But what is the Minotaur it may contain? 28 Kurokawa I et al. Br J Dermatol 2007: 157: 415–416. 29 Kurokawa I et al. J Cutan Pathol 2007: 34: 338–341.

Viewpoint 5 ‘Acne inversa’ indicates that this is a disease of follicular occlusion similar to acne vulgaris. Pathogenetically, how- Before you read these comments, the authors want you to ever, more can be said about what does not cause ‘hidrade- know that we have a ‘conflict of interest’ – we do not nitis suppurativa’ than what causes it. believe in ‘hidradenitis suppurativa’! While we will not be Similar to other forms of acne, apocrine glands have no engaging in a terminological debate, in our opinion ‘acne role in acne inversa; the glands are guilty by association inversa’ is the appropriate term for this often devastating only and as such are merely ‘innocent bystanders’ (Fig. 1). disease (1). It is a term which – in a nutshell – already puts This is documented in numerous studies (1–6). The early forth part of our answer to the introductory question experimental model for ‘hidradenitis suppurativa’, pub- posed by the editor: What causes hidradenitis suppurativa? lished in 1955 by Shelley and Cahn (7), was designed to

ª 2008 The Authors Journal compilation ª 2008 Blackwell Munksgaard, Experimental Dermatology 11 Kurzen et al.

(a) (b) (c)

(d) (e) (f)

Figure 1. (a) In the early stage of the disease, microcomedones are characteristic (arrowhead). Note that the apocrine glands are unremarkable (arrow) and that the epidermal surface of the axillary skin is thrown into folds. (b) A comedo has ruptured (arrow) and the body initially tries to contain the acute inflammation by surrounding granulation tissue. (c) If unsuccessful in confining the disease, the inflammatory response following rupture of the comedo is more florid. Note that the apocrine glands to the right of the ruptured nodule are not involved. (d) Apocrine glands, however, become secondarily affected when they are close to the focus of inflammation (arrow). (e) Naked hair shafts (arrowheads) may be the only indication that the process started from the hair follicle. Note that the apocrine glands outside of the area of inflammation are not affected (arrow). (f) In the later stage of the disease sinus tracts, often surrounded by prominent inflammation, are the predominant feature. [All sections stained with haematoxylin–eosin; original magnification: (a) 16·; (b) 200·; (c) 25·; (d) 25·; (e) 50·; (f) 16·]. With permission from Sellheyer, K., Krahl, D. ‘Hidradenitis suppurativa’ is Acne inversa! An appeal to (finally) abandon a misnomer. Int. J. Dermatol 2005; 44: 535–540. recapitulate the changes seen in patients with ‘hidradenitis tory products within the microcomedones, leading to their suppurativa’ and relied on the application of belladonna enlargement and later rupture. The hair follicles in ‘hid- adhesive tape to manually depilated axillary skin. It was radenitis suppurativa’ exhibit an abnormal shape, are wider one of the most important papers emphasizing the central in the deep dermis and are also larger in the axilla (13). role of apocrine glands in the disease process but was pla- We interpret this finding as corroborating our biomechani- gued with flaws and assumptions, although the conclusions cal theory. derived from it were considered dogma – at least for many A close relative of ‘hidradenitis suppurativa’ and part of years. In our opinion, the paper did not deserve the the acne tetrad (14), the pilonidal sinus, is equally caused attention it got and the attention was uncritical. in our opinion by biomechanical factors. Known also as Knowing that at centre stage in ‘hidradenitis suppurati- ‘jeep disease’ (15), it has a predilection for soldiers sub- va’ is follicular occlusion, the question remains what causes jected to long-term occlusive sitting conditions (16,17). it or – more specifically – what causes it in the axillae and Outside the intertriginous areas, the human body employs in the groin area, the predominant anatomical locations clever mechanisms to combat mechanical retention. A good (hence the term acne inversa). It is well documented that example is the external auditory canal, where a peculiar pat- obese females are most commonly affected (8–11). Normal tern of keratinocyte migration occurs for the sole purpose of axillary skin displays already a ruffled appearance under the keeping the meatus free from desquamation products (18). microscope (Fig. 1a) as opposed to the even epidermal sur- The epidermal front of the mucocutaneous junction of the face observable in biopsies, e.g. from facial skin. We specu- tympanic membrane simply expresses hyperproliferation- late that constant friction in the axillae and in the groin, associated cytokeratins and such keratinocytes migrate, thus enhanced by obesity, is the major biomechanical factor preventing their accumulation and allowing also cerumen to contributing to microcomedo formation from which the be disposed to the outside. subsequent cascade of pathogenetic events initiates. Micro- Epidemiologically, it is known that most patients with comedones are part of the histopathological spectrum of ‘hidradenitis suppurativa’ are smokers (19–21). A recent early disease (1,12). Especially the folds observable in the intriguing paper by Hana et al. (22), one of the few experi- skin of obese individuals contribute mechanically to the mental studies in the field, convincingly demonstrated the retention of corneocytes, hair shafts and sebaceous secre- induction of infundibular epithelial hyperplasia by nicotine

ª 2008 The Authors 12 Journal compilation ª 2008 Blackwell Munksgaard, Experimental Dermatology Controversies in Experimental Dermatology as a prerequisite for follicular plugging. The authors Klaus Sellheyer employed organotypic cultures and suggested the non-neu- Department of Dermatology, University of Alabama at ronal acetylcholine receptors, prominent around the follic- Birmingham, Birmingham, AL 35294, USA ular infundibulum, as the nicotine-mediating effector E-mail: [email protected] system (22). This would be the first logical explanation of Dieter Krahl why mostly smokers are predominantly affected by the dis- Institut fu¨r Dermatohistologie, D-69120 Heidelberg, ease and also confirms our findings of microcomedones as Germany; an early pathogenetic event. While we do not think that E-mail: [email protected] nicotine is the inducing factor, it certainly is a major contributing factor of ‘hidradenitis suppurativa’. In later References disease stages, nicotine seems to be less important, as the 1 Sellheyer K, Krahl D. Int J Dermatol 2005: 44: 535–540. non-neuronal cholinergic system is not prominent in sinus 2 Yu C C-W, Cook M G. Br J Dermatol 1990: 122: 763–769. tracts (22). By then, however, the disease has already devel- 3 Felding C, Moesgaard J, Clevin L, Fischer S. Ambulat Surg 1995: 3: 3–6. oped a life of its own. Nicotine is also secreted in apocrine 4 Jemec G B E, Hansen U. J Am Acad Dermatol 1996: 34: 994–999. and eccrine sweat (23), which may play a direct role in 5 Jemec G B E, Thomsen B M, Hansen U. APMIS 1997: 105: 378– ‘hidradenitis suppurativa’. It remains speculative if, in addi- 383. tion to nicotine, tar products in cigarette smoke are 6 Heller D B, Haefner H K, Hameed M, Lieberman R W. J Reprod secreted via sweat and find their way to the follicular Med 2002: 47: 695–700. 7 Shelley W B, Cahn M M. Arch Dermatol 1955: 72: 562–565. infundibula, thereby exerting their well-documented 8 Edlich R F, Silloway K A, Rodeheaver G T, Cooper P H. J Emerg comedogenic effects. Med 1986: 4: 369–378. From the above, it is obvious that we are in dire need of 9 Harrison B J, Read G F, Hughes L E. Br J Surg 1988: 75: 972–975. more basic science data on this enigmatic disease, foremost 10 Barth J H, Layton A M, Cunliffe W J. Br J Dermatol 1996: 134: to help our patients. They may otherwise fall into the 1057–1059. 11 Yosipovitch G, DeVore A, Dawn A. J Am Acad Dermatol 2007: 56: trap of alternative or holistic medicine using the lack of 901–906. knowledge about ‘hidradenitis suppurativa’ to its commer- 12 Plewig G. Acne inversa, Acne keloidalis nuchae, abszedierende Fol- cial advantage by selling products with no therapeutic likulitis der Kopfhaut: ein verbindendes Konzept. In: Plewig G, Prinz benefits, such as algae (http://www.akne-inversa.de/projekt/ J, eds. Fortschrritte der Praktischen Dermatologie und Venerologie spiru-studie/index.htm), thereby extending the ordeal of the 2002. Berlin: Springer, 2003: 192–203. 13 Jemec G B, Gniadecka M. Arch Dermatol 1997: 133: 967–970. patients and preventing them to seek adequate treatment. 14 Plewig G, Steger M. Acne inversa (alias acne triad, acne tetrad or Reflecting on ‘What causes hidradenitis suppurativa?’we hidradenitis suppurativa). In: Marks R, Plewig G, eds. Acne and are still astounded that even the current fragmented level Related Disorders. London: Martin Dunitz, 1989: 345–357. of knowledge on ‘hidradenitis suppurativa’ cannot be com- 15 Patey D. Nurs Times 1971: 67: 534–536. prehended by many representatives of allopathic medicine. 16 Clothier P R, Haywood I R. Ann R Coll Surg Engl 1984: 66: 201– 203. An especially baffling example is that of a well-known 17 Chijiwa T, Suganuma T, Takigawa T et al. Mil Med 2006: 171: cosmetic dermatologist from Nashville, TN, who asked for 650–652. copyright permission for histological photographs from our 18 Vennix P P, Kuijpers W, Peters T A, Tonnaer E L, Ramaekers F C. own paper on acne inversa (1). After reproduction of Laryngoscope 1996: 106: 470–475. the photomicrographs, he elaborates in his review on the 19 Ko¨ nig A, Lehmann C, Rompel R, Happle R. Dermatology 1999: 198: 261–264. application of photodynamic therapy for ‘hidradenitis sup- 20 Bassukas I D, Hundeiker M. J Am Acad Dermatol 1997: 36: 1029. purativa’: ‘Because HS [= hidradenitis suppurativa] is an 21 Freiman A, Bird G, Metelitsa A I, Barankin B, Lauzon G J. J Cutan apocrine disorder, and not a sebaceous gland problem Med Surg 2004: 8: 415–423. […]’ (24). This is the exact opposite of what we have 22 Hana A, Booken D, Henrich C et al. Life Sci 2007: 80: 2214–2220. hoped to rectify with our histological pictures. We 23 Balabanova S, Buhler G, Schneider E, Boschek H J, Schneitler H. Hautarzt 1992: 43: 73–76. hope the reader of this commentary is not a cosmetic 24 Gold M H. Dermatol Clin 2007: 25: 67–73. dermatologist!

Commentary 1 making largely depends on the impression of the physician about pathogenesis of HS. This may be connected to the Hidradenitis suppurativa (HS) is a devastating chronic skin high rate of treatment failure (2). HS was initially con- disorder affecting areas rich in apocrine glands. HS is in ceived as an infectious disease process or as a form of acne. most cases recalcitrant to therapy (1). Therapy decision Patients run a life of exacerbations and remissions with

ª 2008 The Authors Journal compilation ª 2008 Blackwell Munksgaard, Experimental Dermatology 13 Kurzen et al.

1000 the same study, it was shown that percentages of natural killer (NK) cells were decreased in parallel with the natural history of HS (Fig. 1). These findings favour the existence P < 0.0001 of significant alterations of innate immune functions of HS patients. They also support indirectly an auto-immune aeti- 750 ology as antigen presentation is based on the integrity of the innate immune system. The concept of HS as an autoimmune disorder is also based on the favourable responses of patients to targeted 500 therapies against TNF-a. Infliximab, a chimeric anti-TNF-a monoclonal antibody, and etanercept, a soluble receptor of TNF-a, have been administered in case-studies with limited number of patients and at a variety of dose regimens. 250 Results of infliximab were contradictory in terms of safety (pg/10000 cells, median +/– 95% CI) CI) 95% median +/– (pg/10000 cells,

α and efficacy in case-studies of five and seven patients each (9,10), whereas etanecept was effective and well tolerated in TNF a series of six patients (11). The only open-label prospective study on the safety and 0 Controls HS efficacy of a specific regimen of etanercept has been published by our group (12) (EudraCT: 2004-004555-19, Figure 1. Release of TNF-a by monocytes isolated from 53 patients http://www.clinicaltrials.gov, NCT00329823). with hidradenitis suppurativa (HS) after stimulation with endotoxin (LPS) of Escherichia coli O111:B4 compared with six healthy volunteers. Etanercept was administered at a dose of 50 mg once weekly for 12 weeks in 10 patients. A more than 50% heavy purulent discharge from the affected areas. The pres- decrease in the Sartorius score was found in six patients at ence of pus creates the impression of an infectious disease week 12 and in seven patients at week 24. A considerable which is compatible with the traditional theory of patho- decrease in the total number of fistulas from the baseline genesis. In that theory, follicular hyperkeratosis leads to was seen over all weeks of follow-up. Eight patients reported occlusion of the apocrine glands with secondary infection recurrence of drainage of pus from the affected areas within from bacterial skin flora (3). As a consequence, antibiotics 4–8 weeks after the end of administration of etanercept. are prescribed. Even though antibiotics may offer relief of There is accumulating indirect evidence for a key role of symptoms, cessation of treatment is usually accompanied the immune system in HS pathogenesis. Derangements of by flare-ups (4). This creates serious doubts if HS is indeed the innate immunity, favourable responses to therapy with an infectious process or not. The fate of treatment with etanercept and exacerbations upon cessation of anti-TNF-a retinoids is similar (5). treatment, support auto-immunity as the underlying cause Surgical excision of the affected areas is accompanied by of HS. high recurrence rates (6,7). Carbon dioxide laser locally applied for the management of 35 patients with Hurley II Evangelos J. Giamarellos-Bourboulis lesions ended in recurrence in 25 patients. Surprisingly, Fourth Department of Internal Medicine, ‘‘Attikon’’ relapse supervened in another anatomical region, different University General Hospital, University of Athens from the one which was operated (8). Re-appearance in Medical School, Athens, Greece; another site after immune triggering is a characteristic of E-mail: [email protected] auto-immune disorders which creates the hypothesis if HS is an autoimmune disorder. This is compatible with the References coexistence of HS with other auto-inflammatory disorders 1 Shah N. Am Fam Physician 2005: 72: 1547–1552. 2 Jemec G B E. Expert Opin Pharmacother 2004: 5: 1767–1770. like Crohn’s disease (1). 3 Wiseman M. Dermatol Ther 2004: 17: 50–54. In a recent study of our group (8), peripheral blood 4 Slade D E M et al. Br J Plast Surg 2003: 56: 451–461. monocytes were isolated from 53 patients with active HS 5 Krbec A C. J Am Acad Nurs Pract 2007: 19: 228–234. and stimulated with bacterial endotoxin (LPS) for the pro- 6 Kagan R J et al. Surgery 2005: 138: 734–741. duction of pro-inflammatory cytokines, namely tumour 7 Lapins J et al. J Am Acad Dermatol 2002: 47: 280–285. 8 Giamarellos- Bourboulis E J et al. Br J Dermatol 2007: 156: 51–56. necrosis factor-alpha (TNF-a) and interleukin-6 (IL-6). 9 Fardet l et al. J Am Acad Dermatol 2007: 56: 624–628. Cells isolated from six healthy volunteers were controls. It 10 Sullivan T P et al. Br J Dermatol 2003: 149: 1046–1049. was found that monocytes of patients with HS were weaker 11 Cusack C, Buckley C. Br J Dermatol 2006: 154: 726–729. producers of cytokines compared with healthy controls. In 12 Giamarellos-Bourboulis E J et al. Br J Dermatol 2008: 158: 567–572.

ª 2008 The Authors 14 Journal compilation ª 2008 Blackwell Munksgaard, Experimental Dermatology Controversies in Experimental Dermatology

Commentary 2 this system, for example dysregulation and abnormal expression of inflammatory mediators or their receptors in Although recognized more than 150 years ago, hidradenitis KCs, can lead to the pathogenesis of chronic inflammatory suppurativa (HS) is still a mysterious disease. Ever since skin diseases, such as HS. inflammation of the apocrine glands was recognized as a The significance of bacterial involvement in the patho- secondary event in the pathogenesis of the disease, it has genesis of HS is also controversial. It is likely that chronic been speculated that comedonal and ⁄ or poral occlusion(s) inflammation is because of secondary bacterial colonization and bacterial infection play a crucial role. Even though this (5). This is further supported by the fact that routine cul- concept has been questioned almost immediately, there is tures from the surface of the lesions are often negative. some consensus regarding the role of immunity and infec- Still, bacteria are likely to be involved in the pathogenesis tion. Thus, even if I cannot provide an answer to the ques- of the disease as numerous species are most frequently iso- tion posed by the editor, let me elaborate on the issues of lated from lesions (6,9). Most of bacteria identified in HS skin immune response and infection in HS pathogenesis. lesions, such as Propionibacterium acnes and coagulase-neg- The involvement of immune response in HS remains ative staphylococci, are part of the normal microflora, but controversial. Immunological investigations of patients with have also gained attention as pathogens. These findings HS suggested no abnormalities of the immune system (1). highlight a possible polymicrobial nature and predomi- By contrast, other authors showed increased peripheral nance of anaerobic bacteria, supporting the role of bacterial suppressor T-cell activity (2), indicative of a cellular infections as a possible pathogenic event in HS. However, immune response. This is further supported by the interpreting the results of previous studies is difficult, as presence of activated, HLA-DR-positive as well as Leu-8- potential differences amongst recently discovered phyloge- positive immunoregulatory lymphocytes (3). These results netic groups and ⁄ or ecotypes have not been taken into indicate that the lymphocytic infiltrate is definitely the account. Notably, the existence of philogenetically distinct result of in vivo activation of lymphoid cells. Indeed, the P. acnes clusters have recently been demonstrated (10). significant fall of the T-helper ⁄ suppressor and NK cell ratio Importantly, these clusters differ in the production of over time after the initiation supports the existence of a secreted proteins (11), and induce different immune precipitating, cell-mediated immune response with only a responses in KCs and sebocytes (12,13). short eliciting period (3,4). More recent studies have shown These findings challenge our current understanding of that dysfunctional neutrophils and monocytes may also be the pathogenic nature of bacteria involved in HS pathogen- involved in the pathogenesis of HS, still, no primary abnor- esis and raise the exciting possibility that bacterial strains, malities of the innate or acquired immune system can be or group of strains, with greater potential to cause oppor- held to be causal in every case (3–6). tunistic infection in HS may exist. This may explain, in Because of increasing evidence suggesting that keratino- part, the apparent controversy with respect to the role of cytes (KCs) not only participate in cutaneous immune bacterial infection in HS. responses but may in fact play key initiation roles (7), ´ one must also consider the contribution of KCs to HS Istvan Nagy pathogenesis. KCs are able to recognize a wide variety of Institute for Plant Genomics, Human Biotechnology and micro-organisms through their pattern recognition recep- Bioenergy, Bay Zolta´n Foundation for Applied Research, tors (PRRs) and have evolved mechanisms to distinguish Szeged, Hungary; between skin commensals and pathogens. Signalling E-mail: [email protected] through specific PRR combinations provides selectivity and specificity to immune response. As a result, KCs produce a References wide range of antimicrobial peptides, proinflammatory 1 Dvorak V C et al. Arch Dermatol, 1977: 113: 450–3. 2 McDaniel D H, Welton W A. Arch Dermatol, 1984: 120: 437. cytokines ⁄ chemokines and inducible enzymes (8). The 3 Boer J, Weltevreden E F. Br J Dermatol, 1996: 135: 721–5. secretion of antimicrobial peptides is indeed crucial, as skin 4 Giamarellos-Bourboulis E J et al. Br J Dermatol, 2007: 156: 51–6. lesions characterized by low levels of such host-defence 5 Jansen I P et al.. J Eur Acad Dermatol Venereol, 2001: 15: 532–40. peptides are more susceptible to infections. By exhibiting 6 Lapins J, J C et al. Br J Dermatol, 1999: 140: 90–5. chemoattractant activity, KC-derived cytokines ⁄ chemokines 7 Schroder J M et al. Exp Dermatol, 2006: 15: 913–29. 8 Pivarcsi A et al. Curr Immunol Rev, 2005: 1: 29–42. and antimicrobial peptides can recruit T cells, neutrophils 9 Jemec G B et al. Dermatology, 1996: 193: 203–6. and dendritic cells into sites of infection, thus providing an 10 McDowell A et al. J Clin Microbiol, 2005: 43: 326–34. improved immune response against pathogens (7,8). These 11 Valanne S et al. Microbiology, 2005: 151: 1369–79. findings indicate a close interdependence of KCs and 12 Nagy I et al. Microbes Infect, 2006: 8: 2195–205. inflammatory infiltrate as well as a balance between the 13 Nagy I et al. J Invest Dermatol, 2005: 124: 931–8. innate and acquired immune systems. Any perturbation in

ª 2008 The Authors Journal compilation ª 2008 Blackwell Munksgaard, Experimental Dermatology 15 Kurzen et al.

Commentary 3 these patients frequently respond to retinoid treatment, whereas the majority of HS patients do not benefit from Hidradenitis suppurativa (HS) has originally been consid- retinoid. ered a disorder of the apocrine glands.(1) However, over the By contrast, there are the classical obese smoking years more and more authors suggested HS to be a misno- HS ⁄ AI patients with chronic inflammatory lesions, which mer as the disease seems to start rather with follicular occlu- support the concept that nicotine and obesity present a sion than with a primary involvement of apocrine sweat significant risk factor. (5) Compared with the normal- glands. Since 1989, the term acne inversa (AI) is commonly weight patients, we rarely notice any comedones in this used instead of HS to underline that this disorder is actually patient population at all, even in early lesions or healthy a defect of follicular epithelium (2–4). This concept does, surrounding skin. however, not explain sufficiently the causes of follicle A third type of patient, although less frequently encoun- involvement and possible subsequent follicle rupture. tered, presents with typical HS ⁄ AI lesions in conjunction When discussing with colleagues of different medical dis- with granulomatous colitis (Crohn’s disease). We have ciplines, we often are under the impression that the experts never seen comedones in this collective. Are theses patients treating this disease have split in two camps – rigorous really suffering from the same type of HS or AI as is supporters of the term HS, and equally stern supporters of present in slim patients with a prominent history of acne the term AI. The ongoing conflict between these two hos- vulgaris, or the obese smoking female patients? tile camps may actually have put basic HS research into a Based on clinical observation and experience, we hypoth- kind of ‘hibernation’, rather than having promoted scien- esize that we do not deal with the same pathobiological tific breakthroughs in this obscure, but very important entity in all patients that meet currently accepted diagnostic chronic inflammatory disease. criteria for HS or AI. Further studies with an accurate and As a provocative approach to possible causes of HS or refined definition of distinct patient subcollectives is likely AI is favoured on these pages, we therefore would like to to also provide new answers to the central question that modify the question raised by the editor into one that drives these CONTROVERSIES. often rises in clinical routine: Do all patients that we treat for HS ⁄ AI really suffer from the same disease? Falk G. Bechara, Peter Altmeyer From the point of view of daily clinical routine, we Department of Dermatology and Allergology, Ruhr-Univer- have long wondered about the extremely different patient sity Bochum, Germany; St Josef Hospital, Gudrunstr. 56, D populations that are referred to our clinic with a diagno- 44791 Bochum, Germany; sis of HS or AI. Some of our patients report acne vulgaris E-mail: [email protected] to have occurred in their medical history. Theses patients often show clinical signs or remnants of a previous severe References acne, most often presenting as typical scarring of the face 1 Brunsting H A. Arch fur Dermatol Syph (Berlin) 1939: 39: 108–120. 2 Plewig G, Steger M. Acne inversa (alias acne triad, acne tetrad or and back. Others still have active acne lesions. These hidradenitis suppurativa). In: Marks R, Plewig G, eds. Acne and patients often do not show obesity (which has been Related Disorders. London: Martin Dunitz, 1989: 345–57. claimed to be a characteristic of HS patients). Instead, 3 Slade D E M et al. Br J Plast Surg 2003: 56: 451–461. they usually present comedones within or surrounding the 4 Wiseman M C. Dermatol Ther 2004: 17: 50–54. inflammatory lesions. Interestingly, in our experience, 5 Jansen T et al. J Eur Acad Dermatol Venereol 2001: 15: 532–540.

Viewpoint 6 USA). By contrast, different triggering factors are well known, amongst which tobacco smoking, sweating, obesity Hidradenitis suppurativa (HS) is a chronic inflammatory, and colonization with Staphylococcus aureus seem to be the disabling disease that has been shown to emerge from the most important (3). pilosebaceous unit of the intertriginous areas and is hence In the following, I would like to place full emphasis on preferentially called acne inversa in the European literature what I consider a still insufficiently appreciated aspect of (1). HS is certainly a multifactorial disease based on a HS, i.e. why and how tobacco-smoking is involved in HS genetic background with different triggering factors. The pathogenesis: genetic background that predisposes the individual for HS 1 Between 80–90% of HS patients are active smokers, has so far escaped elucidation. First attempts to localize the which designates HS, together with, e.g. pustular palmopl- HS susceptibility locus to chromosome 1q (2) have not antar psoriasis, a clearly tobacco-related skin disease (3,4). been confirmed by other groups (U. Radhakrishna, Omaha, In a questionnaire-based study, we have noticed that HS

ª 2008 The Authors 16 Journal compilation ª 2008 Blackwell Munksgaard, Experimental Dermatology Controversies in Experimental Dermatology

Figure 1. Proposed sequence of events in the pathogenesis of acne (a) inversa: Nicotine fuels the vicious circle of ACh production and Staphylococcus aureus growth (a) promoting chemotaxis of neutrophilic granulocytes and degranulation of mast cells (b). In addition the infundibular epidermis becomes increasingly hyperplastic leading to follicular obstruction (b). Influx of neutrophilic granulocytes and mast cell degranulation promote a spongiotic infundibulitis. The acute phase is followed by a chronic phase characterized by the appearance of lymphocytes and macrophages (c) culminating in the rupture of the hair follicle (d). Draining sinus are formed in an attempt to eliminate residual hair-shafts. ACh Acetylcholine, ASG apokrine sweat gland, ChAT choline acetyl transferase, ESG eccrine sweat gland, GC foreign body giant cell, HF hair follicle, LC lymphocyte, MC mast cell, NG neutrophilic granulocyte, Mph macrophage.

patients develop new lesions after surgical therapy only if (b) they continued smoking (4). 2 Nicotine is the main toxin in cigarette smoke and reaches serum levels of 50–500 nm (5), while in axillary sweat nicotine concentrations of up to 150 nm have been measured (6). After smoking one cigarette, nicotine can be detected for up to 7 days in axillary sweat, indicating a slow metabolization rate (7). Thus, due to the density of sweat glands and the occlusive milieu in the intertriginous areas, nicotine is present on the surface in relevant phar- macological concentrations, higher than elsewhere on the skin. This observation would explain the distribution of HS lesions and the interdependency to sweating, provided nicotine really is a major pathogenic factor in HS. 3 As a natural alkaloid occurring in plants, nicotine has (c) a selective antimicrobial activity. An important exception in this is S. aureus, the growth of which is actually favoured by nicotine (8). Consequently, HS patients are, like smokers in general, heavily colonized with S. aureus (4,9). This vicious circle is closed by the fact that S. aureus has been shown to stimulate the expression of choline-acetyltransferase (ChAT), the rate-limiting enzyme in the synthesis of acetyl- choline (ACh) (10). ACh, in turn, can stimulate S. aureus growth – analogous to nicotine (Fig. 1a,b). 4 Indeed, the production of ACh in human skin (as suggested by ChAT immunoreactivity) is limited to kerati- nocytes, sebocytes, endothelial cells, lymphocytes and neu- trophilic granulocytes, finding its peak in peri-infundibular basal keratinocytes of HS patients (blue line in Fig. 1a,b) (11). (d) 5 The natural ligands of nicotine are nicotinic acetylcho- line receptors (AChR) that can be found in a highly complex composition on all cells putatively participating in the path- ogenesis of HS: keratinocytes, sebocytes, mast cells, neutro- philic granulocytes, lymphocytes and macrophages (12). 6 Despite their somewhat misleading name, not all nAChR are stimulated by nicotine. While nicotine leads to calcium influx in homopentameric a7 nAChR and to sodium and ⁄ or potassium flux after binding a3* nAChR

ª 2008 The Authors Journal compilation ª 2008 Blackwell Munksgaard, Experimental Dermatology 17 Kurzen et al.

(13), in both a9 and a10 nAChR, ion flux is actually inhib- more complex, but even those may still be prominently be ited after nicotine exposure. Therefore, in the presence of, influenced by nicotine, e.g. by impairment of macrophage e.g. both a7 and a10 nAChR the dominating effect is hard and lymphocyte functions. to predict and may be determined by AChR subunit Clearly, further studies are needed to verify the presented density (remains to be determined) (12). novel hypothesis of HS as a disorder of over-stimulation of 7 Nicotine has been shown to promote inflammatory defined nicotinergic AChR, and to test the new therapeutic reactions, especially in the presence of mast cells and neutr- strategy that is invited by this HS pathogenesis scenario: ophilic granulocytes, both of which have been shown to be the systematic use of anticholinergic agents in the manage- present in early lesions of HS, causing a spongiotic infun- ment of HS. dibulitis (1). Specifically, nicotine provokes mast cell degranulation (14) and enhances chemotaxis and survival Hjalmar Kurzen of neutrophilic granulocytes (15,16). Department of Dermatology, Venerology and Allergology, 8 To terminate an inflammatory reaction induced by Medical Faculty of Mannheim, University of Heidelberg, micro-organisms, cell debris or by ‘foreign’ material (hair Theodor-Kutzer-Ufer 1-3, D-68167 Mannheim, Germany; keratin!) in the chronic phase of HS lesions, both lympho- E-mail: [email protected]; cytes and macrophages secrete and need proinflammatory [email protected] cytokines like TNF-a, IL-4, Il-2 or IL6 (Fig. 1c,d). Nicotine has been shown to suppress the LPS-induced secretion of References these cytokines (10,17). In addition, chemotaxis of macro- 1 Sellheyer K, Krahl D. Int J Dermatol 2005: 44: 535–540. 2 Gao M et al. J Invest Dermatol 2006: 126: 1302–1306. phages is diminished in the presence of cigarette smoke 3 Jansen I et al. J Eur Acad Dermatol Venereol 2001: 15: 532–540. extract, an effect attributed to nicotine (18). 4 Kurzen H et al. Int J Coloproct 2000: 22: 76–80. 9 Nicotine is a highly potent inducer of epidermal 5 Alkondon M et al. Neuropharmacology 2000: 39: 2726–2739. hyperplasia as we could show in an in vitro model. This 6 Kintz P et al. J Chromatogr B: Biomed Sci Appl 705: 357–361. corresponds to the prominent, approximately threefold 7 Balabanova S, Krupienski M. Hautarzt 1995: 46: 255–258. 8 Pavia C S et al. J Med Microbiol 2000: 49: 675–676. increase in epidermal thickness observed in HS specimens 9 Durmaz R et al. New Microbiol 2001: 24: 143–147. (compare Fig. 1a–d). At the same time, the density of 10 Kawashima K et al. Life Sci 2003: 74: 675–696. nAChR is most pronounced at the place where HS patho- 11 Hana A et al. Life Sci 2007: 25: 2214–2220. genesis is thought to originate from: the hair follicle infun- 12 Kurzen H et al. Horm Metab Res 2007: 39: 125–135. dibulum (11). 13 Millar N. Biochem Soc Trans 2003: 31: 869–874. 14 Blandina P et al. J Physiol 1980: 301: 281–293. Altogether, there is overwhelming evidence for a crucial 15 Aoshiba K et al. J Lab Clin Med 1996: 127: 186–194. role of nicotine and the non-neuronal cholinergic system 16 Sorensen L T et al. Surgery. 2004: 136: 1047–1053. in the pathogenesis of HS, providing a molecular mecha- 17 Pavlov V A et al. Mol Med 2003: 8: 125–134. nism especially for early events in HS pathogenesis, which 18 Ortega E et al. Comp Immunol Microbiol Infect Dis 1994: 17: is follicular obstruction and infundibulitis. Late events are 77–84.

ª 2008 The Authors 18 Journal compilation ª 2008 Blackwell Munksgaard, Experimental Dermatology